No title

Editorial

In 2006, WHO’s World Health Report detailed the perilous state of health-care provision in sub-Saharan Africa. The statistics are familiar to anyone with an interest in global health. Africa had only 2·3 health-care workers per 1000 population, compared with 18·9 in Europe, and this workforce had to deal with 24% of the global disease burden, with just 1% of the total global funding for health. Since published, these statistics have formed the introduction to many subsequent publications on health in Africa, but so far, practical solutions have been in short supply. A key problem is training and retention of medical staff, including doctors. The Lancet Commission, Health professionals for a new century, has shown this to be a global problem. A health policy paper from the subSaharan African Medical Schools Study in this week’s issue of The Lancet has laid bare the difficulties involved in educating medical students in sub-Saharan Africa. Fitzhugh Mullan and colleagues’ thorough survey of 105 medical schools details problems in accreditation, postgraduate education, and coordination between ministries of health and education; all these shortcomings contribute to inadequate provision of doctors, especially in rural areas. Absence of faculty and infrastructure, and retention of graduates were the main problems. The paper is a worrying read, but there is a glimmer of hope. On March 7, 2011, 240 people from 30 medical schools in Africa and 20 institutions from the USA met in Johannesburg for the first annual symposium of the Medical Education Partnership Initiative (MEPI). Never before have so many stakeholders come together with the aim of improving medical education in sub-Saharan Africa. MEPI, whose main funding comes from a new initiative of the US President’s Emergency Fund for AIDS Relief (PEPFAR), aims to turn medical education in sub-Saharan Africa around, starting with deploying US$130 million in grants to medical schools in the region. Significantly, all of the awards are made directly to the African institutions in keeping with the Obama administration’s foreign-aid principle of country ownership. 11 5-year medical education focused programmatic grants of up to $2 million per year, and eight grants supported by the US National Institutes of Health have been awarded. Partnerships between African grantees and other medical schools in Africa or the USA will implement this initiative. www.thelancet.com Vol 377 March 26, 2011

Workforce development is the linchpin of MEPI, and retention of graduates is a key aim of the programme. The emigration of African medical graduates has contributed greatly to the continent’s shortage of doctors; in 2005, the International Organization for Migration estimated that within 5 years of graduation around one in five African doctors have migrated to a high-income country. To encourage more graduates to stay where needed MEPI schools will use a combination of factors including the use of community-based education, early exposure to rural practice, creation of clear career paths, and support for regionally relevant research. MEPI’s aims of improving infrastructure and increasing faculty should improve the quantity and quality of graduates. Another boon for health-care provision in Africa came hot on the heels of the MEPI symposium, with the inaugural meeting of the African Society for Laboratory Medicine (ASLM) in Addis Ababa, Ethiopia. Funded by a $4·2 million grant from PEPFAR, ASLM’s aim is to coordinate top quality laboratory services across the continent. In a similar vein to medical education, laboratory services in Africa lack infrastructure and meaningful accreditation. There is also no professional body of qualified staff. But at the end of their meeting, the ASLM delegates were hopeful that they had set out along a path to revitalise laboratory services. The plans to professionalise services, train staff, and have a stepwise and pragmatic approach to accrediting laboratories, should lead to a marked improvement in service provision. This will hopefully break the seemingly never ending downward spiral linking poor laboratory services, with lack of use, and lack of investment. The creation of the African Journal of Laboratory Medicine, promotion of research, and the development of networks to provide global expertise should also help support and retain staff. Although logistical difficulties lie ahead for both MEPI and ASLM, their foundation should be celebrated as efforts to provide African solutions to health care within Africa. Indeed, March, 2011, has the potential to be an inflexion point in the history of African healthcare provision. Hopefully those responsible for change can continue to ride the wave of enthusiasm, which will often not be easy, and work together towards an African future in which good health care is the norm for all. „The Lancet

Sean Sprague/Lineair/Still Pictures

Health care: an African solution

See Health Policy page 1113

For The Lancet Commission see http://www.lancet.com/ education-of-healthprofessionals For WHO’s World Health Report 2006 see http://www.who.int/ whr/2006/whr06_en.pdf (accessed March, 2011) For more on the Medical Education Partnership Initiative see http://www.fic.nih. gov/programs/training_grants/ mepi/index.htm (accessed March 21, 2011) For the International Organization for Migration, World Migration Report 2005 see http://publications.iom.int/ bookstore/index.php?main_ page=product_info&cPath=37& products_id=176 For more on the African Society for Laboratory Medicine see http://www.afslm.org (accessed March 21, 2011) For The Lancet’s special issue on human resources for health see Lancet 2008; 371: 623–96

1047

Editorial

Reuters

Key indicators of health in the USA

For the IOM report see http:// www.iom.edu/~/media/Files/ Report%20Files/2011/LeadingHealth-Indicators-for-HealthyPeople-2020/Leading%20 Health%20Indicators%20 2011%20Report%20Brief.pdf

The US Government’s Healthy People initiative aims to improve the health of Americans. Last week, the Institute of Medicine released Leading Health Indicators for Healthy People , prioritising 12 health indicators and 24 health objectives among 42 topics and nearly 600 objectives. Those health indicators are access to care services, quality of health-care services, healthy behaviours, physical environment, social environment, chronic disease, mental health, injury, maternal and infant health, tobacco use, substance abuse, and responsible sexual behaviour. Several objectives are targeted for reduction: deaths from cancer (563 875 in 2007) and coronary artery disease (one of every six US deaths), tobacco use (one of five preventable deaths), adolescent pregnancies (10% of US births), central-line-associated bloodstream infections (the third most common health-care associated infection), the proportion of adults with hypertension (74·5 million), the proportion of obese children (one in seven lowincome, preschool-aged children), and the proportion of people who have major depression (one of 13).

Increases are planned in objectives focused on health literacy (90 million of US adults are considered health illiterate), the proportion of adults who get sufficient sleep (50–70 million have chronic sleep insufficiency), the proportion of sexually active people who use condoms (use is as low as 17% among white adults), and the proportion of adults who are physically active (10% are not active at all and 60% do not exercise regularly). Objectives for lesbian, gay, bisexual, and transgender health are addressed for the first time in the report. They include an increase in their health insurance coverage (they are twice as likely to be uninsured when compared with a heterosexual partner) and a reduction in obesity (lesbians and bisexual females are more likely to be overweight). The report’s objectives must be implemented and work together with the Patient Protection and Affordable Care Act of 2010 to improve Americans’ access to health-care services and the quality of health care. This will require coordinated interplay of health players at the federal, state, and local level. „The Lancet

Science Photo Library

Managing acute upper gastrointestinal bleeding

For the toolkit see http://aomrc. org.uk/projects/uppergastrointestinal-bleedingtoolkit.html

1048

On March 15, a new toolkit (commissioned by the UK’s National Patient Safety Agency) for service planning for acute upper gastrointestinal bleeding was released. More than 50 000 patients in the UK have such a bleed each year—mortality is about 10%. The report highlights that out-of-hours care is a particular problem, and that patients are not always seen by appropriately trained staff, the necessary equipment is not available, or that patients are not transferred to a specialist unit. Treatment aims to stop the bleed and reduce risk of rebleeding. If an upper gastrointestinal bleed with blood loss is suspected, the patient should be stabilised before diagnostic endoscopy; the toolkit recommends that urgent-risk patients undergo endoscopy within 6–12 h. Depending on the site and type of vessel, endoscopic management can be mechanical (eg, banding, clipping) or by thermal or electrical coagulation, usually followed by local injection of epinephrine or a sclerosant. Radiological interventions should also be readily available. For rebleeding that cannot be stopped by repeat endoscopy, radiological intervention might be needed to

localise the site of bleeding and to continue treatment. In severe cases, surgery might be needed to stop the bleed. If the patient is in a high-risk category for surgery, radiological intervention is the preferred approach. The toolkit recommends that these interventions must be available 24 h a day, 7 days a week. The patients often at risk include those with a peptic ulcer. Patients with chronic liver disease are also at high risk of bleeding because of complications due to portal hypertension. Patients with other comorbidities, particularly the elderly population with cardiovascular or lung disease, for example, are at high risk of bleeding complications and death, because the upper gastrointestinal bleed can worsen the underlying condition by decreasing haemodynamic stability. At present, the mortality rate of acute upper gastrointestinal bleeding, especially out of hours, is unacceptable. The toolkit should bring management to the fore and help hospitals to institute appropriate management pathways and care for this acute medical emergency. „The Lancet www.thelancet.com Vol 377 March 26, 2011

Comment

Acute MI: triple-markers resurrected or Bayesian dice?

www.thelancet.com Vol 377 March 26, 2011

would have identified 9·8% of patients as eligible for early discharge, 0·9% of whom went on to have a major adverse cardiac event within 30 days. Use of the protocol could potentially have saved 1–2 hospital bed-days per low-risk patient. ASPECT was well designed to achieve its objectives, and shows that it is possible to achieve an acceptably high sensitivity when triple-marker testing is used in the appropriate population. However, that selection of the appropriate population was pivotal to the success of the accelerated discharge protocol. Triple-marker testing alone had a relatively low sensitivity, at just 82·9%. The overall sensitivity of the protocol increased to an acceptable level because of the application of Bayesian principles, with biomarker testing only in patients with a low pretest probability of disease (ie, patients with a TIMI risk score of 0 and a normal ECG). Thus, predicting just over 80% of major adverse cardiac events in an already low-risk population yields an even lower net risk. Most people will probably consider this net risk to be statistically acceptable. However, if properly informed, low-risk patients might feel differently about the relative merits of waiting for definitive 6-h laboratorybased troponin testing or going home after 2 h on the basis of results from a test that correctly identifies serious coronary disease, when present, in just over eight of ten occasions. This issue is particularly pertinent in view of the recent development of highly

Published Online March 23, 2011 DOI:10.1016/S01406736(11)60392-9 See Articles page 1077

Science Photo Library

Acute coronary syndromes are the acute manifestations of a disease that will ultimately kill around one in six people,1 a disease that has been feared for centuries and is still revered by physicians. The disease can kill instantly, yet the symptoms and signs alone simply cannot be relied on to differentiate an acute coronary syndrome from much less threatening disorders.2 Even William Osler, one of the most esteemed diagnosticians in history, said: “One must be a professional Ulysses in craft and wisdom not sometimes to err in estimating the nature of an attack of severe heart pain. There is no group of cases so calculated to keep one in a condition of wholesome humility.”3 Most people who seek emergency medical attention for symptoms compatible with an acute coronary syndrome do not actually have the syndrome. We do, however, invest substantial time and money establishing that through diagnostic investigations. These investigations usually mandate hospital admission, meaning that such patients account for over a quarter of all acute medical admissions.4 The need for an effective rapid rule-out strategy to facilitate early discharge from the emergency department has been appreciated for over 20 years. Despite extensive research, however, none has been widely adopted. One potential strategy that has gained considerable interest over the past decade is triple-marker testing. Creatine kinase-MB fraction and myoglobin rise early after the onset of infarction, while the rise in troponin is late and sustained. In theory, the strategy should detect infarction in patients who present both early and late after symptom onset. However, some of the studies reporting high sensitivities and negative predictive values had important verification bias,5 while other studies had inadequate sensitivity.6 In The Lancet, Martin Than and colleagues7 report the ASia-Pacific Evaluation of Chest pain Trial (ASPECT), a multinational prospective diagnostic cohort study. The study, which included 3582 patients, investigated the diagnostic accuracy of an accelerated diagnostic protocol that would enable early discharge for patients who met the predefined criteria of a Thrombolysis In Myocardial Infarction (TIMI) risk score of 0 (out of 7), no ischaemic ECG changes, and normal point-of-care triplemarker panels at presentation and 2 h later. This strategy

1049

Comment

sensitive troponin assays, some of which can have a sensitivity of around 90% (for acute myocardial infarction) at the time of presentation and possibly up to 100% within 3 h,8,9 and by research into other promising biomarkers such as heart fatty acid binding protein and copeptin.10,11 It therefore remains important that triple-marker testing is compared with some of these more recent alternatives. Now, more than ever, it will also be important to compare the relative merits of point-of-care testing with laboratory-based assays that have much higher analytical sensitivity and precision. Finally, the recent Randomised Assessment of Triage with Panel Assay of Cardiac markers (RATPAC) trial12 showed that, although triple-marker testing increased the proportion of patients successfully discharged from the emergency department and reduced the median length of initial hospital stay, such testing was also associated with increased mean length of hospital stay and greater use of coronary care, which might be a function of the low specificity and positive predictive value of the biomarker panel. The findings of a costeffectiveness analysis are expected shortly.13 In the ASPECT trial, the biomarkers alone had a positive predictive value of only 20·1%. Although it is not imperative that the overall specificity of the accelerated diagnostic protocol is high (ultimately, the specificity of 10% still potentially means that 10% of “healthy” patients are eligible for early discharge when they would otherwise have been admitted), the low specificity of the biomarker panel (56% in ASPECT) might be more of a problem. It could be harder for clinicians to ignore increases in biomarkers that supposedly indicate myocardial necrosis, thus prompting over-treatment and over-investigation. Ultimately, ASPECT has successfully established that an accelerated diagnostic protocol incorporating triple-marker testing successfully identifies a group of patients at very low risk of major adverse cardiac events who could reasonably be considered for early discharge. The field must now ask whether the strategy defined is indeed optimal, whether more sensitive and specific assays might improve performance, and whether these promising data will stand up

1050

to subsequent analyses of cost-effectiveness and patients’ preference. Richard Body Cardiovascular Sciences Research Group, University of Manchester, Manchester M13 9WL, UK [email protected] I have attended two advisory group meetings for Roche Diagnostics (no fees were paid and travel expenses were not claimed). I have done research supported by collaborative agreements with Alere Diagnostics (involving free transport and testing of samples), Roche Diagnostics (involving donation of reagents for serum testing), and Randox Diagnostics (involving loan of equipment and donation of reagents to test plasma samples). Siemens Diagnostics will donate reagents for a research project I am leading. I have received honoraria for speaking engagements with Bristol-Myers Squibb and PASTEST, and have spoken at meetings sponsored by Roche Diagnostics and Randox Diagnostics (no honoraria). I received an honorarium for assisting Bristol-Myers Squibb to prepare educational presentations. Roche Diagnostics and Randox Diagnostics have arranged travel and accommodation for presentations at company-sponsored symposia in Europe (with Randox Diagnostics, this is still pending). I attended a lecture and subsequent meal sponsored by Brahms Diagnostics. 1 2 3 4 5

6

7

8 9 10 11 12

13

Allender S, Peto V, Scarborough P, Kaur A, Rayner M. Coronary heart disease statistics. July, 2008. http://www.ws3.heartstats.web.baigent.net/ datapage.asp?id=7998 (accessed March 17, 2011). Body R, Carley S, Wibberley C, McDowell G, Ferguson J, Mackway-Jones K. The value of symptoms and signs in the emergent diagnosis of acute coronary syndromes. Resuscitation 2010; 81: 281–86. Silverman ME, Murray TJ, Bryan CS, eds. The quotable Osler. Revised edn. Philadelphia: American College of Physicians, 2008. Goodacre S, Cross E, Arnold J, Angelini K, Capewell S, Nicholl J. The health care burden of acute chest pain. Heart 2005; 91: 229–30. Straface A, Myers J, Kirchick H, Blick K. A rapid point-of-care cardiac marker testing strategy facilitates the rapid diagnosis and management of chest pain patients in the emergency department. Am J Clin Pathol 2008; 129: 788–95. Brown AM, Sease KL, Robey JL, Shofer FS, Hollander JE. The impact of B-type natriuretic peptide in addition to troponin I, creatine kinase-MB, and myoglobin on the risk stratification of Emergency Department chest pain patients with potential acute coronary syndrome. Ann Emerg Med 2007; 49: 153–63. Than M, Cullen L, Reid CM, et al. A 2-h diagnostic protocol to assess patients with chest pain symptoms in the Asia-Pacific region (ASPECT): a prospective observational validation study. Lancet 2011; published online March 23. DOI:10.1016/S0140-6736(11)60310-3. Reichlin T, Hochholzer W, Bassetti S, et al. Early diagnosis of myocardial infarction with sensitive cardiac troponin assays. N Engl J Med 2009; 361: 858–67. Keller T, Zeller T, Peetz D, et al. Sensitive troponin I assay in early diagnosis of acute myocardial infarction. N Engl J Med 2009; 361: 868–77. Keller T, Tzikas S, Zeller T, et al. Copeptin improves early diagnosis of acute myocardial infarction. J Am Coll Cardiol 2010; 55: 2096–106. Body R, Newbury M. Heart fatty acid binding protein for rapid diagnosis of acute myocardial infarction in the emergency department. Emerg Med J 2009; 26: 519–22. Goodacre S, Bradburn M, Cross E, Collinson P, Gray A, Hall A. The randomised assessment of treatment using panel assay of cardiac markers (RATPAC) trial: a randomised controlled trial of point-of-care cardiac markers in the emergency department. Heart 2010; 97: 190–96. Fitzgerald P, Goodacre S, Cross E, Dixon S, on behalf of the RATPAC investigators. Cost-effectiveness of point-of-care biomarker assessment for suspected myocardial infarction: The RATPAC trial (Randomised Assessment of Treatment Using Panel Assay of Cardiac markers). Acad Emerg Med (in press).

www.thelancet.com Vol 377 March 26, 2011

Comment

Size still matters…but not in the way we once thought

www.thelancet.com Vol 377 March 26, 2011

weight are needed, whereas accurate measurement of waist and hip circumferences needs some degree of training and can be especially difficult in individuals who are obese. Second, the analysis indicated that most of the cardiovascular risk associated with being overweight is mediated by deleterious effect on other risk factors— namely blood pressure, history of diabetes, and total and HDL cholesterol. After adjustment for these risk factors, the strength of association between body size and cardiovascular disease was substantially attenuated. Thus the addition of body size to risk-prediction models (that include intermediary factors) did not improve their accuracy. Finally, the study dispelled previous hope that assessment of body size could replace the cost, time, and inconvenience of blood lipids assay. It had been suggested that knowledge attained from an individual’s age, sex, blood pressure, history of diabetes, smoking status, and a measure of body size would be equivalent to the information obtained from a blood test.8,9 Such an option could be important, particularly in resource-poor settings. However, findings of the Emerging Risk Factors Collaboration’s study showed that inclusion of BMI and waist circumference could only marginally compensate for the lack of information about total and HDL cholesterol in a risk-prediction model. Despite the extensive analyses by the Collaboration, important questions remain, such as to what extent

Published Online March 11, 2011 DOI:10.1016/S01406736(11)60239-0 See Articles page 1085

Science Photo Library

In recent months, three large-scale consortia have published studies about the relation between bodymass index (BMI) and mortality in high-income populations.1–3 The size of these studies has enabled quantification of the J-shaped relation between BMI and mortality from cardiovascular disease, cancer, and all causes in diverse populations, more reliably than was possible before. In The Lancet, a fourth report4 adds to the findings of these studies. This new analysis, by the Emerging Risk Factors Collaboration, used individual records from 58 (predominantly European and US) cohorts. The information comes from more than 220 000 individuals and more than 14 000 cardiovascular events. Importantly, what distinguishes this Collaboration from the Prospective Studies Collaboration,1 the National Cancer Institute Cohort Consortium,2 and the European Prospective Investigation into Cancer3 is the scope of the adiposity information. However, all four studies had information about measures of central obesity in addition to BMI and a large number of covariates. This information allowed, for the first time, a comparison of the usefulness of these measures for prediction of subsequent cardiovascular risk, when the variables were analysed separately or in combination. The analysis by the Emerging Risk Factors Collaboration had three key findings. First, BMI, waist circumference, and waist-to-hip ratio had a similar strength of association with coronary heart disease and ischaemic stroke. For every increment of one standard deviation, in each of the three measures, the risk of a cardiovascular event was 25–30% higher after adjustment for age, sex, and smoking status. This finding contradicts previous hypotheses that measures of central obesity are more strongly associated than are global indices (such as BMI) with subsequent cardiovascular risk,5,6 despite evidence that visceral fat is more metabolically active than is fat in other depots (such as subcutaneous fat).7 The finding also effectively argues against the idea that waist circumference or waist-to-hip ratio should replace BMI as the anthropometric measure of choice in clinical practice, at least for cardiovascular risk assessment.4 Maintaining the status quo has its advantages; BMI is the easiest of the three units to measure because only height and

1051

Comment

does a specific middle-age risk factor represent the status of that factor long before and after its measurement? The answer might be less of an issue with BMI than for the intermediate risk factors that might develop. This issue is especially pertinent in children and young adults with high BMI because they are more likely to develop risk factors for cardiovascular disease by middle age than are those with low BMI. Moreover, because the relations between body anthropometry and risk factors for cardiovascular disease were cross-sectional, additional consideration of changes in body size might be relevant for prediction of subsequent risk. Do the new data from the Emerging Risk Factors Collaboration mean that we should stop measuring BMI? On the contrary, BMI continues to be useful as an indicator of adiposity, despite its obvious and occasional misrepresentation of muscular people and lack of sensitivity to body shape and composition. BMI used with good clinical judgment is highly appropriate in adults because it is so strongly associated with chronic disease risk, although we caution that it is correlated with height in children.10 Many overweight or obese adolescents, young adults, and middle-aged individuals with few risk factors for cardiovascular disease will develop that risk relatively soon, so BMI should serve as an early warning, both to them and their general practitioners. But identification of which overweight individuals without risk factors for cardiovascular disease will go on to develop those risk factors, and ultimately clinical cardiovascular disease,

remains a challenge—here, blood tests continue to be helpful. *Rachel R Huxley, David R Jacobs Jr Division of Epidemiology and Community Health, School of Public Health, University of Minnesota, Minneapolis, MN 55454, USA (RRH, DRJ); and The George Institute for Global Health, University of Sydney, Sydney, NSW, Australia (RRH) [email protected] We declare that we have no conflicts of interest. 1 2 3 4

5

6 7 8 9

10

Prospective Studies Collaboration. Body-mass index and cause-specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies. Lancet 2009; 373: 1083–96. Berrington de Gonzalez A, Hartge P, Cerhand JR, et al. Body-mass index and mortality among 1·46 million white adults. N Engl J Med 2010; 363: 2211–19. Pischon T, Boeing H, Hoffmann K, et al. General and abdominal adiposity and risk of death in Europe. N Engl J Med 2008; 359: 2105–20. The Emerging Risk Factors Collaboration. Separate and combined associations of body-mass index and abdominal adiposity with cardiovascular disease: collaborative analysis of 58 prospective studies. Lancet 2011; published online March 11. DOI:10.1016/S01406736(11)60105-0. Yusuf S, Hawken S, Ounpuu S, et al, on behalf of the INTERHEART Study Investigators. Obesity and the risk of myocardial infarction in 27 000 participants from 52 countries: a case-control study. Lancet 2005; 366: 1640–49. Schneider HJ, Friedrich N, Klotsche J, et al. The predictive value of different measures of obesity for incident cardiovascular events and mortality. J Clin Endocrinol Metab 2010; 95: 1777–85. Snijder MB, van Dam RM, Visser M, Seidell JC. What aspects of body fat are particularly hazardous and how do we measure them? Int J Epidemiol 2006; 35: 83–92. Lim SS, Gaziano TA, Gakidou E, et al. Prevention of cardiovascular disease in high-risk individuals in low-income and middle-income countries: health effects and costs. Lancet 2007; 370: 2054–62. WHO. Prevention of cardiovascular disease: guidelines for the assessment and management of total cardiovascular risk. 2007. http://www.who.int/ cardiovascular_diseases/guidelines/PocketGL.ENGLISH.AFR-D-E.rev1.pdf (accessed Feb 18, 2011). Cole TJ. Weight/heightp compared to weight/height² for assessing adiposity in childhood: influence of age and bone age on p during puberty. Ann Hum Biol 1986; 13: 433–51.

Tranexamic acid for trauma Published Online March 24, 2011 DOI:10.1016/S01406736(11)60396-6 See Correspondence page 1071 See Articles page 1096

1052

After its publication in July, 2010, the CRASH-2 study1 generated widespread interest in the early administration of the antifibrinolytic agent tranexamic acid to patients with traumatic bleeding. Tranexamic acid is an inexpensive, easily used, and relatively safe drug, and it seemed to have saved lives. However, how it did so was unclear—the blood-transfusion requirements of the tranexamic acid and placebo groups were similar and, survival bias notwithstanding, the mortality benefit might have been attributable to an effect of tranexamic acid on something other than acute traumatic coagulopathy.2

This issue is partly addressed with the publication in The Lancet of a follow-up analysis that used the outcome of death due to bleeding rather than all-cause mortality.3 The CRASH-2 collaborators3 report a 32% reduction in death due to bleeding when tranexamic acid is given within 1 h of injury. Although markers of coagulopathy were not measured, the mortality benefit is probably mediated through antifibrinolytic effects on clot stabilisation.4 While it will not prevent the massive haemorrhage from disrupted vessels or organs that needs surgical intervention, tranexamic acid appears to improve survival through its effect on mild to moderate bleeding. www.thelancet.com Vol 377 March 26, 2011

Early administration is necessary, however, and benefit was only seen in CRASH-2 when tranexamic acid was administered within 3 h of injury. Unlike coagulopathy that is secondary to haemodilution, hypothermia, or acidosis, acute traumatic coagulopathy is a hyperacute process in which systemic fibrinolysis releases D-dimers that are detectable within 30 min of injury.5 While the mechanisms are poorly understood, shock and tissue injury seem to be important initiators.6 Not all severely injured patients develop acute coagulopathy, but those who do are much more likely to die and to die early.7 The earlier that tranexamic acid is administered, the more likely it might be to prevent full activation of fibrinolysis. Once fully activated, fibrinolysis has been shown to continue unabated until endogenous antifibrinolytic elements are restored.8 Importantly, the CRASH-2 collaborators3 report increased mortality due to bleeding in patients receiving tranexamic acid when it is given more than 3 h after injury. The cause of these deaths is unclear. Reports exist of prothrombotic effects of each of the anti-fibrinolytic drugs. Alternatively, it might reflect some factor of the patients who received it late. Whatever the mechanism, the CRASH-2 collaborators3 have cautioned against the use of tranexamic acid when more than 3 h have expired after injury. Who, then, should be treated with tranexamic acid? Most of the 274 study sites in CRASH-2 were in low-income and middle-income countries, where other treatments directed at coagulopathy, such as fresh frozen plasma, platelets, and cryoprecipitate, are less available. Although many patients with acute coagulopathy will die before reaching hospital, tranexamic acid is a practical, affordable, and effective treatment for bleeding trauma patients in such centres, provided they receive it within 3 h of injury. Far less clear is the place for tranexamic acid in high income countries where massive transfusion protocols incorporate fresh-frozen plasma that contains all the endogenous antifibrinolytic elements in plasma.9 Plasma can cause harm as well as benefit, and there is little prospective evidence regarding its efficacy. However, because it is in widespread use, and because late administration of tranexamic acid can be harmful, it is unlikely that many clinicians in major trauma centres will choose tranexamic acid as firstline treatment. www.thelancet.com Vol 377 March 26, 2011

Science Photo Library

Comment

The best place for tranexamic acid in developed trauma systems might actually be in the prehospital environment. Helicopter and road transport direct to major trauma centres has reduced overall injury mortality, but has extended the time before patients reach hospital.10 Prehospital administration of blood products, especially plasma, is uncommon in civilian settings, resulting in little directed management of coagulopathy. By contrast, tranexamic acid can be safely stored in vehicles and simply administered. In view of the new findings from CRASH-2, the best outcomes might be achieved with simple measures for haemorrhage control and early inhibition of coagulopathy with tranexamic acid, followed by rapid transport for surgery or angiography and tailored management of coagulopathy in hospital. CRASH-2 was an extraordinary achievement, with randomisation of more than 20 000 patients in 40 countries. It has established tranexamic acid as an effective hospital-based treatment for traumatic haemorrhage, provided that the drug is given within 3 h of injury. In trauma systems that have advanced prehospital services and that use other hospital-based treatments for coagulopathy, CRASH-2 raises more questions—and more possibilities—that are worth investigating. *Russell L Gruen, Biswadev Mitra The National Trauma Research Institute (RLG) and Emergency and Trauma Centre (BM), The Alfred Hospital, Monash University, Melbourne, VIC 3004, Australia [email protected]

1053

Comment

We declare that we have no conflicts of interest. 1

2 3

4 5

CRASH-2 trial collaborators. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet 2010; 376: 23–32. Levy J. Antifibrinolytic therapy: new data and new concepts. Lancet 2010; 376: 3–4. The CRASH-2 collaborators. The importance of early treatment with tranexamic acid in bleeding trauma patients: an exploratory analysis of the CRASH-2 randomised controlled trial. Lancet 2011; published online March 24. DOI:10.1016/S0140-6736(11)60278-X. Paran H, Gutman M, Mayo A. The effect of aprotinin in a model of uncontrolled hemorrhagic shock. Am J Surg 2005; 190: 463–66. Davenport R, Curry N, Manson J, De’Ath H, Coates A, Rourke C. Hemostatic effects of fresh frozen plasma may be maximal at red cell ratios of 1:2. J Trauma 2011; 70: 90–95.

6 7 8

9 10

Hess J, Brohi K, Dutton R, Hauser C, Holcomb J, Kluger Y. The coagulopathy of trauma: a review of mechanisms. J Trauma 2008; 65: 748–54. Mitra B, Cameron P, Mori A, Fitzgerald M. Acute coagulopathy and early deaths post major trauma. Injury 2010; published online Dec 7. DOI:10.1016/j.injury.2010.10.015. Bolliger D, Szlam F, Levy J, Molinaro R, Tanaka K. Haemodilution-induced profibrinolytic state is mitigated by fresh-frozen plasma: implications for early haemostatic intervention in massive haemorrhage. Br J Anaesth 2010; 104: 318–25. Bolliger D, Görlinger K, Tanaka K. Pathophysiology and treatment of coagulopathy in massive hemorrhage and hemodilution. Anesthiology 2010; 11: 1205–19. Lockey D, Deakins C. Pre-hospital trauma care: systems and delivery. Contin Educ Anaesth Crit Care Pain 2005; 5: 191–94.

Disasters and a register for foreign medical teams

Getty Images

The tsunami in Asia and earthquakes in Pakistan, Iran, Indonesia, and Haiti at the start of the 21st century, and now the earthquake and tsunami in Japan, have emphasised people’s continuing willingness to respond to large-scale disasters overseas. However, these individual acts of altruism are tempered by criticisms about lack of preparedness, coordination, and appropriate skills.1–3 In the UK, many of these volunteers work in the National Health Service (NHS), and a sudden exodus of highly skilled staff can put considerable strain on their institutions. To address these issues, a formal register of UK surgeons, anaesthetists, emergency physicians and nurses, and other supporting medical, nursing, and paramedical staff has been established.4

Patient is transferred out of recovery room the day after surgery at the Israeli army hospital on Jan 19, 2010 in Port-au-Prince, Haiti

1054

The register has been developed with the UK Government’s Department of Health and Department for International Development, and with non-governmental organisations including Medical Emergency Relief International (Merlin). Although the idea to create a register has been considered after each major event,5 only now has sufficient momentum been gathered to see its implementation. Governments, non-governmental organisations, and UN agencies can select health-care workers from the register and be assured that they are ready to go and are fit for purpose. The register is supported by the Faculty of Pre-Hospital Care of the Royal College of Surgeons of Edinburgh, the Royal College of Surgeons of England, the UK College of Emergency Medicine, the British Association of Immediate Care Schemes, the British Association of Plastic Reconstructive and Aesthetic Surgeons, the Academy of Medical Royal Colleges in the UK, the UK’s Royal College of Nursing, and the British Medical Association. Existing collaboration with other countries will be strengthened. An important role of the register will be to foster training. The core competencies for deployment to acute surgical emergencies (most noticeably earthquakes) are probably more easily identified and agreed on than are those for longer-term development work. Therefore the register will concentrate its remit on these emergencies. This focus also chimes with WHO’s initiative to update its guidelines for the use of foreign field-hospitals. After experts met in Cuba in December, 2010, an ad-hoc working group (chaired by ADR) was formed to explore registration of foreign medical teams, both before and after deployment to sudden-onset disasters. www.thelancet.com Vol 377 March 26, 2011

Comment

There will be liaison with the employers of volunteers, particularly the NHS, to define their commitment and to ensure that when teams are despatched they come from various sites to minimise disruption. As yet, no upfront external funding is attached to this initiative, and those who wish to be deployed will be volunteers and hopefully seconded (and paid) by their employer. This system has already worked well in the NHS for past emergencies overseas. Those who wish to be supportive but not deployed overseas can contribute equally by offering to provide voluntary cover for their colleagues. To test the ongoing willingness and capacity of institutions and individuals to enter into such voluntary arrangements, deployment to specific surgical emergencies also provides a clearly defined commitment: 24–48 h notice for deployment for 2–3 weeks is being requested. How the register performs in this specialist and timelimited framework could help to develop the more complex arrangements that are needed for longer-term secondments overseas and that are now supported in principle by government.6 In tandem with, and complementary to, the trauma register is the International Health Links Centre’s Humanitarian Response Register,7 which is designed to assist organisations in identifying suitable volunteers when planning their medium-term and long-term responses, and thus to bridge the gap between the immediate response to sudden-onset natural disasters and the sustained response needed for protracted emergencies. Awareness has increased of the need for accountability and standards when aid is offered to another country,8 particularly in a severe emergency when the usual checks and balances for the individual’s appropriateness are weakened or absent. This awareness is particularly important in medicine when interventions might be life-saving or lifechanging. Many amputations might have been done

in Haiti immediately after the earthquake,9 but few data exist.10 How can we improve the response when the reasons for surgery are largely undocumented and therefore unknown? Emergency humanitarian medical assistance is only part of medical practice and therefore needs training, accreditation, and accountability. The UK International Emergency Trauma Register is a collective effort to support these aims and those of the health-care workers who will help their patients in the time of greatest need. *Anthony D Redmond, Timothy J O’Dempsey, Bertrand Taithe Humanitarian and Conflict Response Institute, University of Manchester, Manchester M13 9PL, UK (ADR, BT); and Liverpool School of Tropical Medicine, University of Liverpool, Liverpool, UK (TJO’D) [email protected] We declare that we have no conflicts of interest. 1 2 3 4 5 6

7 8 9

10

Abolghasemi H, Radfar MH, Khatami M, Nia MS, Amid A, Briggs SM. International medical response to a natural disaster: lessons learned from the Bam earthquake experience. Prehosp Disast Med 2006; 21: 141–17. Brennan RJ, Waldman RJ. The south Asian earthquake six months later— an ongoing crisis. N Engl J Med 2006; 354: 1769–71. Krin CS, Giannou C, Seppelt IM, et al. Appropriate response to humanitarian crises. BMJ 2010; 340: c562. UK-MED. UK international emergency trauma register database. Dec 27, 2010. http://www.uk-med.org/trauma.html (accessed Feb 27, 2011). Redmond AD. Medical response to disasters overseas. BMJ 1992; 304: 653. Department of Health. Global health partnerships: the UK response to health in developing countries. The Government response. March 13, 2008. http://www.dh.gov.uk/prod_consum_dh/groups/dh_digitalassets/@dh/@ en/documents/digitalasset/dh_083510.pdf (accessed Feb 27, 2011). International Health Links Centre. Promoting effective health partnerships worldwide. Jan 18, 2011. http://www.ihlc.org.uk/news/index.htm (accessed Feb 27, 2011). McQueen KA, Hyder JA, Taira BR, Semer N, Burkle FM Jr, Casey KM. The provision of surgical care by international organizations in developing countries: a preliminary report. World J Surg 2010; 34: 397–402. Pan American Health Organization, WHO. Earthquake in Haiti: PAHO/WHO situation report on health activities post earthquake. May 18, 2010. http:// www.who.int/hac/crises/hti/haiti_health_cluster_bulletin_18may2010.pdf (accessed Feb 27, 2011). Peranteau WH, Havens JM, Harrington S, Gates JD. Re-establishing surgical care at Port-au-Prince General Hospital, Haiti. J Am Coll Surg 2010; 211: 126–30.

Research to achieve health care for all in India Well planned health research is fundamental to the improvement of health in all countries.1 Relevant health research has contributed to a doubling of life expectancy in India over the past 60 years since independence. However, India still has the largest disease burden of any country,2 the mitigation of www.thelancet.com Vol 377 March 26, 2011

which will require existing gaps in health research to be addressed. The national health policy of India has strengthening of health research as one of its aims.3 On the basis of an understanding of how the research produced in India relates to the disease burden and health-system priorities, who is producing this

Published Online January 12, 2011 DOI:10.1016/S01406736(10)62034-X See Series Lancet 2011; 377: 252, 332, 413, 505, 587, 668, and 760

1055

Comment

For published papers in PubMed see http://www.ncbi. nlm.nih.gov/pubmed

research, and what the recent major health research initiatives are, we suggest here how policy and funding should respond to enhance research in order to achieve health care for all in India. The proportion of published papers from India in PubMed increased from 0·4% of the global total in 1988 to 1·8% in 2008. However, the proportion of public health research, which is crucial for achieving health care for all in India, continues to be small, at 5% of the total health research published.4 Additionally, many of the leading causes of disease burden across communicable diseases, non-communicable diseases, and injuries continue to be under-represented in this published research output, indicating that even among the limited papers on public health research, a large proportion do not address public health priority conditions in India.4 Distinct from published papers, an analysis of public health research reports produced in India also showed that the leading chronic non-communicable diseases and injuries were under-represented between 2001 and 2008.4 Health-system research was more evident in public health research reports than in published public health papers. However, the scarcity of Public health research reports 2001–08 Public health research papers 2007 Basic science research papers 2007

For-profit private organisations

Not-for-profit non-governmental research institutions Non-medical technology institutions Agencies of government ministries other than health

Agencies of health ministry

University departments other than medical

Medical academic institutions 0

10

20

30 40 Reports or papers (%)

50

60

Figure: Institutions producing public health and basic science research in India Health ministry agencies include Indian Council of Medical Research network, and agencies of other government ministries include networks of Council of Scientific and Industrial Research and Indian Council of Social Science Research. Public health research reports 2001–08 include those in the public domain on the internet, and papers from 2007 include those in PubMed.4

1056

research about the routine health-information system in both reports and in published papers is concerning because it is crucial to track the response of the health system to the health needs of the population.4–6 Only one in four public health research reports were rated as being of adequate quality. The quality was higher for reports produced by collaborations between Indian and international organisations, indicating that there is merit in promoting such collaborations for more useful research output. An analysis has shown that evaluations of populationlevel health interventions and policies in India are commissioned mostly by international organisations, suggesting that governmental organisations in India should pay more attention to commissioning evaluations of key interventions and policies.7 The study design and analytical approach were generally poor in these evaluations, a weakness that should be urgently addressed. Although most clinical research in India is done by academic medical institutions, the organisations producing basic science and public health research are varied (figure).4 Efforts to enhance health research in India would have to involve this variety of stakeholders. The Indian Council of Medical Research has been the nodal organisation for health research in India since 1949. In recognition of the need for wider coordination in health research between various sectors, the Department of Health Research has been established under India’s Ministry of Health, which includes the Indian Council of Medical Research.8 This department aims to: encourage innovation related to diagnostics, treatment methods, and prevention; translate innovations into products or processes by facilitating their evaluation; and introduce innovation through health-systems research. In addition to important ongoing efforts to enhance public health training and research in India (eg, the National Institute of Epidemiology and some universities), a major new initiative that aims to strengthen training and research capacity relevant for health programmes and policies is the Public Health Foundation of India, which was launched by the country’s Prime Minister.9 Another promising new initiative is the Biomedical Research Career Programme for India, jointly funded by the Indian Government’s Department of Biotechnology and the Wellcome Trust,10 which is planning to increase www.thelancet.com Vol 377 March 26, 2011

Comment

its emphasis on public health research. Although a full analysis of all present funding for health research in India and what it is spent on is not available, the funding from both domestic and international sources has increased substantially in India over the past decade.4 What then are the key goals that policy should address to boost research towards health care for all in India? First, a national research tracking mechanism should be developed to guide funding and commissioning of highquality research into the major under-represented causes of disease burden and into neglected health-system issues. Second, a systematic plan is needed to make research initiatives more interactive with policies and implementation of health programmes, so that research is more relevant for the health system and policy, and the knowledge generated is used more often by policy makers. Third, rigorous evaluation research should become an essential component of all major population health programmes and policies, to understand how these could be refined to improve health outcomes and how the underserved segments of the Indian population could be better reached to improve health equity. For these goals to be achieved, the major national organisations attempting to strengthen health research in India should come together to provide effective stewardship. These organisations should collaboratively develop mechanisms that enable agreement on tangible medium-term and long-term targets for health research in the country, a plan of action, and methods to track progress in research utilisation to achieve health care for all in India. Although solutions for India will have to be tailored to its circumstances, there are useful lessons to be learnt from the systematic efforts of other countries aimed at matching research with public health priorities to more effectively improve population health.11,12

*Lalit Dandona, V M Katoch, Rakhi Dandona Public Health Foundation of India, New Delhi 110070, India (LD, RD); Institute for Health Metrics and Evaluation, University of Washington, Seattle, Washington, USA (LD); and Department of Health Research and Indian Council of Medical Research, Ministry of Health and Family Welfare, Government of India, New Delhi, India (VMK) [email protected] We declare that we have no conflicts of interest. 1 2 3 4 5

6 7

8 9 10 11

12

WHO. World report on knowledge for better health. 2004. http://www. who.int/rpc/meetings/world_report_on_knowledge_for_better_health. pdf (accessed June 20, 2010). WHO. The global burden of disease: 2004 update. 2008. http://www.who. int/healthinfo/global_burden_disease/GBD_report_2004update_full.pdf (accessed June 20, 2010). Ministry of Health and Family Welfare, Government of India. National health policy. 2002. http://mohfw.nic.in/np2002.htm (accessed June 20, 2010). Dandona L, Raban MZ, Guggilla RK, Bhatnagar A, Dandona R. Trends of public health research output from India during 2001–2008. BMC Med 2009; 7: 59. Health Metrics Network. Framework and standards for country health information systems. 2008. http://www.who.int/ healthmetrics/documents/hmn_framework200803.pdf (accessed June 20, 2010). Raban MZ, Dandona R, Dandona L. Essential health information available for India in the public domain on the internet. BMC Public Health 2009; 9: 208. Dandona L, Raban MZ, Dandona R. Composite analysis of evaluations of health system/policy interventions in India: 2001–2008. First Global Symposium on Health Systems Research; Montreux, Switzerland; Nov 16–19, 2010. Department of Health Research, Ministry of Health and Family Welfare, Government of India. Mandate. http://www.dhr.gov.in/madate.htm (accessed June 20, 2010). Reddy KS. Boosting public health capacity in India. Natl Med J India 2006; 19: 122–25. Wellcome Trust–DBT India Alliance. Biomedical Research Career Programme for India. http://www.wellcomedbt.org (accessed June 20, 2010). Canadian Institutes of Health Research. Charting the course: a pan-Canadian consultation on population and public health priorities. 2002. http://www.cihr-irsc.gc.ca/e/documents/charting_the_course_e.pdf (accessed June 20, 2010). National Health and Medical Research Council, Australian Government. Report of the review of public health research funding in Australia. 2009. http://www.nhmrc.gov.au/_files_nhmrc/file/research/phr/Nutbeam.pdf (accessed June 20, 2010).

The case for a global rare-diseases registry Rare diseases are a clinically heterogeneous group of about 6500 disorders,1 and in fewer than 200 000 individuals in the USA.2 They are commonly diagnosed during childhood, often inherited, and can have deleterious long-term effects. Although any one condition is rare, their cumulative public health burden is substantial, with 6–8% of people having a rare disease at some point during life.3 www.thelancet.com Vol 377 March 26, 2011

Because of the rarity, no single institution, and in many cases no single country, has sufficient numbers of patients to do generalisable clinical and translational research. Geographic spread of patients has been a major impediment to recruitment into clinical trials. Most rare diseases do not have a specific International Classification of Diseases code, which hampers research that uses existing databases.3 Before the USA, the

Published Online August 2, 2010 DOI:10.1016/S01406736(10)60680-0

1057

Comment

European Union, and Asian countries passed orphandrug legislation more than 20 years ago, the drug industry gave little attention to the development of drugs for these diseases. Although these laws increased the pace of orphan-drug development,4 most rare diseases still have no medical therapy. In recognition of these barriers and the moral and public health imperatives to advance knowledge on the best ways to improve the health and wellbeing of patients with rare diseases, recent conferences in the USA5 and Europe6 called for wide expansion of access to registries for such patients. The US meeting called for the creation of the infrastructure for a global registry. Once the population has been defined, various data types can be added. Data can be entered by patients, clinicians, researchers, or be imported from electronic health records. Scientists and drug companies are more likely to research a rare disease if they find a registry in place. Registries enable the formation of infrastructures for various types of research, education, and outcomes improvement (panel).7,8 Less than a fifth of rare diseases have registries, and most of these are operated by patients’ organisations Panel: Research functions to enable a patients’ registry for rare diseases • Knowledge dissemination: distribution of information to patients and their clinicians on new therapies, best practices, and safety issues • Patients’ recruitment: providing patient-population information for designing trial protocols that optimise size and length of trials • Clinical epidemiology: population descriptive statistics, natural history of disorders, medical practice variation • Clinical effectiveness: evaluation of the effects of preventive, diagnostic, and curative interventions delivered in real-world settings • Safety monitoring: orphan drugs are generally not tested in large phase 3 studies, which makes the need for postmarketing safety surveillance via registries even more important than with conventional drugs4 • Quality and outcomes improvement: enhancing patients’ outcomes by standardising practice and reducing practice variation • Genotype/phenotype association studies: the registry provides phenotypic data which can be linked to genetic and other exposure data • Linkage to biospecimens and biorepositories: to detect phenotypic correlates of cell and tissue biology

1058

or researchers.6 Although most registries are countryspecific, there are a few international efforts (eg, in cystic fibrosis9 and neuromuscular diseases10) that are showing the benefits of combining data across international boundaries. We believe that now is the time to design and develop the infrastructure to foster global rare-disease registries. The increasing mobility of populations and the globalisation of lifestyles and food products make it clear that disease knows no boundaries.11 Some rare diseases occur so infrequently (<1 per 1 000 000 population) that only by forming international populations can sufficient numbers of patients be accrued. Because funding has been a key obstacle to establishing and maintaining registries, economies of scale that can be developed by forming a global rare-disease infrastructure would improve access to registries for many patients. Registries are infrastructure, not research projects, and as for so many global concerns, there is no single funding source. A federated model in which several registries for the same disease are linked will most probably be needed to form a global infrastructure. A federated model requires that individual registries are developed or, for those already in existence, transformed to ensure that they are interoperable (ie, data are defined in the same way, use the same standards, and are stored in the same vocabularies). For registry developers, there is no established forum for sharing experiences. Each time a new registry is developed, it starts from scratch.6 Information on best informatics practices and common data templates would go a long way toward reducing the start-up costs associated with developing a registry. Some data elements might be common to all rare diseases (eg, sociodemographics, diagnosis, genetics, growth, medications, services), which raises the possibility of creating a core dataset that can be incorporated into all rare-disease registries. A single individual, group, or even country will not lead the movement toward formation of a global rare-disease registry. As in the open-source software community, an open-science community for rare diseases is needed. Such a community would ensure that the conditions necessary for data exchange are addressed by defining common datasets, data standards, and vocabularies, and would provide a forum for exchange of experiences and knowledge. The biggest hurdle to our vision of a global registry is not technical, but rather the www.thelancet.com Vol 377 March 26, 2011

Comment

cultural obstacles to collaboration and data sharing across academic institutions and international boundaries. Overcoming these hurdles is extremely important. A global infrastructure for a rare-disease registry will inject new energy into the effort to deliver more fully on the promise of orphan-drug legislation. Such a registry will draw new interest in rare diseases from academic researchers and the drug industry because it will enable the design of more effective clinical trials and effectiveness studies and the recruitment of patients much faster and at much lower cost. *Christopher B Forrest, Ronald J Bartek, Ya a Rubinstein, Stephen C Groft Children’s Hospital of Philadelphia, Philadelphia, PA 19104, USA (CBF); Department of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, PA, USA (CBF); Friedreich’s Ataxia Research Alliance, Springfield, VA, USA (RJB); and Office of Rare Diseases Research, National Institutes of Health, Bethesda, MD, USA (YR, SCG) [email protected] We declare that we have no conflicts of interest.

1 2 3 4 5

6

7

8 9 10 11

European Commission. Rare diseases. http://ec.europa.eu/health/ph_ threats/non_com/rare_diseases_en.htm (accessed April 8, 2010). Office of Rare Diseases Research, National Institutes of Health. Rare diseases and related terms. http://rarediseases.info.nih.gov/ RareDiseaseList.aspx (accessed April 8, 2010). Schieppati A, Henter J-I, Daina E, Aperia A. Why rare diseases are an important medical and social issue. Lancet 2008; 371: 2039–41. Haffner ME. Adopting orphan drugs—two dozen years of treating rare diseases. N Engl J Med 2006; 354: 445–47. Office of Rare Diseases Research, National Institutes of Health. Advancing rare disease research: the intersection of patient registries, biospecimen repositories and clinical data. http://rarediseases.info.nih.gov/ ScientificConferences.aspx?PageID=5&ID=1021 (accessed April 15, 2010). Wrobel P. Workshop on patients registries for rare disorders: need for data collection to increase knowledge on rare disorders and optimize disease management and care. Brussels, Belgium; March 18–19, 2009. http:// www.epposi.org/web/publications/_docs/publications/2009/Documents. aspx (accessed May 16, 2010). Gliklich RE, Dreyer NA. Registries for evaluating patient outcomes: a user’s guide. Jan 1, 2007. http://www.effectivehealthcare.ahrq.gov/index.cfm/ search-for-guides-reviews-and-reports/?pageaction=displayproduct&prod uctid=12 (accessed May 16, 2010). Watson MS, Epstein C, Howell RR, et al. Developing a national collaborative study system for rare genetic diseases. Genet Med 2008; 10: 325–29. McCormick J, Mehta G, Olesen HV, Viviani L, Macek M Jr, Mehta A. Comparative demographics of the European cystic fibrosis population: a cross-sectional database analysis. Lancet 2010; 375: 1007–13. Butcher J. Cooperation is key, say neuromuscular-disease researchers. Lancet Neurol 2007; 6: 298–99. Dator DP. The globalization of medical care: managing rare diseases from elsewhere. J Urol 2010; 183: 1280–81.

Cardiology: a call for papers Why not send your best cardiology papers to The Lancet? The Lancet is planning a special issue to coincide with the European Society of Cardiology’s meeting to be held in Paris, Aug 27–31, 2011. We will consider highquality original research papers that describe the results of randomised trials and will influence clinical practice. HotLine papers will be considered for fast-track review at the journal to allow publication immediately after

the presentation. For HotLine papers only, the deadline for submission is July 18, 2011. The deadline for other submissions is May 23, 2011. Please state in your covering letter that the submission is in response to this call for papers.

To submit a paper go to http:// ees.elsevier.com/thelancet

Stuart Spencer The Lancet, London NW1 7BY, UK

A thank you to all our peer reviewers in 2010 Today we publish as a webappendix our annual list of clinical and statistical peer reviewers from many different countries and specialties who helped us choose and improve papers in 2010. Those who reviewed more than five papers are marked with an asterisk. By publishing their names, we hope that some small recognition of an important academic task can be given to those who give their time, knowledge, and expertise. While the academic community ponders on the best way to do peer review, a much www.thelancet.com Vol 377 March 26, 2011

more urgent question for those who design research assessment exercises is: how can this time-consuming yet undervalued activity be more formally appreciated and incorporated into the assessment of an academic’s work? As journal editors, we highly regard and respect our trusted reviewers. Thank you all!

See Online for webappendix

Sabine Kleinert The Lancet, London NW1 7BY, UK

1059

Comment

AFP/Getty Images

Photolibrary

Corbis

Offline: Where was Europe? Over 600 people gathered in Seattle last week to take stock of global health. This was the first—the second actually, but this time we did it properly—Global Health Metrics and Evaluation (GHME) conference. Do you begin scientific meetings enthusiastically? A race to check in to the hotel, find breakfast and the bar, and then register. An intense trawl of the programme to plan each day. Difficult choices over which sessions to sacrifice over others. And then the opening plenary, often with music and garlanded speeches. That’s when it hits you, that feeling you knew was coming, but somehow had managed to bury in the vaults of your memory. The whole utter dread of the moment envelops you as you realise you have at least another 3 days of mediocre presentations, broken sleep, and obesogenic snacking, all stealing precious hours from your rapidly dwindling life as you feebly take notes you know you will never look at again. GHME was not that kind of conference.

be read as a radical critique of WHO as it exists today. Just who is WHO accountable to, he asked. The standard answer is: our member states. Wrong. WHO is accountable to the peoples of member states, not their governments. Indeed, a crucial role for WHO is not simply to respond to the whims of countries, but to defend the health of their peoples despite those governments. Since the era of Brundtland, WHO has failed to live up to this important accountability role. While the agency’s timidity might satisfy some powerful governments, it fails citizens. And its strategy of member-state appeasement does not protect WHO from the burden of bankruptcy. With donors failing to support even a compliant UN, hundreds of jobs are now being shed across the organisation. WHO is being hollowed out. The current meanness of member states will hurt those same states in the future. But in the meantime, WHO needs not only to recover its voice, but also sharpen its edge.

*

*

The preamble was the same. But from the opening morning, this was a grade A show. More data to reveal the continuing decline in maternal mortality. Evidence showing that malaria prevention programmes are having a large impact on under-5 deaths. A better understanding of how maternal and paternal education improves the lives of children. 35 million iatrogenic injuries that take place every year. A neoplastic transformation that is sweeping the world. Children younger than 5 years, even in subSaharan Africa, facing a double burden of undernutrition and overweight. In Latin America, alcohol, systolic blood pressure, body-mass index, and tobacco (along with violence in some countries) are risk factors ravaging populations. And the global burden of dental disease is falling.

A troubling trend. The scientists in Seattle came mostly from American institutions, especially the Institute of Health Metrics and Evaluation, Harvard, and Hopkins. There were emerging strong voices from Latin America, Asia, and sub-Saharan Africa. The London School of Hygiene and Tropical Medicine also did well (it made it to the top 100 list of Higher Education World Reputation Rankings this month, and came first in the otherwise depressingly awful UK university funding settlements with government last week). But the biggest surprise was continental Europe. Its absence, that is. Presumably there are some scientists concerned with global health issues in France, Germany, Spain, Italy, the Netherlands, Scandinavia, and so on. If so, they were few and far between in Seattle.

LSHTM

* Julio Frenk is the Dean of the Harvard School of Public Health. He also has a past. Aside from being Minister of Health in Mexico, Julio once worked with Gro Harlem Brundtland while she was Director-General of WHO in the late 1990s. Their most famous initiative was the World Health Report , a project that ranked countries according to their health-system performance (Brazil never quite forgave them). In Seattle, Julio Frenk reviewed the lessons of WHR . His argument could 1060

* And a final note. Howard Bauchner is to be the new Editorin-Chief of JAMA. Since 2003, he has edited the Archives of Disease in Childhood with great skill. Everyone at The Lancet wishes his predecessor, Cathy DeAngelis, well, at the same time as we warmly welcome Howard to his new role. Richard Horton [email protected]

www.thelancet.com Vol 377 March 26, 2011

World Report

Japan: the aftermath Health and aid workers in Japan face multiple challenges in the wake of the earthquake and tsunami that have devastated the country’s northeast coast. Justin McCurry reports from Tokyo.

www.thelancet.com Vol 377 March 26, 2011

areas such as those struck by the quake and tsunami. Many of those who were able to flee the tsunami are battling for survival in cramped conditions with insu cient food, water, and heating and, in the most serious cases, without life-saving drugs.

“‘It’s the sheer scale of it...There must still be pockets of people out there who haven’t been reached.’” The case of Senen General Hospital in Takajo, a town in Miyagi prefecture, is typical of other medical facilities in the tsunami zone. About half the 100 patients were severely ill before the calamity, and the remaining 80 have been transferred to shelters. For the first few days after the disaster the hospital had no running water and patients survived on frozen noodles and vegetables retrieved from a damaged freezer. Employees at other hospitals have reported similarly desperate conditions, with the lack of water the most pressing concern. In some areas, non-critical patients have been transferred to hotels or are being treated in corridors and cafeterias. Although some hospitals are still using existing drug stocks and other medical supplies, others have run out or are running low. At a hospital in Futaba, just 10 km from the power plant, troops discovered 128 elderly patients, some of them comatose, who seemed to have been left to fend for themselves by medical sta . 14 patients died, two of them on a bus while they were being taken to an evacuation centre. Yet the humanitarian and health crises have been overshadowed by fears surrounding the Fukushima Daiichi nuclear power plant, 150 miles north of Tokyo. All six of its reactors have had failures of their cooling

systems, raising fears of a meltdown that could release radioactive material over a wide area. The team of technicians, firefighters, and soldiers working around the clock to make the plant safe are at greatest risk of being exposed to dangerous levels of radiation. But the threat of meltdown— however remote—and evidence that small amounts of radiation have leaked into the atmosphere and the food supply, sparked an additional health scare this week. Government o cials and international health experts were quick to reassure the public that tainted milk, spinach, and other vegetables produced in the Fukushima area would have to be consumed in enormous

Published Online March 22, 2011 DOI:10.1016/S01406736(11)60413-3

Most a ected prefectures Districts with assessed damage or flooding A ected nuclear power plants

Iwate Sea of Japan

Sendai

Miyagi

Earthquake epicentre

Fukushima

Fukushima Daiichi and Daini nuclear power plants Japan Tokyo

Pacific Ocean

0

50

100 150 km

Map showing regions most a ected by the earthquake and tsunami as of March 16

1061

USAID

In the immediate aftermath, the aid and medical response to the earthquake and tsunami that struck Japan on March 11 has been complicated by the sheer scale of the devastation, widespread damage to supply routes, and concerns about radiation leaks from a stricken nuclear power plant. As emergency supplies of fuel, water, food, blankets, and other essentials finally began to get through to the estimated 350 000 people living in 2500 evacuation centres in the northeast of the island, o cials were issuing reassurances about food, milk, and tap water found to have been contaminated with radioactive iodine-133. Just over 1 week after the magnitude 9·0 earthquake unleashed a powerful tsunami that washed away entire communities on the coasts of Iwate, Miyagi, and Fukushima prefectures, Japan is only just beginning to comprehend the scale of the humanitarian crisis in its midst. Japan’s police agency says that the death toll has reached 8649, with 12 877 people still missing. But with authorities in Miyagi reporting more than 15 000 people missing in their jurisdiction alone, the final total is expected to be much higher. 1 million homes are still without water, and the quake and tsunami destroyed more than 14 000 buildings and damaged 100 000 more. The disaster has inflicted serious damage on the region’s health infrastructure. Many hospitals collapsed or were flooded, and doctors, nurses, and other health workers are either among the victims or living in evacuation shelters. One of the defining characteristics of the tragedy is the large number of elderly victims and survivors. More than 20% of Japan’s population is 65 years or older, and is about 30% in remote

AFP/Getty Images

World Report

An 80-year-old woman is rescued from her destroyed home, 9 days after the quake

quantities to pose a health risk. But that has not stopped the government from considering the distribution of potassium iodide tablets—which can prevent radioactive iodine-131 from being absorbed by the thyroid gland, causing cancer. The tablets were the target of panic-buying sprees in the USA and elsewhere, prompting WHO to o er advice on how to avoid radiation and when to take potassium iodide pills. WHO said that the contaminated products posed no threat to health but described the radiation scare as “serious”, given that it had spread well beyond the 30 km exclusion zone to farms as far as 90 miles from the atomic power plant. Abnormally high quantities of iodine-131 and caesium-137, another radioactive material, have also been identifi ed in tap water in Tokyo, though again, in quantities too small to pose a threat to health, o cials said. Iodine-131 decays within a matter of weeks, but can pose a shortterm risk to health, according to the International Atomic Energy Agency. “There is a short-term risk to human health if radioactive iodine food is absorbed into the human body”, the agency said. “If ingested, it can accumulate in and cause damage to the thyroid. Children and young people are particularly at risk.” A potentially greater concern is caesium-137, which caused widespread damage to the food supply in the wake of the Chernobyl disaster 1062

in 1986. The substance takes much longer to break down in the environment and can increase the chances of developing cancer. Even when homes are rebuilt, communities reformed, and the health infrastructure restored to its pre-quake levels, Japan’s health system is ill prepared to address long-term mental health problems triggered by the disaster. Experts say that thousands of victims will be in need of long-term trauma counselling to cope with the loss of their relatives, friends, and homes. They include an estimated 100 000 children who were displaced by the disaster and were particularly vulnerable to post-traumatic stress and long-term behavioural and other mental health problems. “Children we talk to say that whenever there’s a tremor they are scared that something is going to happen”, says Stephen McDonald of Save the Children, which has set up an operations base in Sendai. Children who have been caught up in disasters can develop behavioural and mental health problems unless they receive counselling at an early stage. Left untreated, those initial fears can impinge on their development as adults, said McDonald. The already dire situation in the northeast has been compounded by a shortage of doctors and nurses in some hospitals. In towns and villages covered by the 30-km radius from the nuclear power plant, where 210 000 people have either been evacuated or told to stay indoors, providing even basic treatment is a daily struggle. At one hospital only 60% of sta are on duty and, although there are enough medical supplies on site, inpatients’ meals are having to be restricted. Across the stricken region, hospitals are sending non-critical patients home or to evacuation centres, and more serious cases are slowly being found beds at other facilities. The stories of misery are relentless, from the elderly patients with high blood pressure and diabetes who have been unable to take their medication, to younger

people with diarrhoea from drinking river water in desperation. At the start of this week, health workers on the ground said priority should be given to ensuring safe drinking water and the disposal of sewage to prevent outbreaks of typhoid and cholera. “For the people a ected by the earthquake as well as the tsunamis, there are a number of issues: the cold, as well as a lack of food and water”, said Eric Ouannes, general director of the Japan o ce of Médecins Sans Frontières, which has 11 people in the region. “And the most urgent need is blankets to protect the most vulnerable sections of the population against the cold.” Ouannes said he had seen chronic diseases usually associated with the elderly: hypertension, cardiac diseases, and diabetes. “We are trying to restart the treatment. We have also seen some cases of hypothermia on top of the other diseases, as well as dehydration. But again, it’s a limited number of cases with regards to the vast number of people who have been displaced or who have lost their homes.” Doctors who tour evacuation shelters in Fukushima, Miyagi, and Iwate prefectures said they were bracing for the possible spread of gastroenteritis, diarrhoea, and other illnesses associated with dirty water. The relief e ort is being hampered by damage to road and sea routes, and the loss of power supplies and phone networks. “Everywhere is closed, there are no telecommunications, no heat, and no light”, said Patrick Fuller of Red Cross International, who is based near Otsuchi, an Iwate prefecture town where more than half the 19 000 population is thought to have died. Nothing could have prepared him or his Japanese colleagues for the size and complexity of the humanitarian crisis, he said. “It’s the sheer scale of it. You could have a thousand helicopters and it still wouldn’t be enough. There must still be pockets of people out there who haven’t been reached.”

Justin McCurry www.thelancet.com Vol 377 March 26, 2011

World Report

Gairdner Foundation names global health award winner On March 23, the Canadian Gairdner Foundation announced that a leading child health expert was the winner of its prestigious annual global health award. Priya Shetty reports.

www.thelancet.com Vol 377 March 26, 2011

Bryce pays tribute to Black’s tremendous dedication to an area that was for many years at the bottom of the pile. “Bob is steadfast”, she says. “He championed child health and nutrition long before their recent adoption as global priorities; in fact, his work and that of his colleagues and students are responsible for much of the current attention to maternal, newborn and child survival.”

“‘He championed child health and nutrition long before their recent adoption as global priorities...’” Although Black may have struggled initially to convince the world of the crisis in child health, the data have long attested to that fact. According to WHO estimates, twothirds of the 9 million child deaths in developing countries each year are caused by infectious diseases such as pneumonia, diarrhoea, and malaria; undernutrition contributes to a third of these 6 million child deaths. Indeed, one of Black’s major achievements has been to champion zinc supplementation. Zinc defi ciency increases the risk of diarrhoea, pneumonia, and malaria and the deaths of 0·4 million children younger than 5 years in developing nations. Black’s studies in Bangladesh, India, Peru, and Zanzibar, showing that daily zinc supplementation could greatly reduce the severity of childhood illnesses such as diarrhoea and pneumonia, has led to substantial policy changes—UNICEF and WHO now recommend adjunct therapy with zinc for the treatment of all diarrhoea. Black fi rst began investigating the links between diarrhoea and nutrition at the International Centre

for Diarrhoeal Disease Research, Bangladesh, where he worked for 2 years after focusing on enteric diseases at the US Centers for Disease Control and Prevention in the 1970s. In Bangladesh, he began a series of longitudinal studies to elucidate the links between diarrhoea and nutrition, and realised how important the connection was: most of the children who died of diarrhoea were those who were most undernourished. The 2010 Gairdner award for global health went to Nicholas White, a renowned malaria expert at the Mahidol Oxford Tropical Medicine Research Unit at Mahidol University in Bangkok, Thailand. Both White and Black have striking similar ethos. Throughout their careers, the search for low-cost, large-scale interventions that are practical and pragmatic has always won out over satisfying an academic or intellectual scientifi c curiosity. Thus, they would rather advocate for research into devising more e ective ways to deliver an existing treatment that is proven to work than to fund research into an altogether new treatment. Maternal and child health has begun to feature more prominently in the global consciousness over the past decade. But Black is not complacent. Although he welcomes the growing attention that child health is getting, he says that “substantial reductions in child mortality will only be possible with more focus on the two major direct causes of child death—pneumonia and diarrhoea— and on the underlying conditions of stunted growth and micronutrient defi ciencies. These problems are still receiving insu cient attention.”

Published Online March 23, 2011 DOI:10.1016/S01406736(11)60414-5 See World Report page 1064

The Gairdner Foundation

Robert E Black, who is professor of public health at the Johns Hopkins Bloomberg School of Public Health in Baltimore, MD, USA, is the 2011 recipient of Canada’s Gairdner Global Health Award for his work to reduce childhood diarrhoea, which kills more than 1 million children in low-income countries every year. The Canadian Gairdner Foundation honours leaders in international health and biomedicine with awards of CAN$100 000 each. The Gairdner Foundation was established by Toronto stockbroker James Arthur Gairdner in 1957. Since its inception, it has awarded prizes to 303 scientists, 76 of whom have gone on to win a Nobel prize. Fittingly, Black is being honoured with the illustrious award at a time when child health is higher up the global agenda than it has ever been before. Black has been working in the specialty for 30 years now, and at the beginning was one of just a handful of people who recognised the staggering neglect of child and maternal health, and the importance of nutrition in early life. Jennifer Bryce, a leader in child health, is a colleague of Black’s at Johns Hopkins, and the pair work closely on assessing large-scale intervention programmes for women and children. Bryce has worked with Black since the mid-1990s when she was responsible for assessing at WHO’s Department of Child and Adolescent Health and he was a member of the Technical Advisory Group. Bryce was so impressed by how “independent and tough” Black was, that when she received funds from the Bill & Melinda Gates Foundation to do a multicountry assessment of the Integrated Management of Childhood Illness strategy, Bryce brought Black on board as a technical adviser.

Robert E Black

Priya Shetty 1063

World Report

And the 2011 Gairdner International Award winners are... Published Online March 23, 2011 DOI:10.1016/S01406736(11)60415-7

Canada’s Gairdner Foundation honoured researchers in immunology and genetics on March 23 with some of the world’s largest annual international research awards. Paul C Webster reports.

The Gairdner Foundation

See World Report page 1063

The Gairdner Foundation

Jules A Ho man

The Gairdner Foundation

Shizuo Akira

The Gairdner Foundation

Howard Cedar

The Gairdner Foundation

Aharon Razin

Adrian Bird

1064

The 2011 Gairdner International Awards valued at CAN$100 000 each went to Jules A Ho mann, Distinguished Class Research Director, CNRS, Strasbourg, France, and Shizuo Akira, Director of the WPI Immunology Frontier Research Center, Osaka, Japan, for their contribution to establishment of toll-like receptors (TLRs) as pathogen recognition receptors as well as clarifi cation of their signalling pathways. Howard Cedar and Aharon Razin, both of the Department of Developmental Biology and Cancer Research, Hebrew University, Jerusalem, Israel, and Adrian Bird of the Wellcome Trust Centre for Cell Biology, University of Edinburgh, UK, were also honoured for their discoveries in the specialty of DNA methylation and related gene expression. Next to the Nobel Prize in Medicine, the Canada Gairdner Awards are the most prestigious global medical research awards, according to the Gairdner Foundation. The Foundation began recognising basic science pioneers in 1959. In 2008, the Government of Canada endowed CAN$20 million to support the awards. 303 international awards have been given to scientists from 14 countries. The Foundation has a reputation for spotting winners and breakthroughs early: in the 52 years since the Gairdner Awards were first presented, 76 Gairdner winners have gone on to win Nobel prizes. Canada Gairdner Award winners have included the discoverers of breakthroughs including the vaccine that eradicated smallpox and the identifi cation of the structure of DNA. The awards are chosen using a twopart selection process involving a Medical Review Panel, composed of leading midcareer scientists from across Canada, which reviews all nominations.

Their recommendations are passed to a Medical Advisory Board, composed of 20 senior scientists from across Canada and around the world. All adjudication is done on a voluntary basis. The awardees are chosen by secret ballot of the Medical Advisory Board. This year, notes John Dirks, President and Scientifi c Director of the Gairdner Foundation, four of the winners are pioneering geneticists. “On the 10th anniversary of the first model of the human genome people are questioning whether it has resulted in what was first visualised”, Dirks observes. “The work of the researchers we are honouring typifi es the huge explosion in understanding of literally hundreds of diseases that has emerged.”

“...76 Gairdner winners have gone on to win Nobel prizes.” In honouring the work of Cedar, Razin, and Bird, the Foundation says that the three have made seminal contributions in recognising that the placement of methyl groups on DNA could a ect the way regulatory molecules interact with DNA sequences, how the static DNA code could di er between cell types, and how the environment could interact with our DNA code. Cedar explains that the award raises the stature of a specialty of basic science with important emerging implications for cancer research. “We weren’t interested in the disease, we were interested in basic biology”, he says about the initial investigations of DNA methylation. “But it turns out that methylation presents the opportunity to prevent the disease.” Bird says that investigations into methylation changes in utero are especially intriguing. “Our approach is to look at the basic elements in understanding the mechanisms of

epigenetic change.” 10 years after the genome was published in draft form, Cedar describes his work as “doing the annotations” to the original book. The Gairdner Foundation describes the work of Ho mann and Akira on the discovery of TLRs as “revolutionising our understanding of innate resistance to infection in the animal kingdom by discovering the role of toll-like receptors and the array of microbial compounds that they recognise and respond to. Understanding TLRs has opened up an entirely new era in the understanding of the role of innate immunity.” Jack Gauldie, director of the Institute for Molecular Medicine and Health at McMaster University in Hamilton, Ontario, who serves on the Gairdner Foundation’s Medical Review Panel as well as its Medical Advisory Board, describes the TLR discoveries as “fundamental: This is the way that almost all species recognise when they are being attacked. This is how we actually recognise pathogen exposure”. Although the overall area of research is quite new, “it is really very primal in how all organisms stay alive”. Bird hopes the Gairdner Awards will help to propel further research into DNA methylation, an area that he describes as something as a “Cinderella subject—sometimes in fashion, sometimes out of fashion”. As well as identifying promising specialties of basic biomedical research, Bird says the Gairdner Foundation has been highly successful in engaging Canadian society with basic science. Award winners are expected to participate in a suite of events with Canadian students, researchers, and policy makers, he notes: “This is the way to get science into the bloodstream of a country.”

Paul C Webster www.thelancet.com Vol 377 March 26, 2011

Perspectives

Book A clone speaks from his soul I am a clone, the son of a clone, and—like many of those who share that distinction—an evolutionary dead end. My own body cells copy themselves, of course, without benefi t of sex; and of all the untold billions of sperm they have made since adolescence, not one has succeeded in fi nding a fertile egg with which to fuse and to generate some modifi ed version of myself. My mother, as it happens, was an identical twin (although I do not think that is why I became a geneticist); and even if I did sometimes, as a child, fi nd it hard to tell her and her sister apart, their uncanny similarity never struck me as posing a deep theological question. For some people, it did. Unnatural: The Heretical Idea of Making People explains why. It also tells us why the Catholic Church prohibits babies made by IVF (in-vitro rather than invagina fecundation) but accepts GIFT (gametes, sperm and egg, inserted into the Fallopian tube, preferably with sperm extracted from the vagina) in a desperate attempt to avoid the theological pitfalls of the test-tube. It all has to do with the soul. When and how (or indeed why) that structure appears has bothered philosophers since the Greeks. According to Thomas Aquinas the Y chromosome must play a part, for boy fetuses gain theirs at 6 weeks of gestation, while females have to wait twice as long. Other theologians disagree. For pious Catholics, the soul is inserted when sperm meets egg—which is where my mother’s spiritual status became ambiguous. Identical twins arise through the splitting of an embryo well after fertilisation—but if that is so, where does the second soul come from? Were my mother and my Aunt Peggy blessed with only half a soul each, or did God have a stock of spares ready to insert when needed? Perhaps the magical structure only www.thelancet.com Vol 377 March 26, 2011

made its entry when the fetus was recognisably human—or even when it was born? Such ancient issues still impinge on life and death, on contraception and abortion, and on the supposed rights of cells—stem cells most of all—to a life of their own. Their Church has of late been more ambiguous on the timing of that event: in 1974 the Holy See opined that even if ensoulment is delayed (which it doubted), the early embryo

“Unnatural ends by reminding us that science has not, in fact, discovered the secret of life, and might find in the end that it is not worth looking for.“ is still “a human life, preparing and calling for a soul…It su ces that the presence of the soul be probable (and the contrary will never be established)”. In this book, Philip Ball pours well-merited scorn on such slippery reasoning: “As a statement of bald superstition, this one takes some beating”. The Bible was the first genetics text of all, and although biology has progressed a little since then, theology—as Unnatural shows— remains trapped between a dimly remembered past and a largely imagined future. The Saviour saw the feeble body as detached from the eternal soul, while the plant-breeding monk proved that the genetic message is disengaged from the ephemeral existence of its carriers. Both embody the idea of resurrection; the renewal of a being defi led by death. Even so, as science has progressed, ethics has found it hard to keep up. In 1163 the Council of Tours prohibited dissection on the grounds that it would be impossible to reassemble the pieces on Judgment Day. Eight centuries later the philosophers are still debating such weighty matters.

They often fi nd an excuse to do so by claiming that science can do far more than it really can. From Mary Shelley to Johann Wolfgang von Goethe to Aldous Huxley (and even to his biologist brother Julian who wrote a bizarre short story in which a savage chief cloned his cells and distributed them as holy relics) and from oral contraception to Dolly the Sheep (to Bill Clinton “a matter of mortality and of spirituality”), the smallest advance in biology generates large and horror-struck claims of playing God, of eugenics, or of worse. This has led to ethical, social, and political upheaval; to wars, unwanted children, the death of women through illegal abortion, and to what Unnatural, in its energetic attack on irrationality, describes as “theocon” interference with biological research. Ball chases the story from Zosimos of Panopolis through Plato to George W Bush’s appointee on his Ethics Advisory Board who said of the birth of Louise Brown, the fi rst IVF baby, that “this blind assertion of will against our bodily nature… can only lead to self-degradation and dehumanization”. It takes in a vast corpus of deliberate self-delusion on the way (not to speak of some of the classics of English literature). The thread that unites them all is a fear of the unnatural; of making something that contains within itself the potential to be at least in part human. In this, his sixteenth book, Ball ranges widely through the history of those who have persuaded the gullible that artifi cial life is just around the corner. They range from the bronze fly that chased away the other biting insects of Naples to Thomas Edison’s phonographic doll that was accused by some of being almost human because it could talk. For bioethicists, if Frankenstein had not been real, it would have been

Unnatural: the Heretical Idea of Making People Philip Ball. Bodley Head/Random House, 2011. Pp 384. £20·00. ISBN 1847921523

1065

Perspectives

necessary to invent him. Ball puts it well: the biggest hazard faced by the (perhaps too well-funded) bioethics industry is grandiosity. Its employees are addicted to overstatement, to hysteria, and to an ingrained inability to separate fact from fiction. Francis Crick himself claimed that “We have discovered the secret of life!” when he ran into the Eagle pub in

Cambridge to announce the structure of DNA, and some of his intellectual descendants have not calmed down since then. Unnatural ends by reminding us that science has not, in fact, discovered the secret of life, and might find in the end that it is not worth looking for. 10 years on from the completion of Crick’s doubly helical agenda, the scientific journals have

been filled with rather rueful accounts of quite what that molecule has, and has not, told us about the workings of the cell or of evolution with scarcely a word about the sequence that codes for the soul. For the ethicists, though, it’s Brave New World all over (yet) again.

Steve Jones [email protected]

In brief Film Lost childhoods In November, 2009, Kevin Rudd who was at that time the Australian Prime Minister, said he was “deeply sorry” for the hardship and abuse endured by thousands of children as a result of the child migration programme from the UK that the Australian Government had supported during the postwar decades, until as late as 1970. A few months later, Gordon Brown followed suit on behalf of the British Government, calling child migration an “ugly stain” on the UK’s recent history. Last month, Prime Minister David Cameron reiterated those sentiments and announced a fund for reuniting former child migrants in Australia, Canada, and other past British colonies with their relatives. He singled out the dedicated work of Margaret Humphreys, who founded the Child Migrants Trust. Oranges and Sunshine, a raw and disturbing fi lm about the child migrants sent to Australia, is therefore timely, although conceived some years before these political

Oranges and Sunshine Directed by Jim Loach, screenplay by Rona Muro. Icon Films, 2011. On general release in the UK from April 1, 2011. See http:// www.iconmovies.co.uk/ orangesandsunshine See Online for webvideo

Icon Films

For Child Migrants Trust see http://www.child migrantstrust.com

1066

announcements. It centres on Humphreys (persuasively played by Emily Watson), a mother of two and a social worker from Nottingham in the UK, whose job brought her into contact with a former child migrant in 1986. The plot is based on her struggle in the UK and in Australia to get recognition for the injustice done to well over 100 000 children, which she documented in her 1994 book Empty Cradles. The fi lm’s title comes from a ploy used to entice orphaned and “unwanted” children to Australia. Instead of a harsh life in a British children’s home, they would have endless oranges and sunshine. A lie was also used in many cases: children were told that their mothers had died, while mothers were encouraged to think that their children were being sent away to loving adoptive families. A lucky few were; most were used as cheap labour; some were cruelly abused in outback institutions run by the Catholic Church, whose reputation will no doubt get a further

beating from this fi lm. Later in life, many of these children would su er a crisis of identity, often leading to depression and even suicide. None of these facts are skated over, but neither are one’s emotions manipulated, whether in the bleakness of a British winter or the sun-baked bush of western Australia. The lead performances are of extraordinary subtlety, especially David Wenham’s as Len, a strapping former child migrant who has done well in business in Australia yet who desperately needs to meet his English mother—whatever her reasons may have been for letting him go as a boy. Len’s sparring but trusting relationship with the buttoned-up Margaret gives the film both grit and grace. Sunshine and Oranges does not have a message, despite its shocking source material, and is all the better for this. Directed with sensitivity by Jim Loach (son of the director Ken Loach), the film packs a punch; but it makes you think, as well as feel outraged. Social workers will be particularly appreciative. Honest politicians may find themselves refl ecting on the fact that while compassionate personal initiative may be key to the Big Society, it was Nottingham Social Services that supported Margaret Humphreys’s initial research with 2 years’ paid leave.

Andrew Robinson [email protected]

www.thelancet.com Vol 377 March 26, 2011

Perspectives

A visit to the doctor was a special event in Lucica Ditiu’s childhood. “In our household there was a high regard for doctors as being important and powerful individuals in the way they looked after people. A medical visit was one of the only times we would wear our best clothes!” Ditiu went on to become a doctor and is now the newly appointed Executive Secretary of the Stop TB Partnership—a network of global partners working towards eliminating tuberculosis as a public health problem. A native Romanian whose parents were factory workers, Ditiu grew up during a turbulent period when the country was under the rule of Nicolae Ceausescu. “I grew up largely on the streets because my parents were poor”, Ditiu recounts. But she studied hard and scored highly in the medical entrance exams that secured her a place at Bucharest’s University of General Medicine and Pharmacy. After graduating, in 1992, she completed specialty training in pulmonology. But Ditiu found it hard to survive on her US$30 per month salary and so she also worked as a journalist for a glossy Romanian magazine. “I love people and I love communicating and nearly gave up medicine to do journalism full time, but as I got more immersed in the fi eld of tuberculosis, and saw for myself that DOTS treatment when organised and done properly leads to good health outcomes, I did not give up and it is why I am still in tuberculosis today.” Her commitment is inspirational, says Giovanni B Migliori, one of her mentors and a tuberculosis expert for WHO: “Lucica’s story shows that it is possible for a young woman, from an emerging country and not a traditional academic to reach a top position in the fight against tuberculosis. To me it is a message of hope for a young generation of specialists.” Ditiu began her career with WHO in 2000, as a medical o cer for tuberculosis in Albania, Kosovo, and Macedonia. She has been the driving force behind many initiatives that have raised the profile of tuberculosis control in the region, such as establishing national tuberculosis programmes and guidelines, strengthening tuberculosis/HIV integration, and creating a special Laboratory Task Force. Working closely with the Global Fund to Fight AIDS, Tuberculosis and Malaria since its creation in 2002, Ditiu also provided technical support to countries with their applications for financing. Her dedication went as far as finishing a Global Fund application whilst delivering her baby and signing the proposal just after giving birth. “My doctor thought I was mad, but we got $45 million which made such a di erence in Romania”, she says. Today, however, Ditiu laments how much there is still to do in the region where vertical programmes, weak infrastructure, and poor infection control have all hampered progress. “What is really missing in Europe is that medical education does not focus on how you communicate, engage, and respect the patient, and follow them through their www.thelancet.com Vol 377 March 26, 2011

diagnosis and treatment. The medical sta are disengaged from the patient. Our tuberculosis patients are not always the nicest in the world but there has to be more focus by the medical community to understand their needs and work with them. Then there is the absence of a strong tuberculosis community and this is a problem globally. We need to make patients understand the power they have to mobilise action at a national level, and how they can help each other to stop stigmatisation and more practically, deliver treatment.” Ditiu believes it is time to focus on tuberculosis as a broader health and development problem. “A few years ago tuberculosis and tobacco were fashionable, then tuberculosis and health systems. Next year it will be tuberculosis and blue eyes! We can’t keep randomly linking. Tuberculosis is a disease that needs to be looked at contextually in the country. It is time for us to broaden the mindset around tuberculosis, demystify it, and strengthen the approach towards a patient who has tuberculosis amongst other illnesses,” she urges. To gain global attention for tuberculosis, Ditiu wants the partnership to have a unified voice and focus on advocacy and communication. ”I want to see more engagement from our partners. The big guns in tuberculosis must share the challenges and problems as well as the benefi ts of the partnership. We need to strengthen our collaboration with the Global Fund and ensure countries are really supported, but we should look to engaging new potential donors and partners in tuberculosis control like the BRICS countries who will put money in their own countries directly rather than handing it over to Geneva”, she explains. This year’s theme for World TB Day is innovation. “It is of course about new tools and research—discovering and using them—but is also about using what we have now and thinking more creatively”, Ditiu says. “Look at case detection which has stagnated around 60% for the past several years. How do we reach the di cult cases and engage them into treatment?” Her experience leading the TB REACH initiative—awarding grants of up to $1 million to organisations for innovative ways of reaching the poor and vulnerable—has yielded impressive solutions. Ditiu’s humane approach to TB care and control is her unique strength. Muhwa Jeremiah Chakaya, from Kenya’s Ministry of Health and a Stop TB member, remarks: “Lucica has bags of passion and mountains of energy, just what we need at the Stop TB Partnership. I hope her passion will spread like wildfire, for the enemy, tuberculosis, is regrouping and becoming more ferocious. This enemy can only be tamed by a passionate and truly energised group of warriors, and I am glad that Lucica will be among the generals in this war.”

Stop TB Partnership

Profile Lucica Ditiu: putting people at the forefront of tuberculosis care

Published Online March 24, 2011 DOI:10.1016/S01406736(11)60397-8 For Stop TB Partnership see http://www.stoptb.org

Pamela Das [email protected]

1067

Perspectives

The Art of Medicine Paying attention to the bipartite brain

2011 The M.C. Escher Company-Holland. All rights reserved. www.mcescher.com

What I am going to say will appal many right-thinking scientists. I believe two things that put me beyond the pale. First, I believe science is inextricably bound up with philosophy—with questions such as who we are, what the world is like, and how we should act—and can’t be free of it. The contrary position is itself just another philosophical position, although a dangerously unexamined one. Worse, I believe that we need crucially to attend to an aspect of the structure of the brain: its bipartite nature. If you are still reading, let me explain, since it has consequences for the way we practise medicine and think of ourselves, our minds, and bodies. I have been studying the di erence between the cerebral hemispheres for 20 years. To do so is not to “dichotomise”. Nature dichotomised before us, and even intensifi ed the division over the course of evolution—a strange thing to do with an organ the purpose of which is to make connections, unless the division is a sign of something crucially important at stake. We can’t a ord to ignore it. All that has been discovered in the past 40 years about the way in which the two hemispheres work shows that the di erence does not lie, as once was thought, in what either hemisphere does. Each hemisphere, we now know, is crucially involved in absolutely everything, including language, maths, reasoning, emotion, and visual imagery. It turns out that it’s not what it does, but the way it does

M C Escher, Drawing Hands (1948)

1068

it, that matters. Each hemisphere has a di erent “take” on the world, and that di erence has many important consequences. Why has this come about? Imagine a chick, trying to pick out a seed of corn against the background of gravel on which it lies. For this it requires a certain kind of attention: narrowly focused, guided towards something that is known already to be of interest. But it also needs to have a completely di erent attention to the world at large, for whatever else may come along, without preconception, be it friend or foe. Unless it can do both at once, it will end up being someone else’s lunch while it gets its own. It is virtually impossible to run these two types of consciousness simultaneously in one “brain”. Birds and animals, like us, solve this problem by using the left hemisphere to provide narrow attention to what we can use, our food or prey, and the right to provide the broad sustained awareness of the whole picture. It is no accident that for most of us it is the left hemisphere that controls the hand with which we grasp things, and bring to bear those aspects of language that are concerned with pinning something down, so that we can grasp it. Attention is reciprocally related to what exists: it’s not just that we attend di erently depending on what we fi nd, but that what we fi nd depends on the kind of attention we pay. This means that the two hemispheres produce nothing less than two kinds of world. In the one, the world is present: we experience—the live, complex, embodied, world of individual, unique beings, forever in flux, a net of interdependencies, forming and reforming wholes, a world with which we are deeply connected. In the other the world is “re-presented”: we “experience” our experience in a special way, rather like reading a map. As with a map, less can be more. Here there are, for the fi rst time, static, separable, bounded, but essentially fragmented entities, grouped into classes, on which predictions can be based. This kind of attention isolates, fi xes, and makes each thing explicit by bringing it under the spotlight of attention. In doing so it renders things inert, mechanical, lifeless: it denatures them. But it also enables us to learn more easily, and to see how to make things by putting them together. This gives us power. We need both types of attention. But their relation is not symmetrical. Several di erent lines of argument converge to show that the right hemisphere is aware of, and understands, more than the left: but the left is more able to articulate and use what it knows. The right hemisphere grounds what then gets to be processed, at an intermediate level, by the left hemisphere, before returning to the right hemisphere for integration into www.thelancet.com Vol 377 March 26, 2011

Perspectives

the rest of what we know, in order to make sense of it. Take language. From David McNeill’s subtle, painstaking observations of the relation between language and gesture, it seems that thought originates in processes which are best carried out in the right hemisphere, and is then passed to the left for syntactical and semantic processing. But expression is hugely enriched by what is then added by the right hemisphere—tone, humour, metaphorical understanding, and the implicit meaning that makes up most of what we communicate when we talk. So with everything, it seems: grounded in the right, “unpacked” for practical purposes in the left, then reintegrated into the whole picture by the right again. I am reminded of the relation between Newtonian and quantum mechanics. The “laws” that work well in the middle ground, at the every day level, and which seem so certain there, turn out to be false outside that context. They are just useful approximations, that help us manipulate the world, but radically mistake its fundamental nature. Each hemisphere delivers a vital aspect of experience, and nothing good comes from relying over much on one alone. But that itself is the view of only one of the hemispheres. The other thinks that it can go it alone. The catch is that the left hemisphere exists to make things certain. It has a tendency to prioritise its theory over experience, “re-presentation” over the presence of things. It has a tendency, familiar to all who have cared for patients with a right hemisphere stroke, to deny its own limitations. It is not a reliable guide to the nature of the world, ourselves or our bodies: it is literally most at home with machines. Yet its mechanistic vision is seductive. When we say we understand something, what we mean is that we have likened it to something else that we think we know better. In the case of the brain, or consciousness— even of our bodies—it is not clear what that model should be. If we don’t think about it, we just assume the machine model. The more we see that way, the more that is all that we will see, and hence the better fi t the world seems to be to our model. We close down perception to a certainty, instead of opening it up to possibility. But the left hemisphere is not “science”. The origins of science lie in open-mindedness, fl exibility in applying models, empirical observation of the world of experience, what comes to us through our senses, understood as part of the context in which it lies. Science, just as much as the arts, starts with the processes best served by the right hemisphere. Later, pieces of information—data— are “verifi ed” by repetition of certain procedures out of context, a vitally important stage. But the information is not any more than temporarily “verifi ed”: it gives us something to work on according to a certain model. But change the model, change the context, and it too must change. There just is no certainty. www.thelancet.com Vol 377

What has this got to do with medicine? Everything. Medicine is not about bodies, brains, or minds, but about human beings, their lives and experience. If practised attentively, it will lead to lesser, rather than greater, certainty about what on earth those bodies, brains, and minds might really be like. Certainty is the badge of ignorance. But in order to understand human beings, we need to know something of what are still called the humanities. Some 50 years ago, C P Snow famously alerted us to the growing division between the two cultures of science and the arts. Does not the current interest in all things “neuro” among those in the world of the arts, and the willingness of doctors and scientists, myself included, to engage in discussions with artists, signal a healthy rapprochement? I fear not. It reminds me, sadly, of the bad old days of colonialism, with the “natives” aspiring to dress like the white-suited colonials in order to be taken seriously. It is largely a one-way street, for all the rhetoric that may suggest otherwise. And the arts must share the blame. According to their postmodernist philosophy, there is no truth. Apparently this doesn’t stop us clutching at science to fi ll the vacuum, and tell us what we are. Clearly both are mistaken: there may be no single truth, but no single truth does not mean no truth. It alerts us to be appropriately tentative, careful, though nonetheless determined in our search. But what we fi nd cannot by nature ever be proved. It can only be tested in the cauldron of human experience. Medical education needs urgently to be brought back to the humanities out of which it once arose. Doctors are likely to be e ective in proportion to the degree that they are able to see the broader context in which the complaint brought before them lies—nothing less than the whole world of the patient in front of them. I remember with chagrin how, on “take”, the wards would fi ll with patients who had chest pain or abdominal pain, the majority sent home without a diagnosis. No-one thought of—possibly, it occurs to me now, no-one even knew how to—sit down with them and ask about their lives. Don’t get me wrong: detailed scientifi c knowledge is hugely important. We rely on such minute information to inform the bigger picture. But it is a necessary, not su cient, condition, of being a good physician. Without a way of understanding and interpreting it at a deeper level, more detailed knowledge will achieve precisely nothing, and will lead us ultimately to let our patients down. It will close our reality down into what we imagine to be certain, where an appropriate awareness of the limitations of our knowledge would have liberated us and our patients into a world much richer than we can suspect.

Iain McGilchrist [email protected] http://www.iainmcgilchrist.com

Further reading McGilchrist I. The master and his emissary: the divided brain and the making of the western world. New Haven CT: Yale University Press, 2009 McGilchrist I. Reciprocal organization of the cerebral hemispheres. Dialog Clin Neurosci 2010; 12: 317–34 McNeill D. Hand and mind: what gestures reveal about thought. Chicago: University of Chicago Press, 1992 Rogers LJ, Andrew RJ. Comparative vertebrate lateralization. Cambridge: Cambridge University Press, 2002 Vallortigara G. Comparative neuropsychology of the dual brain: a stroll through animals’ left and right perceptual worlds. Brain and Language 2000; 73: 189–219

1069

victorpatterson.com

Obituary

John Raymond McClean General practitioner and Irish civil rights activist. Born on Jan 16, 1932, in Coleraine, UK, he died on Jan 29, 2011, in Derry/Londonderry, UK, aged 78 years. You might say, borrowing the form but not the content of Karl von Clausewitz’s celebrated aphorism, that Raymond McClean saw politics as the continuation of medicine by other means. The condition of the housing he encountered while doing home visits as a general practitioner (GP) in the Northern Irish city of Derry during the 1960s fired his sense of injustice. This was greatly magnified by the Bloody Sunday shootings of 1972, the culmination of a civil rights march in which McClean was himself taking part. Dismayed not only by the event but also by its aftermath he stood for election to the local council, and went on to become Derry’s first Social Democratic and Labour Party mayor. The Derry in which McClean practised as a physician in the 1960s was very different from the city as it is today. “He was shocked by the housing conditions, with whole families living in single rooms”, says his GP colleague Aine Abbott. “At that time there were a lot of injustices.” Foremost among these was the discrimination against the town’s Catholic and nationalist community. As Joe McEvoy, another of McClean’s GP colleagues, recalls, “There’s a story that he saw a lady in her home, found that she was suffering from TB and advised her to go to bed. She was sitting on a sofa. She said this is my bed. He found there were 26 people living in the one house.” It was the social conditions of Derry that prompted McClean to join the Northern Irish civil rights movement, 1070

and Sunday Jan 30, 1972, found him taking part in a march through the city. As he recounted in his book The Road to Bloody Sunday, the marchers began in something of a carnival mood, anticipating that they would probably be stopped and that there might be “minor rioting”. That this was going to be a march unlike any other began to dawn when he heard shooting. Because he was well known as a doctor, he was called to help. “I spent much of the time passing between the two houses where the dead and the dying were being treated, encouraging the young first-aid workers and others, and wondering at the outward calm they all maintained.” He also wrote of the confused welter of emotions he felt: a mix of fear, anger, and inadequacy. “When the ambulances were loaded and away we stood around the little square in small groups, smoking, subdued and talking in hushed tones. There was an air of unreality hanging over all of us.” The following day McClean received a phone call requesting him to attend the victims’ post mortems as an observer on behalf of the Church. The notes he took of the bleak clinical details as each body was examined appear in an appendix to his book. His offer to give evidence to the Widgery Tribunal set up to report on the shootings was turned down. It was his first inkling that the report was to be wholly inadequate, as is now generally accepted in the wake of the 2010 Saville Inquiry report into Bloody Sunday. McClean studied medicine at Dublin’s Trinity College, graduated in 1958, and then served for 3 years as a medical officer in the Royal Air Force. His wife, whom he met shortly after university, recalls that his aim was always to become a GP. “But Northern Ireland was over-doctored at that time”, she says. “We thought like many others that we might have to emigrate.” They didn’t, but McClean did have to take a job in occupational health for a time. He then worked in several general practices in Derry, the last of which, set up in 1995, was Bayview Medical Centre. Located in the heart of the city, a block or so from the River Foyle, it was here that he was joined by McEvoy and Abbott. “Ray was an extremely gifted communicator and very interested in people”, says Abbott. “He inspired trust. He had good relationships with patients and with colleagues. A generous man who tended to underplay his own abilities, but an inspirational figure. He made a huge difference in a quiet way.” As McEvoy puts it, “If he thought something should be tackled, he tackled it.” McClean’s enduring motive, whether it was the needs of individual patients or a concern with the civil rights of the entire community, was social justice. Having proved a popular councillor he was asked to consider becoming a Westminster MP. He declined, says his wife. Derry was where he wanted to be. “And he was a GP to the core.” In addition to his wife Sheila, he leaves a son and a daughter.

Geo Watts [email protected]

www.thelancet.com Vol 377

Correspondence

www.thelancet.com Vol 377 March 26, 2011

trial. Indeed, far from protecting the interests of patients participating in research, requirements for written informed consent and the resultant delay in starting treatment could be lethal. Second, the delay in starting treatment can obscure a real treatment benefi t from the administration of a time-critical treatment. In the CRASH-2 trial, the requirement for written informed consent probably means that the trial has underestimated the benefi cial e ect of tranexamic acid in trauma patients with bleeds, which would be given without delay in normal clinical practice. In the context of research involving people who are incapable of giving informed consent, the Declaration of Helsinki7 states that, if no patient representative is available and the

Published Online March 24, 2011 DOI:10.1016/S01406736(11)60317-6 See Comment page 1052 See Articles page 1096

Submissions should be made via our electronic submission system at http://ees.elsevier.com/ thelancet/

Proportion of patients that will benefit from treatment 80 70 Cumulative proportion (%)

Clinical trials are important in improving the safety and e ectiveness of emergency care. Many such trials seek to assess the e ects of timecritical treatments for life-threatening disorders such as traumatic brain injury, severe haemorrhage, or respiratory distress. In general, before patients can be enrolled in such trials, current regulations require that they or their legal representatives provide written informed consent.1,2 Although the requirement for written informed consent can sometimes be waived3— eg, if the patient is unconscious, treatment is urgent, and no relative is available—written consent is usually required in emergency-care research, despite the delays to treatment that this will usually entail. Analysis of data from the Medical Research Council (MRC) CRASH Trial,4 a multicentre randomised controlled trial of corticosteroid administration in acute severe head injury, provides an estimate of the delay associated with the requirement for written consent. On average, compared with hospitals that waived the need for consent, initiation of treatment was delayed by 1·2 h (95% CI 0·7–1·8) in hospitals where written consent from relatives was required.5 This delay would not occur if the treatment was given (or withheld) outside the context of a clinical trial, in normal clinical practice. The delay can be life threatening. The CRASH-2 trial6 showed that giving tranexamic acid to trauma patients with bleeds results in a signifi cant and clinically important reduction in overall mortality (relative risk 0·91, 95% CI 0·85–0·97), as well as in mortality specifi cally ascribed to bleeding (0·85, 0·76–0·96). Further analyses have shown that these benefi cial e ects depend importantly on the promptness with which treatment with tranexamic acid is started. Taking account of the

average delay (1·2 h) associated with the need to obtain written consent in the MRC CRASH trial, we used CRASH-2 data to provide an estimate of the consequences for survival of a (more conservative) 1-h delay resulting from the requirement to obtain written consent. We used a logistic regression model, with mortality due to bleeding as the outcome variable and “treatment group”, “time to treatment”, and an interaction term as explanatory variables. The interaction term from this model estimates how the e ect of treatment (odds ratio) changes with time to treatment. We then calculated the risk of death for patients by treatment group, according to time to treatment. The results are shown in the figure. The green and blue lines in the bottom part of the figure give the risk of death in treated and untreated patients, respectively. The e ect of a treatment delay in treated patients was estimated by applying the odds ratio corresponding to a 1-h treatment delay to the risk of death in the untreated group. The red line gives the estimated risk of death in patients in whom treatment is delayed. The point at which the green and red lines intersect the blue line gives the time when the trial treatment no longer provides patient benefi t. Using these data and data from the CRASH-2 trial on the proportion of patients who arrive at hospital within a given time since injury, we estimate that a 1-h treatment delay reduces the proportion of patients who benefi t from the trial treatment from 63% to 49%. Whereas the relative risk of death from bleeding with tranexamic acid was estimated in the CRASH-2 trial as 0·85 (95% CI 0·76–0·96), the corresponding relative risk in the presence of a 1-h delay is 0·96 (0·86–1·08). The delay from consent rituals in emergency situations has important consequences. First, it results in avoidable mortality and probably morbidity in participants in the

60

E ect of delay

50 40 30 20 10 0

9

Risk of death

8 7 Risk of death (%)

E ect of consent rituals on mortality in emergency care research

6

E ect of delay

5 4 No intervention Intervention not delayed Intervention delayed by 1 h

3 0

0

1

2 3 Hours from injury to arrival at hospital

4

5

Figure: E ect of 1-h delay in start of treatment with tranexamic acid on proportion of patients who will benefi t and on risk of death

1071

Correspondence

research cannot be delayed, the study may proceed without informed consent provided that the specifi c reasons for involving patients with a disorder that renders them unable to give informed consent have been stated in the research protocol and that the study has been approved by a research ethics committee. We argue that the need for an urgent trial treatment, even in patients who are conscious and whose relatives are available, by itself excludes the possibility of fully informed consent. If consent rituals delay the start of a trial treatment such that the treatment e ect could be reduced or obscured, we maintain that seeking consent is actually unethical. There might be other treatments whose benefi ts have been missed or underestimated as a result of insistence on the rituals of informed consent, against the precepts of the Declaration of Helsinki, with resultant avoidable harm to patients. There is little evidence that widely promoted forms of research regulation do more good than harm.8 Informed consent procedures, like other well-intentioned public health interventions, should be assessed rigorously. The lethal e ects we have shown might have been found decades ago had the research ethics community accepted a responsibility to provide robust evidence that its prescriptions are likely to do more good than harm. We declare that we have no confl icts of interest.

*Ian Roberts, David Prieto-Merino, Haleema Shakur, Iain Chalmers, Jon Nicholl [email protected] Clinical Trials Unit, London School of Hygiene and Tropical Medicine, London WC1E 7HT, UK (IR, DPM, HS); James Lind Initiative, Oxford, UK (IC); and School of Health and Related Research, University of She eld, She eld, UK (JN) 1

2

1072

Anon. Medicines for human use (clinical trials) amendment (no. 2) regulations 2006. http://www.legislation.gov.uk/uksi/2006/ 2984/contents/made (accessed Feb 7, 2011). Anon. Medicines for human use (clinical trials) regulations 2004. http://www.legislation.gov. uk/uksi/2004/1031/contents/made (accessed Feb 7, 2011).

3

4

5 6

7

8

Food and Drug Administration. Protection of human subjects: informed consent and waiver of informed consent requirements in certain emergency research. Final rules 21CFR50.24. Fed Register 1996; 61: 51498–533. CRASH Trial Collaborators. E ect of intravenous corticosteroids on death within 14 days in 10,008 adults with clinically signifi cant head injury (MRC CRASH Trial): a randomised placebo-controlled trial. Lancet 2004; 364: 1321–28. CRASH Trial Management Group. Research in emergency situations: with or without relatives’ consent. Emerg Med J 2004; 21: 703. CRASH-2 Collaborators. E ects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with signifi cant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet 2010; 376: 23–32. World Medical Association. Declaration of Helsinki—ethical principles for medical research involving human subjects. http:// www.wma.net/en/30publications/10policies/ b3/index.html (accessed Feb 7, 2011). Chalmers I. Regulation of therapeutic research is compromising the interests of patients. Int J Pharm Med 2007; 21: 395–404.

Risk assessment for recurrent venous thrombosis In the Review entitled “Risk assessment for recurrent venous thrombosis” (Dec 11, p 2032), 1 Paul Kyrle and colleagues report that women who continued hormone intake after a first event of venous thromboembolism (VTE) were at high risk of recurrent VTE. However, findings from a French cohort of postmenopausal women emphasise that the risk of recurrent VTE in users of hormone therapy depended on the route of oestrogen administration. Consistent with previous data,2 women using oral oestrogens after a first thrombotic event were at high risk of recurrent VTE (adjusted hazard ratio 6·4, 95% CI 1·5–27·3). By contrast, transdermal oestrogens were not associated with an increased risk of recurrent VTE compared with non-use (1·0, 0·4–2·4).3 These data provide evidence to support the safety of the transdermal route of oestrogen

administration with respect to the risk of VTE recurrence and add to the current epidemiological evidence that transdermal oestrogens have no effect on thrombotic risk in postmenopausal women. 4,5 In light of these data, the evidence for the implication of continuous oestrogen use in the increased risk of recurrent venous thrombosis should be modulated according to the route of oestrogen administration in postmenopausal women. We declare that we have no conflicts of interest.

*Valérie Olié, Geneviève Plu-Bureau, Marianne Canonico, Pierre-Yves Scarabin [email protected] Cardiovascular Epidemiology Section, Centre for Research in Epidemiology and Population Health, INSERM UMS1018, 94807 Villejuif, France (VO, GPB, MC, PYS); and René Decartes University, Paris V, Paris, France (GPB) 1 2

3

4

5

Kyrle PA, Rosendaal FR, Eichinger S. Risk assessment for recurrent venous thrombosis. Lancet 2010; 376: 2032–39. Hoibraaten E, Qvigstad E, Arnesen H, Larsen S, Wickstrom E, Sandset PM. Increased risk of recurrent venous thromboembolism during hormone replacement therapy—results of the randomized, double-blind, placebocontrolled estrogen in venous thromboembolism trial (EVTET). Thromb Haemost 2000; 84: 961–67. Olie V, Plu-Bureau G, Conard J, Horellou MH, Canonico M, Scarabin PY. Hormone therapy and recurrence of venous thromboembolism among postmenopausal women. Menopause (in press). Olie V, Canonico M, Scarabin PY. Risk of venous thrombosis with oral versus transdermal estrogen therapy among postmenopausal women. Curr Opin Hematol 2010; 17: 457–63. Scarabin PY, Oger E, Plu-Bureau G. Differential association of oral and transdermal oestrogen-replacement therapy with venous thromboembolism risk. Lancet 2003; 362: 428–32.

Paul Kyrle and colleagues1 suggest that patients with D-dimer concentrations of more than 250 g/L are at risk of recurrence of idiopathic venous thrombosis and thus candidates for resuming anticoagulation therapy. Although a meta-analysis of seven randomised controlled trials showed that a positive D-dimer result after stopping anticoagulation therapy was associated with a 2·5-fold increase www.thelancet.com Vol 377 March 26, 2011

Correspondence

I declare that I have no confl icts of interest.

5

Carrier M, Le Gal G, Wells PS, Rodger MA. Systematic review: case-fatality rates of recurrent venous thromboembolism and major bleeding events among patients treated for venous thromboembolism. Ann Intern Med 2010; 152: 578–89.

We applaud Paul Kyrle and colleagues1 for their compre hensive Review of risk assessment for recurrent venous thrombosis (VT). It is clear that “VT begets VT” and that some disorders— eg, cancer—increase the risk of VT recurrence. In addition to risk factors highlighted in the Review, we would like to draw attention to the association between (recurrent) VT and chronic infl ammatory diseases. Infl ammatory bowel disease has been linked to increased risk of VT, especially during disease fl areups.2 Furthermore, atherosclerosis—a chronic infl amma tory disease—shares risk factors and pathophysiological mechanisms with VT, and other chronic infl ammatory diseases such as rheumatoid arthritis and psoriasis have been associated with VT and athero thrombosis.3–5 Therefore, evidence suggests that these chronic infl ammatory diseases should be added to the list of disorders associated with increased risk of recurrent VT. Future studies are likely to shed more light on this interesting association.

Tetsuji Fujita

We declare that we have no confl icts of interest.

[email protected] Department of Surgery, Jikei University School of Medicine, Tokyo 105-8461, Japan

*Ole Ahleho , Gunnar Hilmar Gislasson, Christian Torp-Pedersen, Peter Riis Hansen

1

[email protected]

2

3

4

Kyrle PA, Rosendaal FR, Eichinger S. Risk assessment for recurrent venous thrombosis. Lancet 2010; 376: 2032–39. Verhovsek M, Douketis JD, Yi Q, et al. Systematic review: D-dimer to predict recurrent disease after stopping anticoagulant therapy for unprovoked venous thromboembolism. Ann Intern Med 2008; 149: 481–90. Fancher TL, White RH, Kravitz RL. Combined use of rapid D-dimer testing and estimation of clinical probability in the diagnosis of deep vein thrombosis: systematic review. BMJ 2004; 329: 821. Cosmi B, Legnani C, Tosetto A, et al. Usefulness of repeated D-dimer testing after stopping anticoagulation for a first episode of unprovoked venous thromboembolism: the PROLONG II prospective study. Blood 2010; 115: 481–88.

www.thelancet.com Vol 377 March 26, 2011

Department of Cardiology, Copenhagen University Hospital Gentofte, 2900 Hellerup, Denmark. 1 2

3

4

Kyrle PA, Rosendaal FR, Eichinger S. Risk assessment for recurrent venous thrombosis. Lancet 2010; 376: 2032–39. Grainge MJ, West J, Card TR. Venous thromboembolism during active disease and remission in infl ammatory bowel disease: a cohort study. Lancet 2010; 375: 657–63. Ageno W, Dentali F. Venous thromboembolism and arterial thromboembolism: many similarities, far beyond thrombosis per se. Thromb Haemost. 2008; 100: 181–83. Matta F, Singala R, Yaekoub AY, Najjar R, Stein PD. Risk of venous thromboembolism with rheumatoid arthritis. Thromb Haemost 2009; 101: 134–38.

5

Karabudak O, Ulusoy RE, Erikci AA, Solmazgul E, Dogan B, Harmanyeri Y. Inflammation and hypercoagulable state in adult psoriatic men. Acta Derm Venereol 2008; 88: 337–40.

Authors’ reply In a large case-control study from France, transdermal oestrogens did not increase the risk of a first venous thrombosis.1 Valérie Olié and co-workers now provide evidence that, by contrast with continued use of oral oestrogens, transdermal application of oestrogens after a first episode of venous thrombosis does not confer an increased risk of recurrence. However, we would still be hesitant to treat women with a history of venous thrombosis with transdermal oestrogens until more evidence is provided that this is safe. Clearly, we do not regard all patients with a high D-dimer concentration shortly after withdrawal of anticoagulants as candidates for extended secondary thromboprophylaxis as stated by Tetsuji Fujita. The positive predictive value of a high D-dimer concentration regarding recurrent venous thrombosis is low and thus not helpful for clinical decision making.2 Measurement of D-dimer concentrations might be more useful in identifying patients at low risk of recurrence in whom the risk of bleeding during long-term anticoagulation might be higher than the risk of recurrence after stopping treatment. Residual vein thrombosis as detected by repeated venous ultrasound testing is possibly a good predictor of recurrence. However, as stated in our Review, this approach is hampered by the fact that uniformly acknowledged criteria for defi nition of vein recanalisation are lacking. We agree with Ole Ahleho and coworkers that some chronic inflammatory diseases, in particular inflammatory bowel disease, are associated with an increased risk of venous thrombosis. Indeed, we have reported that the probability of recurrence 5 years after discontinuation of anticoagulation therapy was 33·4% (95% CI 21·8–45·0)

Science Photo Library

in annual risk of disease recurrence,2 these trials used di ering D-dimer assays. Rapid whole-blood qualitative assays, which can be run and interpreted at the bedside, seem to be commonly used for detecting abnormal D-dimer concentrations, because prompt management of patients with increased D-dimer concentrations might be necessary. These assays have a high sensitivity but a relatively low specifi city for venous thrombosis, and the cut-o value for a positive test result ranges from 80 g/L to 500 g/L.3,4 A normal result from the highly sensitive D-dimer test would safely exclude high-risk patients, but patients with a positive result are not necessarily at high risk of recurrence of venous thrombosis. In a further meta-analysis,5 the proportion of patients who died of recurrent venous thrombosis decreased from 11·3% during anticoagulation therapy to 3·6% after anticoagulation, whereas the case– fatality rate of major bleeding events during anticoagulation was 11·3%.5 In addition to D-dimer testing, repeated venous ultrasound imaging would be warranted to balance the risks and benefi ts of anticoagulation therapy.

1073

Correspondence

in patients with inflammatory bowel disease and 21·7% (18·8–24·6; p=0·01) in those without the disease.3 After adjustment for potential confounders, inflammatory bowel disease was an independent risk factor for recurrent venous thrombosis, with a hazard ratio of 2·5 (1·4–4·2).3 Patients with inflammatory bowel disease could be candidates for extended anticoagulation, but interventional trials to investigate the e cacy and safety of long-term anticoagulant therapy are absent in these patients. We declare that we have no confl icts of interest.

*P A Kyrle, S Eichinger [email protected] Department of Medicine I, Medical University of Vienna, 1090 Vienna, Austria 1

2

3

Canonico M, Oger E, Plu-Bureau G, et al. Hormone therapy and venous thromboembolism among postmenopausal women. Circulation 2007; 115: 840–45. Le Gal G, Bounameuax H. D-Dimer testing to predict recurrence risk in venous thromboembolism: looking for a useful threshold: a rebuttal. J Thromb Haemost 2004; 2: 1670–72. Novacek G, Weltermann A, Sobala A, et al. Infl ammatory bowel disease is a risk factor for recurrent venous thromboembolism.! Gastroenterology 2010; 139: 779–87.

Renal sympathetic denervation for refractory hypertension The Symplicity HTN-2 Investigators’ results of renal sympathetic denervation for medication-resistant hypertension (Dec 4, p 1903)1 show a striking reduction in systolic and diastolic blood pressure at 6 months compared with age-matched, equally hypertensive controls. Despite the short follow-up of this study, excitement exists in the potential for this treatment to reduce blood pressure over a much longer period, analogous to the historic results published by Smithwick and Thompson in 1953 that showed longterm e cacy of sympathectomy for the treatment of hypertension.2 Enthusiasm for the results from Symplicity HTN-2 is tempered by 1074

Change in systolic blood pressure (mm Hg)

Change in diastolic blood pressure (mm Hg)

O ce-based minus home-based (mm Hg)

Homebased minus 24-h ambulatory (mm Hg)

O ce-based minus 24-h ambulatory (mm Hg)

Renal denervation O ce-based (n=49)

–32 (23)

–12 (11)

..

..

..

Home-based (n=32)

–20 (17)

–12 (11)

..

..

..

24-h ambulatory (n=20)

–11 (15)

–7 (11)

..

..

Absolute di erence

12/0

9/5

..

..

..

21/5

O ce-based (n=51)

1 (21)

0 (10)

..

..

..

Home-based (n=40)

2 (13)

0 (7)

..

..

..

–3 (19)

–1 (12)

..

..

Control

24-h ambulatory (n=25) Absolute di erence

..

..

1/0

–5/–1

.. –4/–1

Data are mean (SD).

Table: Intragroup comparison of blood-pressure assessment by means of o ce-based, home-based, and 24-h ambulatory blood pressure monitoring at 6-month follow-up visit within treatment groups

important study limitations, including a concerning finding of incomplete and discrepant results of various blood pressure assessments at 6 months (table). More specifi cally, why the absolute di erence in 6-month blood pressure assessment varied substantially by technique for the treatment group, yet not as strikingly for controls, remains unclear. Although o cebased blood pressure provided the sole assessment of e cacy as a primary endpoint, this intragroup variability of blood pressure assessment requires further explanation, particularly since 35% of controls enjoyed a 10 mm Hg or more decrease in systolic blood pressure at 6 months. Despite these shortcomings, the Symplicity HTN-2 Investigators submit provocative data on the treatment of medication-refractory hypertension with decidedly more striking bloodpressure-lowering results than that seen in many other randomised trials of antihypertensive medications.3 Further analysis of this technology in a large and more diverse population is needed. I declare that I have no confl icts of interest.

Matthew C Bunte [email protected] Robert & Suzanne Tomsich Department of Cardiovascular Medicine, Cleveland Clinic, Cleveland, OH 44195, USA

1

2 3

Symplicity HTN-2 Investigators. Renal sympathetic denervation in patients with treatment-resistant hypertension (The Symplicity HTN-2 Trial): a randomised controlled trial. Lancet 2010; 376: 1903–09. Smithwick RH, Thompson JE. Splanchniectomy for essential hypertension: results in 1,266 cases. JAMA 1953; 152: 1501–04. Wright JM, Musini VM. First-line drugs for hypertension. Cochrane Database Syst Rev 2009; 3: CD001841.

Author’s reply The essence of Matthew Bunte’s communication is that the di erences in trial blood pressures recorded with the three methods used—automated clinic pressure with printout (the primary trial endpoint), patients’ self-recorded home blood pressure, and ambulatory 24-h blood pressure—perhaps should temper any conclusion that the trial was in any sense definitive. In this context, Bunte sees the contrasting absence of di erences in blood-pressure change in the control patients with the di erent methods of measurement as anomalous, and requiring explanation. Explanation of the latter point is straightforward; there was no net change in blood pressure in the control group, providing no scope for di erences between the methods! Bunte correctly records in his table the material di erences between recorded blood-pressure change after renal denervation according to measurement www.thelancet.com Vol 377 March 26, 2011

Correspondence

100

Renal denervation

Control

≥180 mm Hg 160–179 mm Hg 140–159 mm Hg <140 mm Hg

90 Proportion of patients (%)

80 70 60 50 40 30 20 10 0

Baseline

6 months

Baseline

6 months

Figure: Distribution of o ce systolic blood pressures in denervation and control groups of Symplicity HTN-2 Trial at baseline and at 6-month assessment point In 82% of denervated patients, systolic blood pressure at 6 months was less than the specified minimum pressure for trial entry (≥160 mm Hg, for non-diabetics). The corresponding proportion in the control group was 24%.

technique. These di erences might perhaps have arisen in part from the fact that the blood pressure datasets were not complete for the secondary endpoints of home and ambulatory pressures with, in particular, collection of technically satisfactory ambulatory paired records before and 6 months after renal denervation being achieved in only 20 of 49 denervated patients. Data defi ciencies could have distorted the calculated group mean ambulatory blood pressure response, although of course it is common experience that recorded ambulatory clinical trial blood-pressure changes are often less than o ce blood-pressure changes. Bunte correctly notes that in 35% of control patients, without denervation, there was a more than 10 mm Hg decrease in o ce blood pressure at 6 months. These changes in blood pressure in the control group are normally distributed, and certainly di erent from those seen in the denervation group (figure). Additionally, Bunte makes a case for further assessment of renal denervation, and in a more diverse population. We agree. A US-based trial is planned for the near future, larger than the present one, which will include patients of di ering ethnic origin. We look forward to US investigators doing what we expect will be a confirmatory study, www.thelancet.com Vol 377 March 26, 2011

and hope that clinical practice in the USA will not lag behind that in Australia and Europe, where renal denervation with the Symplicity catheter has received regulatory approval and is now entering clinical practice. I have received consulting fees and travel expenses, and my institution has received a research grant, from Ardian.

Murray Esler [email protected] Baker Medical Research Institute, PO Box 6492, St Kilda Road Central Melbourne, Melbourne, VIC 8008, Australia

tolerated statin dose and an LDL cholesterol concentration has been established, how worthwhile it is to measure this annually, since without treatment change it is unlikely to alter. The Quality and Outcomes Framework (QOF) in the UK encourages physicians and practice nurses to treat to laboratory values, with targets that are met being rewarded annually. There would be an immense cost saving for the laboratory and the primary-care practice (which will have many freed-up appointments) if we were to limit checks on biochemical profile, thyroid function, and lipids to before and after statin treatment, and after that only if clinically indicated. Perhaps QOF should be rewritten to include only measurement of bodymass index on a simple weighing scale, since encouragement of lifestyle changes will have a broader e ect on health than blood tests alone. It would demedicalise the issue and put the onus on the patient rather than the practice. Policy makers should focus on population intervention, diet, and exercise, rather than on trying to measure the performance of a practice, which is likely to be beyond the practice’s control. I declare that I have no confl icts of interest.

H de Vries [email protected]

Raised liver enzymes in patients taking statins The Article on statin use in patients with raised liver enzyme concentrations by Vasilios Athyros and colleagues (Dec 4, p 1916)1 and the Comment from Ted Bader (p 1882)2 raise a very important issue for all of us who work in primary care. Apart from the fact that statins can actually improve marginally elevated liver enzymes, for abnormalities that are most likely to be caused by unhealthy lifestyles choices, the findings raise the question of how valid it is to measure liver enzyme concentrations on a routine basis. Likewise I also wonder, once a patient is on a maximum-

City Health Care Partnership, Priory Park East, Kingston-upon-Hull HU4 7DY, UK 1

2

Athyros VG, Tziomalos K, Gossios TD, et al. Safety and e cacy of long-term statin treatment for cardiovascular events in patients with coronary heart disease and abnormal liver tests in the Greek Atorvastatin and Coronary Heart Disease Evaluation (GREACE) Study: a post-hoc analysis. Lancet 2010; 376: 1916–22. Bader T. Liver tests are irrelevant when prescribing statins. Lancet 2010; 376: 1882–83.

Authors’ reply As H de Vries states, there is ongoing debate about monitoring of patients on statins. We suggest that clinicians follow the recommendations of one of several national guidelines. In the UK there is a proliferation of local guidelines. A more national focus is desirable to avoid confusion and 1075

Correspondence

Science Photo Library

ensure equal standards of treatment and safety. Furthermore, eliminating local guidelines will reduce costs. Regarding monitoring once cholesterol targets are achieved, we need to consider several factors. First, patients might gradually “drift” away from lipid goals and the statin dose will therefore need to be adjusted. Sustained adherence to lifestyle measures is often not achieved. Second, there is a trend towards development of diabetes with age, and statins could increase the risk of diabetes.1 Occasional monitoring (with labora tory tests) should detect new-onset diabetes. Third, there is a relatively high prevalence of hypothyroidism in older women,2 some of whom will be taking statins. In turn, hypothyroidism is a cause of hypercholesterolaemia and statin use in patients with hypothyroidism can increase the risk of myopathy.3 Fourth, the liver–statin story is not definitively resolved by a small post hoc analysis.4 Moreover, seven patients of 880 were withdrawn from the GREACE trial because of liver-related adverse e ects attributed to statins.4 This finding has to be considered in the context of millions of patients taking these drugs worldwide. The long-term implications of such adverse e ects are not known. Finally, renal function deteriorates with age (or pathology) and is another predictor of statin-associated myopathy.3 Some statins are safer than others in patients with impaired renal function.5 Therefore, there might be a need to change medication. Although our response does not resolve the issues raised by de Vries, we hope that this correspondence and our paper4 will focus attention on this important topic. We declare that we have no confl icts of interest. Published Online November 19, 2010 DOI:10.1016/S01406736(10)62138-1

Vasilios G Athyros, Asterios Karagiannis, Konstantinos Tziomalos, Thomas D Gossios, *Dimitri P Mikhailidis [email protected] Second Propedeutic Department of Internal Medicine, Medical School, Aristotle University o\f Thessaloniki, Hippokration Hospital, Thessaloniki, Greece (VGA, AK); First Propedeutic Department of Internal

1076

Medicine (KT) and First Cardiology Clinic (TDG), Medical School, Aristotle University of Thessaloniki, AHEPA University Hospital, Thessaloniki, Greece; and Department of Clinical Biochemistry, Royal Free Hospital Campus, University College London Medical School, London NW3 2QG, UK (DPM) 1

2

3 4

5

Athyros VG, Tziomalos K, Karagiannis A, Mikhailidis DP. Preventing type 2 diabetes mellitus: room for residual risk reduction after lifestyle changes? Curr Pharm Des 2010; 16: 3939–847. McGrogan A, Seaman HE, Wright JW, de Vries CS. The incidence of autoimmune thyroid disease: a systematic review of the literature. Clin Endocrinol 2008; 69: 687–96. Antons KA, Williams CD, Baker SK, Phillips PS. Clinical perspectives of statin-induced rhabdomyolysis. Am J Med 2006; 119: 400–09. Athyros VG, Tziomalos K, Gossios TD, et al, for the GREACE Study Collaborative Group. Safety and e cacy of long-term statin treatment for cardiovascular events in patients with coronary heart disease and abnormal liver tests in the Greek Atorvastatin and Coronary Heart Disease Evaluation (GREACE) Study: a post-hoc analysis. Lancet 2010; 376: 1916–22. Athyros VG, Mikhailidis DP, Papageorgiou AA, et al. The e ect of statins versus untreated dyslipidaemia on renal function in patients with coronary heart disease: a subgroup analysis of the Greek atorvastatin and coronary heart disease evaluation (GREACE) study. J Clin Pathol 2004; 57: 728–34.

detect, and manage resident-toresident abuse, and experienced care sta can provide helpful insights into how to develop these strategies.4 However, it would be important not to assume that only those with a past history of paedophilia are likely to be o enders, since vulnerability and inadequate vigilance about access might be more important factors than an agebased “gerontophilia”. Additionally, we are as yet far from understanding how best to meet the sexual and emotional needs of residents in nursing homes.5 The expression of sexuality by nursing home residents often creates discomfort among care sta , and fresh thinking and multidisciplinary strategies are required to address the emotional and sexual wellbeing of older people in residential care. We declare that we have no confl icts of interest.

*Sean Kennelly, Desmond O’Neill, James G O’Brien [email protected]

Elder abuse in residential care Jean-Claude Monfort and colleagues (Jan 22, p 300)1 bring helpful attention to neglected aspects of elder abuse. Elder mistreatment in institutions is poorly catalogued,2 and is especially complex in consisting of both institutional abuse—towards all residents—and acts of elder abuse specifi cally towards a single resident. Monfort and colleagues’ Correspondence raises the further challenge of abuse by other residents, who, along with care sta , are recognised as the most common perpetrators of sexual abuse of nursing home residents.3 Most definitions of elder abuse are premised on their occurring within a relationship where there is an expectation of trust, and the precise nature of the relationship between residents, particularly when one or both might have cognitive impairment, is not at all clear. More attention needs to be paid to developing techniques to prevent,

Department of Medical Gerontology, Trinity College Dublin, Ireland (SK, DON); and Department of Geriatrics, University of Louisville, KY, USA (JGOB) 1 2

3

4

5

Monfort JC, Villemur V, Lezy AM, Baron-Laforet S, Dröes RM. From paedophilia to gerontophilia. Lancet 2011; 377: 300. O’Dwyer C, O’Neill D. Developing strategies for the prevention, detection and management of elder abuse: the Irish experience. J Elder Abuse Negl 2008; 20: 169–80. Ramsey-Klawsnik H, Teaster PB, Mendiondo MS, Marcum JL, Abner EL. Sexual predators who target elders: findings from the first national study of sexual abuse in care facilities. J Elder Abuse Negl 2008; 20: 353–76. Rosen T, Lachs MS, Bharucha AJ, et al. Resident-to-resident aggression in long-term care facilities: insights from focus groups of nursing home residents and sta . J Am Geriatr Soc 2008; 56: 1398–408. Rosen T, Lachs MS, Pillemer K. Sexual aggression between residents in nursing homes: literature synthesis of an underrecognized problem. J Am Geriatr Soc 2010; 58: 1970–79.

Department of Error Mullan F, Frehywot S, Omaswa F, et al. Medical schools in sub-Saharan Africa. Lancet ; 377: – —In this Health Policy paper (published online Nov 11, 2010), the acknowledgments section should have read: “This work was supported by the Bill & Melinda Gates Foundation”. This correction has been made to the online version as of Nov 19, 2010, and to the printed paper.

www.thelancet.com Vol 377 March 26, 2011

Articles

A 2-h diagnostic protocol to assess patients with chest pain symptoms in the Asia-Pacific region (ASPECT): a prospective observational validation study Martin Than, Louise Cullen, Christopher M Reid, Swee Han Lim, Sally Aldous, Michael W Ardagh, W Frank Peacock, William A Parsonage, Hiu Fai Ho, Hiu Fai Ko, Ravi R Kasliwal, Manish Bansal, Sunarya Soerianata, Dayi Hu, Rongjing Ding, Qi Hua, Kang Seok-Min, Piyamitr Sritara, Ratchanee Sae-Lee, Te-Fa Chiu, Kuang-Chau Tsai, Fang-Yeh Chu, Wei-Kung Chen, Wen-Han Chang, Dylan F Flaws, Peter M George, A Mark Richards

Summary

Background Patients with chest pain contribute substantially to emergency department attendances, lengthy hospital stay, and inpatient admissions. A reliable, reproducible, and fast process to identify patients presenting with chest pain who have a low short-term risk of a major adverse cardiac event is needed to facilitate early discharge. We aimed to prospectively validate the safety of a predefined 2-h accelerated diagnostic protocol (ADP) to assess patients presenting to the emergency department with chest pain symptoms suggestive of acute coronary syndrome. Methods This observational study was undertaken in 14 emergency departments in nine countries in the AsiaPacific region, in patients aged 18 years and older with at least 5 min of chest pain. The ADP included use of a structured pre-test probability scoring method (Thrombolysis in Myocardial Infarction [TIMI] score), electrocardiograph, and point-of-care biomarker panel of troponin, creatine kinase MB, and myoglobin. The primary endpoint was major adverse cardiac events within 30 days after initial presentation (including initial hospital attendance). This trial is registered with the Australia-New Zealand Clinical Trials Registry, number ACTRN12609000283279. Findings 3582 consecutive patients were recruited and completed 30-day follow-up. 421 (11·8%) patients had a major adverse cardiac event. The ADP classified 352 (9·8%) patients as low risk and potentially suitable for early discharge. A major adverse cardiac event occurred in three (0·9%) of these patients, giving the ADP a sensitivity of 99·3% (95% CI 97·9–99·8), a negative predictive value of 99·1% (97·3–99·8), and a specificity of 11·0% (10·0–12·2). Interpretation This novel ADP identifies patients at very low risk of a short-term major adverse cardiac event who might be suitable for early discharge. Such an approach could be used to decrease the overall observation periods and admissions for chest pain. The components needed for the implementation of this strategy are widely available. The ADP has the potential to affect health-service delivery worldwide. Funding Alere Medical (all countries), Queensland Emergency Medicine Research Foundation and National Health and Medical Research Council (Australia), Christchurch Cardio-Endocrine Research Group (New Zealand), Medquest Jaya Global (Indonesia), Science International (Hong Kong), Bio Laboratories Pte (Singapore), National Heart Foundation of New Zealand, and Progressive Group (Taiwan).

Introduction Every year, an estimated 5–10% of presentations to emergency departments, and up to a quarter of hospital admissions are attributable to symptoms suggestive of acute coronary syndromes.1 Patients with a missed diagnosis of acute myocardial infarction are at increased risk of a major adverse cardiac event. The need for safe discharge without a substantial risk of a major adverse cardiac event is a priority and a driver of clinician behaviour. Consequently, most patients with symptoms suggestive of acute coronary syndromes undergo lengthy assessment, either in the emergency department or as hospital inpatients, even though 75–85% of these patients ultimately do not have a final diagnosis of acute coronary syndromes.2–4 The assessment processes vary between institutions, with no one process being ideal. www.thelancet.com Vol 377 March 26, 2011

Present recommendations are for serial sampling of cardiac troponin over at least 6 h from the onset of symptoms.5–7 Concerns about accuracy of patients’ recall of events has led many centres to time troponin sampling from the moment of presentation to the emergency department.8 Prolonged assessment contributes to overcrowding in the hospital or department, physician duplication of effort, and clinical risk as patients are treated by different clinical staff.1 Emergency department overcrowding is associated with increased costs and adverse patient outcomes, including increased mortality.9 A reliable, reproducible, and more timely process for the identification of chest pain presentations that have a low short-term risk of a major adverse cardiac event is needed to facilitate earlier discharge.4 Accelerated diagnostic

Lancet 2011; 377: 1077–84 Published Online March 23, 2011 DOI:10.1016/S01406736(11)60310-3 See Comment page 1049 Christchurch Hospital, Christchurch, New Zealand (M Than MBBS, S Aldous MBChB, Prof P M George MBBS); Royal Brisbane and Women’s Hospital, Herston, QLD, Australia (L Cullen MBBS, W A Parsonage MD); Monash University, Melbourne, VIC, Australia (Prof C M Reid PhD); Singapore General Hospital, Singapore, Singapore (Prof S H Lim MBBS); University of Otago, Christchurch, New Zealand (Prof M W Ardagh PhD, Prof A M Richards MD); Cleveland Clinic Foundation, Cleveland, OH, USA (W F Peacock MD); Queen Elizabeth Hospital, Hong Kong, China (H F Ho MD, H F Ko MD); Medanta—The Medicity, Gurgaon, India (R R Kasliwal MD, M Bansal MD); National Cardiovascular Centre Harapan Kita, Jakarta, Indonesia (S Soerianata MD); People’s Hospital, Beijing, China (Prof D Hu MD, R Ding MD); Xuanwu Hospital, Beijing, China (Q Hua MD); Severence Hospital, Seoul, South Korea (Prof K Seok-Min MD); Ramathibodi Hospital, Bangkok, Thailand (Prof P Sritara MD, Prof R Sae-Lee MD); Chang Gung Memorial Hospital, Keelung, Taiwan (T-F Chiu MD); Far Eastern Memorial Hospital, Taipei, Taiwan (K-C Tsai MD, F-Y Chu MD); China Medical University Hospital, Taichung, Taiwan (W-K Chen MD); MacKay Memorial Hospital, Taipei, Taiwan (W-H Chang DrPH); Queensland University of Technology, Brisbane, QLD, Australia (L Cullen); and University of

1077

Articles

Queensland, Brisbane, Australia (D F Flaws MSc) Correspondence to: Dr Martin Than, Department of Emergency Medicine, Canterbury District Health Board, Private Bag 4710, Christchurch, New Zealand [email protected]

Panel 1: The TIMI score for unstable angina or non-ST elevation myocardial infarction15 (1) Age 65 years or older (2) Three or more risk factors for coronary artery disease (family history of coronary artery disease, hypertension, hypercholesterolaemia, diabetes, or being a current smoker) (3) Use of aspirin in the past 7 days (4) Significant coronary stenosis (eg, previous coronary stenosis ≥50%) (5) Severe angina (eg, two or more angina events in past 24 h or persisting discomfort) (6) ST-segment deviation of 0·05 mV or more on first electrocardiograph (7) Increased troponin and/or creatine kinase MB on initial blood tests* The TIMI score had to be zero for the sum of its seven parameters to be categorised as 0. TIMI=Thrombolysis In Myocardial Infarction. *Point-of-care values were used for TIMI score calculation.

protocols (ADPs), clinical decision rules, and prediction rules are terms for processes or methods intended to help clinicians to make bedside diagnostic and therapeutic decisions. They involve variables from the patient’s history and examination, and often incorporate the results of diagnostic tests.6 ADPs for chest pain are well established but emphasise the need to assess the patient for at least 6 h after the onset of symptoms.6,10 Some studies have safely investigated patients with serial biomarkers during 1·5–3 h in a low-risk patient group, but have not defined a reproducible method to identify this low-risk group.11 For an assessment of possible acute coronary syndromes, a maximum of 60 min is recommended for the availability of troponin results.12 Many central laboratories have difficulty in meeting this standard. Point-of-care biomarkers represent a possible solution to meeting this target. The Thrombolysis In Myocardial Infarction (TIMI) score for unstable angina or non-ST elevation myocardial infarction is an externally validated and widely used structured risk assessment method.3,13,14 Its use in conjunction with serial 0–2 h biomarker testing

3853 eligible patients

202 declined consent

3651 consenting eligible patients 21 excluded because TIMI score incomplete

3630 had ADP index test

0 inconclusive result

3260 were ADP positive: not low risk

370 were ADP negative: low risk 30 lost to follow-up

3230 completed 30-day follow-up

418 had 30-day MACE

2812 did not have 30-day MACE

18 lost to follow-up 352 completed 30-day follow-up

3 had 30-day MACE

349 did not have 30-day MACE

Figure 1: Trial profile of participant recruitment and outcomes according to ADP classification 30-day follow-up includes initial hospital attendance. Patients lost to follow-up did not have a MACE during initial hospital attendance. TIMI=Thrombolysis In Myocardial Infarction score for unstable angina or non-ST-elevation myocardial infarction. ADP=accelerated diagnostic protocol. MACE=major adverse cardiac event.

1078

www.thelancet.com Vol 377 March 26, 2011

Articles

(either via central laboratory or point-of-care systems) and electrocardiograph (ECG) has not been prospectively tested. Importantly, there has been little validation of ADPs based in emergency departments outside North America, or in diverse population groups such as the Asia-Pacific population, in whom a mix of ethnic backgrounds and variations in service delivery introduce important differences.15 The ASia-Pacific Evaluation of Chest pain Trial (ASPECT) was a prospective observational validation study designed to assess whether a predefined ADP would identify patients presenting to the emergency department with chest pain, who would be at low risk of harm if they were to be discharged early.

Methods

Participants Enrolment occurred at 14 urban emergency departments in nine countries in the Asia-Pacific region (Australia, China [including Hong Kong], India, Indonesia, New Zealand, Singapore, South Korea, Taiwan, and Thailand). Patients were included if they were at least 18 years old and had at least 5 min of chest pain (or discomfort) suggestive of acute coronary syndromes for whom the attending physician planned to investigate for these syndromes with serial biomarker tests. In accordance with American Heart Association case definitions,16 possible cardiac symptoms included acute chest; epigastric, neck, jaw, or arm pain; or discomfort or pressure without an apparent non-cardiac source. Generally, atypical symptoms (fatigue, nausea, vomiting, diaphoresis, faintness, and back pain) were not used as inclusion criteria in the absence of chest pain. Patients were excluded if they had an ST-segment elevation acute myocardial infarction, there was a clear cause other than acute coronary syndromes for the symptoms (eg, clinical findings of pneumonia), they were unable or unwilling to provide informed consent, staff considered recruitment to be inappropriate (eg, terminal illness), they were transferred from another hospital, they were pregnant, they were recruited on previous presentation, or they were unable to be contacted after discharge. Perceived high risk was not regarded as an exclusion criterion. Recruitment included consecutive eligible cases at each site. Overall enrolment occurred between November, 2007, and July, 2010, but individual sites started and finished at different times according to local logistics. Patients were managed according to local protocols. All data collection occurred prospectively and the data dictionary has been published previously.17 Research nursing staff collected the demographic and risk data from each patient, supervised ECG testing, and drew blood samples for biomarker testing. If a patient was unsure of an answer (eg, family history) a response of no was recorded. Patients were tracked for adverse events at 30 days from initial attendance with hospital records and www.thelancet.com Vol 377 March 26, 2011

Low risk (n=352) Age (years) Men

49·8 (9·2) 220 (62·5%)

High risk (n=3230)

Total

62·8 (14·0)

61·5 (14·1)

2014 (62·4%)

2234 (62·4%)

Ethnic origin White

190(56·4%)

1281 (40·5%)

1471 (42·1%)

Chinese

66 (19·6%)

1108 (35·1%)

1174 (33·6%)

Korean

26 (7·7%)

194 (6·1%)

220 (6·3%)

Indonesian

10 (3·0%)

200 (6·3%)

210 (5·9%)

Indian

9 (2·7%)

122 (3·9%)

131 (3·7%)

Thai

0

70 (2·2%)

70 (2·0%)

Malay

2 (0·6%)

46 (1·5%)

48 (1·3%)

Maori

3 (0·9%)

30 (0·9%)

33 (0·9%)

Aboriginal

1 (0·3%)

8 (0·3%)

9 (0·3%)

30 (8·9%)

102 (3·2%)

132 (3·7%)

Other Unknown

15 (4·2%)

69 (2·1%)

84 (2·3%)

Hypertension

65 (19·9%)

1921 (60·4%)

1986 (56·5%)

Dyslipidaemia

76 (24·0%)

1505 (48·3%)

1581 (46·0%)

124 (39·9%)

1196 (37·6%)

1320 (38·1%)

Previous

106 (32·3%)

1061 (33·2%)

1167 (33·1%)

Current

69 (21·0%)

625 (19·5%)

694 (19·7%) 1138 (41·3%)

Family history of CAD Smoking

Past medical history Angina

18 (8·6%)

1120 (44·0%)

CAD

4 (1·9%)

735 (28·9%)

739 (26·8%)

Acute myocardial infarction

0

625 (24·5%)

625 (22·7%)

Revascularisation

0

541 (21·3%)

541 (19·7%)

Congestive heart failure

2 (1·0%)

281 (11·0%)

283 (10·3%)

Stroke

3 (1·4%)

278 (10·9%)

281 (10·2%)

CABG

0

200 (7·8%)

200 (7·3%)

5 (2·0%)

158 (6·2%)

163 (5·9%)

Arrhythmia Length of initial hospital attendance (h)

26·0 (9·9–37·0)

50·1 (12·6–123·3)

46·0 (12·0–120·8)

Data are mean (SD), number (%), or median (IQR). Data were missing for each category as follows: ethnic origin (84), hypertension (75), dyslipidaemia (148), family history of CAD (118), smoking (54), previous medical history (824), and time in hospital (196). ADP=accelerated diagnostic protocol. CAD=coronary artery disease. CABG=coronary artery bypass graft.

Table 1: Characteristics for low-risk (ADP negative) and high-risk (ADP positive) participants in the ASPECT study (n=3582)

telephone follow-up. Data coordination, monitoring and analysis, and source verification was done through an independent university clinical research organisation at a non-recruitment location in Australia (Centre for Clinical Research Excellence, Monash University, Melbourne). Approval from local ethics committees was obtained, and all patients provided written informed consent.

Procedures The primary endpoint was major adverse cardiac events within 30 days after initial presentation (including initial hospital attendance). The criteria for major adverse cardiac event included any of the following: death (not clearly non-cardiac), cardiac arrest, an emergency revascularisation procedure, cardiogenic shock, ventricular arrhythmia needing intervention, high-degree atrioventricular block needing intervention, 1079

Articles

Number of events*

Patients (of 421) who had event type (%)

Frequency of event type (of 3582 patients in study; %)

ECG* Positive

148 (4·1%)

879 (24·5%)

1027 (28·7%)

363

86·2%

10·1%

Negative

273 (7·6%)†

2282 (63·7%)

2555 (71·3%)

STEMI

53

12·5%

1·5%

Total

421 (11·8%)

3161 (88·2%)

3582 (100%)

Emergency revascularisation

32

7·6%

0·9%

Cardiovascular death

19

4·5%

0·5%

Positive

407 (11·4%)

2606 (72·8%)

3013 (84·1%)

Ventricular arrhythmia

15

3·5%

0·4%

Negative

14 (0·4%)†

555 (15·4%)

569 (15·9%)

0·2%

421 (11·8%)

3161 (88·2%)

3582 (100%)

413 (11·6%)

2701 (75·4%)

3114 (86·9%)

8 (0·2%)†

460 (12·8%)

468 (13·1%)

421 (11·8%)

3161 (88·2%)

3582 (100%)

349 (9·7%)

1391 (38·8%)

1740 (48·6%)

1770 (49·4%)

1842 (51·4%)

3161 (88·2%)

3582 (100%)

NSTEMI

Cardiac arrest

8

1·9%

Cardiogenic shock

7

1·7%

0·2%

High atrioventricular block

4

1·0%

0·1%

NSTEMI=non-ST-segment elevation myocardial infarction. STEMI=ST-segment myocardial infarction occurring after initial recruitment. *421 of 3582 (11·8%) patients had a total of 501 events during initial hospital attendance or 30-day follow-up.

Table 2: Frequency and type of major adverse cardiac event during initial hospital attendance or 30-day follow-up

MACE

No MACE

Total

TIMI‡

Total ECG and TIMI§ Positive Negative Total POC biomarkers¶ Positive Negative Total

72 (2·0%)† 421(11·8%)

ECG and POC biomarkers||

See Online for webappendix

1080

and prevalent (ie, being the cause for the patient’s initial presentation) and incident (ie, occurring during the 30-day follow-up) acute myocardial infarction. Outcomes and investigations were reported with minimum subjectivity with predefined standardised reporting guidelines (webappendix p 1).16–20 The presence of a major adverse cardiac event was adjudicated independently by local cardiologists with these reporting guidelines. Cardiologists were masked to results of the index test biomarkers under investigation and derived TIMI score, but had knowledge of the clinical record, ECG, and serial troponin results from usual care. In accordance with international guidelines, blood troponins at presentation, and then at least 6 h afterwards formed part of the reference standard to establish presence of acute myocardial infarction.7,16 These measurements were part of normal care and were analysed at the recruitment site central hospital laboratory. Webappendix p 2 provides a summary of the characteristics of the laboratory troponins used at each hospital site. Treating clinicians were masked to the results of the index tests, with only central laboratory troponin results used in patient management. Classification of acute myocardial infarction was based on global taskforce recommendations requiring evidence of myocardial necrosis together with evidence of myocardial ischaemia (ischaemic symptoms, ECG changes, or imaging evidence).7 Necrosis was diagnosed on the basis of a rising or falling pattern of the laboratory cardiac troponin concentrations, with at least one value above the 99th percentile, at a level of assay imprecision near to 10%. If the troponin concentration was greater than the reference range, but no rise or fall was recorded, other causes of a raised troponin concentration were considered by the adjudicating cardiologist. If no clear alternative cause of the troponin rise was apparent, and if the clinical presentation was suggestive of acute coronary

Positive

374 (10·4%)

1803 (50·3%)

2177 (60·7%)

Negative

47 (1·3%)†

1358 (37·9%)

1405 (39·2%)

421(11·8%)

3161 (88·2%)

3582 (100%)

418 (11·7%)

2812 (78·5%)

3230 (90·2%)

Total ADP** Positive Negative Total

3 (0·08%)† 421 (11·8%)

349 (9·7%)

352 (9·8%)

3161 (88·2%)

3582 (100%)

MACE=major adverse cardiac event. ECG=electrocardiograph. TIMI=Thrombolysis In Myocardial Infarction score for unstable angina or non-ST-elevation myocardial infarction. POC=point of care. ADP=accelerated diagnostic protocol. *ECG alone; any new ischaemia was positive. †Numbers of patients who were identified as low risk by the diagnostic parameter(s) but had a MACE (ie, false-negative cases). ‡TIMI score of ≥1 was positive and TIMI score of 0 was negative. §ECG and TIMI used. Result was positive if TIMI score was ≥1 or ECG was positive. ¶POC biomarkers: troponin I, creatine kinase MB and change, and myoglobin and change. Any positive parameter created a positive result. ||ECG and POC biomarkers used. Any positive parameter created a positive result. **ADP was negative if TIMI score was 0 and if ECG and POC biomarkers were all negative. If TIMI score was ≥1 or any other parameter was positive, then ADP was positive.

Table 3: Occurrence of MACE during initial hospital attendance or 30-day follow-up according to results of individual and combinations of the ADP test parameters

syndromes, an adjudicated diagnosis of acute myocardial infarction was made. The predefined ADP under investigation was a combination of TIMI risk score of 0, no new ischaemic changes on the initial ECG, and normal point-of-care biomarker panel (at 0–2 h after arrival). All parameters had to be negative for the ADP to be considered negative (and thus for the patient to be identified as low risk). The TIMI score (panel 1) for unstable angina or non-STelevation myocardial infarction had to be zero for the sum of its seven parameters.14 New ECG ischaemic changes, with evidence that these changes were not pre-existing on previous ECGs, had to be absent. They were defined as ST-segment depression of at least 0·05 mV in two or more contiguous leads (including reciprocal changes), T-wave inversion of at www.thelancet.com Vol 377 March 26, 2011

Articles

Biomarkers

ECG*

POC TIMI‡ biomarkers†

POC biomarkers and ECG§

TIMI and ECG¶

ADP||

Sensitivity

35·2% (30·7–39·8)

82·9% (79·0–86·2)

96·7% (94·5–98·0)

88·8% (85·5–91·5)

98·1% (96·3–99·0)

99·3% (97·9–99·8)

Negative predictive value

89·3% (88·0–90·4)

96·1% (95·0–96·9)

97·5% (95·8–98·6)

96·7% (95·5–97·5)

98·3% (96·5–99·2)

99·1% (97·3–99·8)

Specificity

72·2% (70·6–73·7)

56·0% (54·3–57·7)

17·6% (16·3–18·9)

43·0% (41·2–44·7)

14·6% (13·4–15·8)

11·0% (10·0–12·2)

Positive predictive value

14·4% (12·3–16·7)

20·1% (18·2–22·0)

13·5% (12·3–14·8)

17·2% (15·6–18·8)

13·3% (12·1–14·5)

12·9% (11·8–14·5)

Negative likelihood ratio

0·9 (0·8–1·0)

0·3 (0·3–0·4)

0·2 (0·1–0·3)

0·3 (0·2–0·3)

0·1 (0·1–0·3)

0·1 (0·0–0·2)

Positive likelihood ratio

1·3 (1·1–1·5)

1·9 (1·8–2·0)

1·2 (1·1–1·2)

1·6 (1·5–1·6)

1·1 (1·1–1·2)

1·1 (1·1–1·3)

72

4

47 3

14 TIMI=0

8

273 ECG

Figure 2: Occurrence of a major adverse cardiac event during initial hospital attendance or 30-day follow-up in patients with negative results for individual and combinations of diagnostic parameters Figures refer to numbers of patients. TIMI=Thrombolysis In Myocardial Infarction. ECG=electrocardiograph.

least 0·1 mV, or Q-waves greater than 30 ms in width and 0·1 mV or greater in depth in at least two contiguous leads.17,18,20 Patients with abnormal ECG findings (eg, pacing, left ventricular hypertrophy, and left bundle branch block) that were proven to be pre-existing on previous ECGs were defined as low risk. Index test point-of-care biomarkers were measured with whole blood drawn at presentation and 2 h afterwards. Blood was immediately tested for troponin I, creatine kinase MB, and myoglobin. Results were available (to research staff only) within 15 min with the TRIAGE platform or CardioProfilER assay panels (both Alere, San Diego, CA, USA). The following assay results were predefined to be positive on either blood draw: troponin I 0·05 µg/L or greater, creatine kinase MB 4·3 µg/L or greater, or an increase of 1·6 µg/L or more within 2 h; and myoglobin concentration of 108 µg/L or greater or an increase of 25% or more within 2 h. The point cutoffs were based on manufacturer recommendations, with an elevated troponin defined as any detectable concentration of troponin. The levels of change were based on a previous publication21 and peer-group consensus.

Statistical analysis Data were collected with the web-based Open-Clinica data capture system. Baseline characteristics of the study population were analysed with conventional group descriptive statistics. χ² analyses were used to generate two-by-two tables for the calculation of sensitivity, specificity, and positive and negative predictive values. All analyses were done with SPSS (version 18.0.0). www.thelancet.com Vol 377 March 26, 2011

POC=point of care. ECG=electrocardiograph. ADP=accelerated diagnostic protocol. MACE=major adverse cardiac event. TIMI=Thrombolysis In Myocardial Infarction score for unstable angina or non-ST-elevation myocardial infarction. *ECG alone; any new ischaemia was positive. †POC biomarkers: troponin I, creatine kinase MB and change, and myoglobin and change. Any positive parameter created a positive result. ‡TIMI score of ≥1 was positive and TIMI score of 0 was negative. §POC biomarkers and ECG used. Any positive parameter created a positive result. ¶TIMI and ECG used. Result was positive if TIMI score was ≥1 or ECG was positive. ||ADP was negative if TIMI score was 0 and if ECG and POC biomarkers were all negative. If TIMI score was ≥1 or any other parameter was positive, then ADP was positive.

Table 4: Accuracy (95% CI) of POC biomarkers, ECG, and ADP for prediction of MACE

The trial is registered with the AustraliaNew Zealand Clinical Trials Registry, number ACTRN12609000283279.

Role of the funding source The sponsors of this study had no role in the study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Results 3651 consenting eligible patients were enrolled, of whom 3582 completed 30-day follow-up (figure 1). Webappendix p 3 shows the countries and hospitals that recruited patients. Study participants were mostly older men, either white or Chinese, and commonly had cardiovascular risk factors and background cardiovascular past medical history (table 1). A major adverse cardiac event occurred within 30 days in 421 (11·8%) patients. Non-ST-segment acute myocardial infarction (NSTEMI) was the most frequently occurring major adverse cardiac event (table 2). The ADP identified 9·8% (352/3582) of patients as being at low risk of a major adverse cardiac event within 30 days (all ADP parameters were negative). Three (0·9%) of these patients had an event during initial hospital attendance and follow-up (figure 1). Webappendix p 4 outlines the clinical details of these false negatives. The combinations of parameters of the ADP were more effective at identifying patients who had a major adverse cardiac event than were the individual parameters themselves (table 3). The combination of the biomarkers and ECG without the TIMI score did not identify 1081

Articles

47 patients with a major adverse cardiac event at day 30. With use of the ADP including TIMI score, 44 additional patients were correctly identified, which reduced the number of false negatives to three (figure 2). Table 4 shows the statistical analysis of the ADP and its parameters for the prediction of a major adverse cardiac event by day 30. The ADP had a very high sensitivity and negative predictive value (table 4). Secondary analysis showed that patients identified as low risk by negative ADP were associated with a median initial hospital attendance of 26·0 h (IQR 9·9–37·0) and a mean of 43·2 h (95% CI 36·2–51·2), representing 1–2 hospital bed-days.

Discussion Findings from this large, multinational study have prospectively validated that a 2-h accelerated diagnostic protocol, with use of point-of-care biomarkers, ECG, and TIMI score, can safely identify patients at very low shortterm risk of a major adverse cardiac event (panel 2). These patients could potentially be discharged several hours earlier to outpatient follow-up and further investigations than with present practices. The near 10% possible reduction in patients needing prolonged assessment in this large patient group could reduce overcrowding in hospitals and emergency departments and provide earlier reassurance and greater convenience for patients. The potential reduction in initial length of stay accords with the findings of a six centre study in the UK.22 These findings together with those from countries included in our study represent 42% of the world’s population. Extrapolation is difficult, but on the basis of incidence rates of chest pain in the USA of 2·21%, there might be 64 million presentations of chest pain per year across these study nations. If the true incidence was half of this rate, then earlier discharge of 10% of patients could affect 3·2 million presentations. Patients in this study who were identified as low risk had an initial hospital attendance of about 1–2 days; these patients could potentially be discharged within 3–4 h of arrival if follow-up investigations could be arranged as an outpatient. Increasing demand for acute hospital beds is a key challenge for modern health services. The study shows that each of the components of the ADP is essential when used within such an early timeframe after presentation (figure 2, table 3). The use of the TIMI score within the ADP resulted in a lower and more acceptable false negative rate than when only biomarkers and ECG were used for the prediction of 30-day major adverse cardiac event (0·7% vs 11·2%). Troponin assays with lower and more reliable levels of detection have been developed since this study started, but the assay we used was effective in this ADP. The focus of this study was the safety of the ADP when used as a whole; any contemporary troponin could be used either via the central laboratory or point of care as part 1082

Panel 2: Research in context Systematic review We searched Medline from March, 1995, to December, 2010, for full reports of original research and review articles with the terms “acute coronary syndrome”, “chest pain”, “emergency department”, “risk stratification tools”, “point of care”, and “clinical decision rule”. We identified 114 articles. Abstracts were downloaded for all titles of potential relevance. Full papers were downloaded when the abstract was also deemed relevant. To be included in the final analysis, studies had to be prospective, have a large population, and have clearly described their methods and results. The methodology must have allowed the conclusions to be generalised to the emergency department population. Interpretation Together, the results of these studies show that the identification of patients at low risk for major adverse cardiac events is challenging. Increasing research is emerging into the use of accelerated diagnostic protocols (ADP). These protocols typically include the use of a risk stratification method, serial biomarkers, and electrocardiographs, and usually require an assessment period of 6–12 h. The results of our study indicate that a new ADP incorporating a risk stratification method (TIMI score), electrocardiograph, and point-of-care biomarker testing can identify patients at low risk of 30-day major cardiac event at 2 h.

of the ADP. Newer assays, which typically have lower detection limits and higher analytical precision, would probably improve the sensitivity of this ADP for the prediction of a major adverse cardiac event. These newer assays might be used with decision rules under development23 for use in a broad risk population. In this trial, combinations of biomarkers provided cumulative improvement in sensitivity, but a cardiac troponin as a sole biomarker was sufficient alone to produce a high sensitivity of 98·6% (415/421) once ECG and TIMI were added. Although not an a-priori hypothesis, this finding suggests that the ADP might be optimised to include only the cardiac troponin results in conjunction with the ECG and TIMI risk score in the future. Other biomarkers (eg, copeptin and heart fatty acid binding protein) might improve the diagnostic accuracy for acute myocardial infarction; however, their use as part of an ADP has not been reported.24,25 The ADP might be expanded to a broader subset by development of a more specific risk score. The TIMI score was developed from a relatively high-risk population with acute coronary syndromes, but it has been externally validated in more general emergency department populations.2,3,26 A modified TIMI risk score has been derived and validated in an emergency department population previously with laboratory-based troponins,27,28 with a sensitivity of 96·6% reported in the validation www.thelancet.com Vol 377 March 26, 2011

Articles

study. There is no universally accepted definition of a low-risk patient for acute coronary syndromes. This lack of consensus is a serious concern, because according to Bayesian decision making, interpretation of post-test probability after a particular test result is dependent on knowledge of the pre-test probability. The use of a structured and reproducible method is important.29–33 Subjective pre-test probability estimation has much lower inter-rater agreement between clinicians than do structured methods.34 Furthermore, patients presenting to an emergency department are often initially assessed by junior staff, and evidence shows that traditionally taught clinical variables and risk factors are poor predictors of acute coronary syndromes in an undifferentiated population in these clinics.35–37 Patients without chest pain but who presented with atypical symptoms (fatigue, nausea, vomiting, diaphoresis, faintness, and back pain) were not included in this trial, and we were unable to quantify the number of patients presenting with these symptoms. Thus the applicability of the ADP is limited to the selected cohort of patients with chest pain (or discomfort) suggestive of acute coronary syndromes for whom the attending physician planned to investigate for these syndromes. Another limitation of this study is that this was an observational, not an intervention study. Ideally, a management study of the diagnostic protocol would now occur; however, in practice, such studies are rare. The low specificity (11%) of our approach might be regarded as a limitation, but the ADP was used as an exclusion method to predict safety of early discharge of patients and not to establish inpatient management. These patients would otherwise have had extended observation or admission. The low specificity accords with other diagnostic instruments to exclude acute coronary syndromes.10 The goal of a more specific test is to rule-in a diagnosis if positive with sufficient certainty to initiate a change in management. In the setting that we studied, a positive protocol result merely classified patients as requiring management as usual. The optimum balance between specificity and sensitivity is difficult to define. A process yielding a higher specificity is likely to discharge a larger number of patients; however, we believe that the main focus should be on safety and therefore sensitivity. Future research should focus on methods to identify a greater proportion of patients who can be discharged earlier without significant adverse events. Contributors MT had overall responsibility for the trial. MT, LC, CMR, SA, DFF, SHL, WAP, and AMR contributed to the study design. MT, SA, and PMG (New Zealand); LC, WAP, and DFF (Australia); HFH and HFK (Hong Kong); RRK and MB (India); SS (Indonesia); DH, RD, and QH (China); KS-M (Korea); SHL (Singapore); PS and RS-L (Thailand); and T-FC, K-CT, F-YC, W-KC, and W-HC (Taiwan) collected data. MT, LC, CMR, SA, WAP, and MWA analysed data. MT, LC, CMR, MWA, WFP, and AMR wrote the report, which was reviewed by all authors. MT and LC did the literature search.

www.thelancet.com Vol 377 March 26, 2011

Conflicts of interest MT, MB, SHL, RRK, and LC received grants and supplies by Alere Medical. MT, AMR, and LC received honoraria for previous speaking and lecturing from Alere Medical. MT, MB, AMR, SHL, RRK, LC, and W-KC received support for travel to meetings from Alere Medical. MT received provision of administrative support funds from Alere Medical. HFH and HFK received grants from Science International Corporation. HFH received support for travel from Science International Corporation. MWA received unrelated grants from HRCNZ. LC received grants from the Queensland Emergency research Foundation (QEMRF). SA received grants from the National Heart Foundation of New Zealand, and support for travel to meetings from the Christchurch Cardio-Endocrine Research Group. CMR received grants from the National Health and Medical Research Council. WAP has received grants from the QEMRF. He is a board member of Sanofi-Aventis, is a consultant for Hospira, and has been paid to give lectures for Sanofi-Aventis and Roche, all unrelated to this project. WFP has received consultancy payments from Alere for unrelated projects. SS received grants, support for travel to meetings, and fees for participation in review activities from Medquest Jaya Global. DH, RD, QH, KS-M, DFF, RS-L, SS, and PS received support from Alere to travel to meetings. T-FC, K-CT, F-YC, and W-HC received grants for nurses and support for travel from Progressive Group (Taiwan). PMG has received unrelated grants from the Health Research Council New Zealand, National Heart Foundation New Zealand, and National Health and Medical Research Council; and unrelated honoraria from Roche, AstraZenica, and Abbott Laboratories. Acknowledgments We thank the patients who participated in the trial; Angela Brennan, Carl Costolloe, and Philippa Loane for independent third party oversight of the study and source data verification at the Centre for Clinical Research Excellence, Monash University, Melbourne; Queensland Emergency Medicine Research Foundation and National Health and Medical Research Council (Australia), Christchurch Cardio-Endocrine Research Group (New Zealand), Alere Medical (all countries), Medquest Jaya Global (Indonesia), Science International (Hong Kong), Bio Laboratories Pte (Singapore), National Heart Foundation of New Zealand, and Progressive Group (Taiwan) for helping to subsidise the costs of the research infrastructure at study sites; Allan S Jaffe, Jeffrey A Kline, Sarah Lord, Deborah Diercks, Steven Goodacre, Anthony F T Brown, Fred Apple, and Alan Maisel for reviewing the manuscript; Naresh Trehan for administrative support and patient recruitment in India; Rahul Mehrotra for patient recruitment and data collection and verification in India; Darren M Beam for assistance with the data dictionary; Christopher M A Frampton for initial statistical advice; and Joanne M Deely for medical writing and editing. References 1 Goodacre S, Cross E, Arnold J, Angelini K, Capewell S, Nicholl J. The health care burden of acute chest pain. Heart 2005; 91: 229–30. 2 Pollack CV Jr, Sites FD, Shofer FS, Sease KL, Hollander JE. Application of the TIMI risk score for unstable angina and non-ST elevation acute coronary syndrome to an unselected emergency department chest pain population. Acad Emerg Med 2006; 13: 13–18. 3 Chase M, Robey JL, Zogby KE, Sease KL, Shofer FS, Hollander JE. Prospective validation of the Thrombolysis in Myocardial Infarction risk score in the emergency department chest pain population. Ann Emerg Med 2006; 48: 252–59. 4 Hollander JE. The continuing search to identify the very-low-risk chest pain patient. Acad Emerg Med 1999; 6: 979–81. 5 Pollack CV Jr, Antman EM, Hollander JE. 2007 focused update to the ACC/AHA guidelines for the management of patients with ST-segment elevation myocardial infarction: implications for emergency department practice. Ann Emerg Med 2008 52: 344–55 e1. 6 Amsterdam EA, Kirk JD, Bluemke DA, et al. Testing of low-risk patients presenting to the emergency department with chest pain: a scientific statement from the American Heart Association. Circulation 2010; 122: 1756–76. 7 Thygesen K, Alpert JS, White HD, et al. Universal definition of myocardial infarction. Circulation 2007; 116: 2634–53.

1083

Articles

8 9 10

11

12

13 14

15 16

17 18

19

20

21

1084

Aroney CN, Dunlevie HL, Bett JHN. Use of an accelerated chest pain assessment protocol in patients at intermediate risk of adverse cardiac events. Med J Aust 2003; 178: 370–74. Bernstein SL, Aronsky D, Duseja R, et al. The effect of emergency department crowding on clinically oriented outcomes. Acad Emerg Med 2009; 16: 1–10. Steurer J, Held U, Schmid D, Ruckstuhl J, Bachmann LM. Clinical value of diagnostic instruments for ruling out acute coronary syndrome in patients with chest pain: a systematic review. Emerg Med J 2010; 27: 896–902. Newby LK, Storrow AB, Gibler WB, et al. Bedside multimarker testing for risk stratification in chest pain units: the chest pain evaluation by creatine kinase-MB, myoglobin, and troponin (CHECKMATE) study. Circulation 2001; 103: 1832–37. Apple FS, Jesse RL, Newby LK, Wu AH, Christenson RH. National Academy of Clinical Biochemistry and IFCC Committee for Standardization of Markers of Cardiac Damage Laboratory Medicine Practice Guidelines: analytical issues for biochemical markers of acute coronary syndromes. Circulation 2007; 115: e352–55. Conway Morris A, Caesar D, Gray S, Gray A. TIMI risk score accurately risk stratifies patients with undifferentiated chest pain presenting to an emergency department. Heart 2006; 92: 1333–34. Antman EM, Cohen M, Bernink PJ, et al. The TIMI risk score for unstable angina/non-ST elevation MI: a method for prognostication and therapeutic decision making. JAMA 2000; 284: 835–42. Miller CD, Lindsell CJ, Anantharaman V, et al. Performance of a population-based cardiac risk stratification tool in Asian patients with chest pain. Acad Emerg Med 2005; 12: 423–30. Luepker RV, Apple FS, Christenson RH, et al. Case definitions for acute coronary heart disease in epidemiology and clinical research studies: a statement from the AHA Council on Epidemiology and Prevention; AHA Statistics Committee; World Heart Federation Council on Epidemiology and Prevention; the European Society of Cardiology Working Group on Epidemiology and Prevention; Centers for Disease Control and Prevention; and the National Heart, Lung, and Blood Institute. Circulation 2003; 108: 2543–49. Cullen L, Than M, Brown AF, et al. Comprehensive standardized data definitions for acute coronary syndrome research in emergency departments in Australasia. Emerg Med Australas 2010; 22: 35–55. Hollander JE, Blomkalns AL, Brogan GX, et al. Standardized reporting guidelines for studies evaluating risk stratification of ED patients with potential acute coronary syndromes. Acad Emerg Med 2004; 11: 1331–40. Forest RS, Shofer FS, Sease KL, Hollander JE. Assessment of the standardized reporting guidelines ECG classification system: the presenting ECG predicts 30-day outcomes. Ann Emerg Med 2004; 44: 206–12. Cannon CP, Battler A, Brindis RG, et al. American College of Cardiology key data elements and definitions for measuring the clinical management and outcomes of patients with acute coronary syndromes. A report of the American College of Cardiology Task Force on Clinical Data Standards (Acute Coronary Syndromes Writing Committee). J Am Coll Cardiol 2001; 38: 2114–30. Fesmire FM, Percy RF, Bardoner JB, Wharton DR, Calhoun FB. Serial creatinine kinase (CK) MB testing during the emergency department evaluation of chest pain: utility of a 2-hour deltaCK-MB of +1.6ng/ml. Am Heart J 1998; 136: 237–44.

22

23 24 25 26

27

28

29 30 31 32 33 34 35 36

37

Goodacre SW, Bradburn M, Cross E, et al. The Randomised Assessment of Treatment using Panel Assay of Cardiac Markers (RATPAC) trial: a randomised controlled trial of point-of-care cardiac markers in the emergency department. Heart 2011; 97: 190–96. Hess EP, Wells GA, Jaffe A, Stiell IG. A study to derive a clinical decision rule for triage of emergency department patients with chest pain: design and methodology. BMC Emerg Med 2008; 8: 3. Reichlin T, Hochholzer W, Stelzig C, et al. Incremental value of copeptin for rapid rule out of acute myocardial infarction. J Am Coll Cardiol 2009; 54: 60–68. Valle HA, Riesgo LG, Bel MS, et al. Clinical assessment of heart-type fatty acid binding protein in early diagnosis of acute coronary syndrome. Eur J Emerg Med 2008; 15: 140–44. Campbell CF, Chang AM, Sease KL, et al. Combining Thrombolysis in Myocardial Infarction risk score and clear-cut alternative diagnosis for chest pain risk stratification. Am J Emerg Med 2009; 27: 37–42. Body R, Carley S, McDowell G, et al. Can a modified thrombolysis in myocardial infarction risk score outperform the original for risk stratifying emergency department patients with chest pain? Emerg Med J 2009; 26: 95–99. Hess EP, Perry JJ, Calder LA, et al. Prospective validation of a modified thrombolysis in myocardial infarction risk score in emergency department patients with chest pain and possible acute coronary syndrome. Acad Emerg Med 2010; 17: 368–75. Dolan JG, Bordley DR, Mushlin AI. An evaluation of clinicians’ subjective prior probability estimates. Med Decis Making 1986; 6: 216–23. Kassirer JP, Kopelman RI. Cognitive errors in diagnosis: instantiation, classification, and consequences. Am J Med 1989; 86: 433–41. Cahan A, Gilon D, Manor O, Paltiel O. Probabilistic reasoning and clinical decision-making: do doctors overestimate diagnostic probabilities? Q JM 2003; 96: 763–69. Attia JR, Nair BR, Sibbritt DW, et al. Generating pre-test probabilities: a neglected area in clinical decision making. Med J Aust 2004; 180: 449–54. Phelps MA, Levitt MA. Pretest probability estimates: a pitfall to the clinical utility of evidence-based medicine? Acad Emerg Med 2004; 11: 692–94. Iles S, Hodges AM, Darley JR, et al. Clinical experience and pre-test probability scores in the diagnosis of pulmonary embolism. Q JM 2003; 96: 211–15. Goodacre S, Locker T, Morris F, Campbell S. How useful are clinical features in the diagnosis of acute, undifferentiated chest pain? Acad Emerg Med 2002; 9: 203–08. Body R, Carley S, Wibberley C, McDowell G, Ferguson J, Mackway-Jones K. The value of symptoms and signs in the emergent diagnosis of acute coronary syndromes. Resuscitation 2010; 81: 281–86. Han JH, Lindsell CJ, Storrow AB, et al. The role of cardiac risk factor burden in diagnosing acute coronary syndromes in the emergency department setting. Ann Emerg Med 2007; 49: 145–52.

www.thelancet.com Vol 377 March 26, 2011

Articles

Separate and combined associations of body-mass index and abdominal adiposity with cardiovascular disease: collaborative analysis of 58 prospective studies The Emerging Risk Factors Collaboration*

Summary

Background Guidelines differ about the value of assessment of adiposity measures for cardiovascular disease risk prediction when information is available for other risk factors. We studied the separate and combined associations of body-mass index (BMI), waist circumference, and waist-to-hip ratio with risk of first-onset cardiovascular disease. Methods We used individual records from 58 cohorts to calculate hazard ratios (HRs) per 1 SD higher baseline values (4·56 kg/m² higher BMI, 12·6 cm higher waist circumference, and 0·083 higher waist-to-hip ratio) and measures of risk discrimination and reclassification. Serial adiposity assessments were used to calculate regression dilution ratios. Results Individual records were available for 221 934 people in 17 countries (14 297 incident cardiovascular disease outcomes; 1·87 million person-years at risk). Serial adiposity assessments were made in up to 63 821 people (mean interval 5·7 years [SD 3·9]). In people with BMI of 20 kg/m² or higher, HRs for cardiovascular disease were 1·23 (95% CI 1·17–1·29) with BMI, 1·27 (1·20–1·33) with waist circumference, and 1·25 (1·19–1·31) with waist-to-hip ratio, after adjustment for age, sex, and smoking status. After further adjustment for baseline systolic blood pressure, history of diabetes, and total and HDL cholesterol, corresponding HRs were 1·07 (1·03–1·11) with BMI, 1·10 (1·05–1·14) with waist circumference, and 1·12 (1·08–1·15) with waist-to-hip ratio. Addition of information on BMI, waist circumference, or waist-to-hip ratio to a cardiovascular disease risk prediction model containing conventional risk factors did not importantly improve risk discrimination (C-index changes of –0·0001, –0·0001, and 0·0008, respectively), nor classification of participants to categories of predicted 10-year risk (net reclassification improvement –0·19%, –0·05%, and –0·05%, respectively). Findings were similar when adiposity measures were considered in combination. Reproducibility was greater for BMI (regression dilution ratio 0·95, 95% CI 0·93–0·97) than for waist circumference (0·86, 0·83–0·89) or waist-to-hip ratio (0·63, 0·57–0·70).

Lancet 2011; 377: 1085–95 Published Online March 11, 2011 DOI:10.1016/S01406736(11)60105-0 See Comment page 1051 *Members listed at end of paper Correspondence to: Emerging Risk Factors Collaboration Coordinating Centre, Department of Public Health and Primary Care, University of Cambridge, Strangeways Research Laboratory, Cambridge CB1 8RN, UK [email protected]

Interpretation BMI, waist circumference, and waist-to-hip ratio, whether assessed singly or in combination, do not importantly improve cardiovascular disease risk prediction in people in developed countries when additional information is available for systolic blood pressure, history of diabetes, and lipids. Funding British Heart Foundation and UK Medical Research Council.

Introduction National and international guidelines have provided differing recommendations about the value of clinical measures of adiposity for prediction of cardiovascular disease risk in primary prevention.1 Recommendations range from omission of adiposity measures, to inclusion of such measures as additional screening tests, to formal inclusion of such measures as risk factors in prediction models. For example, whereas WHO2 and the US National Heart, Lung and Blood Institute3 recommend assessment of both body-mass index (BMI) and waist circumference in people with a BMI of 25·0–34·9 kg/m², several often used cardiovascular disease risk scores omit adiposity measures (eg, Framingham, SCORE, PROCAM, Reynolds), but others include BMI (eg, QRISK).4 This divergence in guideline recommendations might, in part, indicate uncertainties in relation to data from previous studies. For example, in a large multinational retrospective case-control study, waist-to-hip ratio was three times more strongly related to risk of acute www.thelancet.com Vol 377 March 26, 2011

myocardial infarction than was BMI.5 However, these suggestions have not been tested by powerful prospective studies with assessment of BMI, waist circumference, and waist-to-hip ratio in the same people.6–12 Prospective studies of adiposity have often lacked concomitant measurement of lipids and other conventional risk factors, which has impeded assessment of adiposity measures in the context of standard risk prediction scores.8,13 Furthermore, because studies have often reported on measures of association (such as relative risks) rather than on specific measures of predictive ability (eg, measures of risk discrimination and reclassification), they might not have been able to make an optimum assessment of predictive ability (as opposed to aetiological importance).14,15 Finally, reliable comparison of the long-term reproducibility of BMI, waist circumference, and waist-to-hip ratio has been lacking. The objective of this report was to analyse individual data from 221 934 participants in 58 prospective studies to produce reliable estimates of associations of BMI, 1085

Articles

Coronary heart disease (39 studies, 150 296 participants, 5460 cases) Adjusted for age, sex, and smoking status Adjusted for age, sex, smoking status, and baseline values of intermediate risk factors* 2·5

3·0

HR (95% CI)

2·0

1·5

1·0

Ischaemic stroke (21 studies, 89 413 participants, 2582 cases) 3·0 2·5

HR (95% CI)

2·0

1·5

1·0

20

25 30 BMI (kg/m2)

35

70

80 90 100 110 Waist circumference (cm)

120

0·7

0·8 0·9 1·0 Waist-to-hip ratio

1·1

Figure 1: HRs for coronary heart disease and ischaemic stroke across quantiles of baseline BMI, waist circumference, and waist-to-hip ratio Regression analyses were stratified, where appropriate, by sex. Adjusted study-specific log HRs were combined by multivariate random-effects meta-analysis. Y-axes are shown on a log scale. Reference groups are the second deciles in the plots for coronary heart disease and the first quintiles in the plots for ischaemic stroke. HR=hazard ratio. BMI=body-mass index. *Intermediate risk factors were systolic blood pressure, history of diabetes, and total and HDL cholesterol.

waist circumference, and waist-to-hip ratio with firstonset cardiovascular disease outcomes. We quantified the incremental gain in cardiovascular disease prediction with these adiposity measures, singly and in combination, under a wide range of circumstances. We also compared the reproducibility of adiposity measures by use of serial measurements taken over several years in up to 63 821 people.

Methods

Study design

See Online for webappendix

1086

Details of the Emerging Risk Factors Collaboration have been described previously.16–19 This analysis involved individual records from 58 prospective studies with the following four features (webappendix pp 12 and 27–30): (1) participants were not selected on the basis of having previous vascular disease; (2) concomitant information was provided at baseline for weight, height, and waist and hip circumference; (3) cause-specific mortality or

vascular morbidity, or both, were recorded by use of well defined criteria; and (4) at least 1 year of follow-up had been accrued. In registration of fatal outcomes, all contributing studies used codings from the International Classification of Diseases to at least three digits (or used study-specific classification systems), and ascertainment was based on death certificates. 43 of the 58 contributing studies also used medical records, autopsy findings, and other supplementary sources to help classify deaths. 50 studies used definitions of myocardial infarction based on WHO criteria. 43 studies reported diagnosis of strokes on the basis of brain imaging, and attributed stroke subtype. This study was approved by the Cambridgeshire ethics review committee.

Statistical analysis We assessed BMI, waist circumference, and waist-to-hip ratio in relation to three fatal or first-ever non-fatal www.thelancet.com Vol 377 March 26, 2011

Articles

outcomes: coronary heart disease; ischaemic stroke; and coronary heart disease and any cerebrovascular disease (ie, cardiovascular disease). Because waist-to-height ratio was highly correlated with waist circumference (r=0·95, 95% CI 0·94–0·96), waist-to-height ratio was not reported separately. Analyses were done in two stages by use of estimates of association calculated separately within each study before pooling across studies by random-effects meta-analysis.20 Hazard ratios (HRs) were calculated with Cox proportional hazard regression models, stratified by sex, with censoring of deaths from non-cardiovascular disease causes. The proportional hazards assumptions were met. To avoid overfitting of the statistical models, studies with fewer than ten incident cases of an outcome were excluded from analysis of that particular outcome (although such studies were able to contribute full data for analyses of cross-sectional correlates of adiposity measures). Participants contributed only first non-fatal outcome or death recorded at age 40 years or older (ie, deaths preceded by non-fatal coronary heart disease or stroke were not included). For the four contributing individually matched nested case-control studies within prospective cohorts, odds ratios were calculated with conditional logistic regression models. To characterise shapes of associations, study-specific HRs calculated within overall quantiles of baseline adiposity values were pooled on a log scale by multivariate random-effects meta-analysis and plotted against mean values of the relevant adiposity measure within each quantile. We estimated 95% CIs from the variances that correspond to the amount of information underlying each group (including the reference group).21 In the figures presented, sizes of data markers are proportional to the inverse of the variance of the log HRs. Since associations were nearly log-linear (except at low values of BMI), we calculated HRs associated with 1 SD higher baseline value: 4·56 kg/m² higher BMI, 12·6 cm higher waist circumference, and 0·083 higher waist-to-hip ratio. We also calculated HRs with clinically defined categories of BMI and waist circumference in combination. We investigated effect-modification with formal tests of interaction, and calculated p values for interaction with continuous variables, when appropriate. Diversity between studies was investigated by grouping studies with recorded characteristics and meta-regression. HRs were initially adjusted for age, sex, and smoking status only. To explore potential biological pathways underlying associations, HRs were further adjusted for systolic blood pressure, history of diabetes, and total and HDL cholesterol. Extent of heterogeneity was indicated by the I² statistic.22 Regression dilution ratios were obtained by regression of serial measurements of the relevant characteristic on baseline values. For cardiovascular disease risk prediction models, a single Cox model, stratified by study and sex, was derived with baseline values of risk factors. Resulting predictions were assessed by use of measures of discrimination for censored time-to-event data (Harrell’s www.thelancet.com Vol 377 March 26, 2011

C-index),23 and net reclassification improvement and integrated discrimination improvement was calculated by use of four standard 10-year risk categories (<5%, 5% to <10%, 10% to <20%, and ≥20%).24 The C-index and its changes were calculated in two stages, first with estimation within each study and then pooled with metaanalysis techniques. Analyses were done with Stata (version 11.0), and two-sided p values and 95% CIs are presented.

Role of the funding source The sponsor of the study did not participate in the study design, data collection, data analysis, data interpretation, or writing of the report. DW and JD had full access to all data in the study and had final responsibility to submit the report for publication.

Results Individual records were available for 221 934 participants who had no known history of cardiovascular disease (ie, myocardial infarction, angina, or stroke, as ascertained in each study) at the baseline examination. 155 938 (70%) of these participants also had data on smoking status, systolic blood pressure, history of diabetes, and total and HDL cholesterol. Mean age of participants at baseline was 58 years (SD 9), 124 189 (56%) were women, and 129 326 (58%) were in Europe, 73 707 (33%) were in North America, 9204 (4%) were in Australia, and 9697 (4%) were in Japan (webappendix pp 3–7). Adiposity measures had broadly similar distributions across studies and were approximately linearly Adjusted for age, sex, and smoking status

Adjusted for age, sex, smoking status, and intermediate risk factors*

HR (95% CI)

HR (95% CI)

I² (95% CI)

I² (95% CI)

Coronary heart disease (39 studies, 143 710 participants, 5259 cases) BMI

1·29 (1·22–1·37)

65% (52–75)

1·11 (1·05–1·17)

45% (20–62)

Waist circumference

1·32 (1·24–1·40)

64% (50–75)

1·12 (1·06–1·19)

49% (25–65)

Waist-to-hip ratio

1·30 (1·22–1·38)

65% (51–75)

1·14 (1·09–1·18)

14% (0–42)

Ischaemic stroke (21 studies, 85 169 participants, 2431 cases) BMI

1·20 (1·12–1·28)

39% (0–64)

1·06 (0·99–1·13)

26% (0–57)

Waist circumference

1·25 (1·18–1·33)

21% (0–54)

1·11 (1·05–1·17)

9% (0–43)

Waist-to-hip ratio

1·25 (1·18–1·32)

21% (0–53)

1·14 (1·09–1·20)

0% (0–47)

Cardiovascular disease (39 studies, 144 795 participants, 8347 cases) BMI

1·23 (1·17–1·29)

72% (61–79)

1·07 (1·03–1·11)

47% (23–64)

Waist circumference

1·27 (1·20–1·33)

69% (57–78)

1·10 (1·05–1·14)

49% (26–65)

Waist-to-hip ratio

1·25 (1·19–1·31)

67% (54–76)

1·12 (1·08–1·15)

8% (0–38)

HRs are presented per 4·56 kg/m² higher BMI, 12·6 cm higher waist circumference, and 0·083 higher waist-to-hip ratio (ie, 1 SD higher baseline values). HRs were adjusted as shown, and stratified, where appropriate, by sex. Analyses were restricted to participants with BMI of 20 kg/m² or higher and complete information on age, sex, smoking status, and intermediate risk factors. HR=hazard ratio. BMI=body-mass index. *Intermediate risk factors were systolic blood pressure, history of diabetes, and total and HDL cholesterol.

Table 1: Associations of baseline values of BMI, waist circumference, and waist-to-hip ratio with risk of coronary heart disease, ischaemic stroke, and cardiovascular disease, adjusted for baseline values of confounders and intermediate risk factors

1087

Articles

A

B

Coronary heart disease Number of cases

HR (95% CI)

40–59

2644

1·41 (1·30–1·53)

60–69

1999

1·23 (1·15–1·31)

≥70

3049

1·12 (1·05–1·19)

Female

3070

1·24 (1·14–1·35)

Male

4680

1·26 (1·18–1·34)

40–59

2644

1·50 (1·37–1·63)

60–69

1999

1·28 (1·20–1·37)

≥70

3049

1·13 (1·06–1·21)

Female

3070

1·31 (1·21–1·43)

Male

4680

1·24 (1·17–1·32)

Bottom third

2127

1·24 (1·12–1·38)

Middle third

2602

1·26 (1·17–1·37)

Top third

3021

1·27 (1·18–1·37)

40–59

2644

1·43 (1·34–1·53)

60–69

1999

1·27 (1·19–1·36)

≥70

3049

1·14 (1·06–1·21)

Female

3070

1·30 (1·20–1·41)

Male

4680

1·24 (1·17–1·31)

Bottom third

2127

1·24 (1·15–1·33)

Middle third

2602

1·24 (1·18–1·31)

Top third

3021

1·23 (1·17-1·30)

BMI

Interaction p value

Ischaemic stroke

Number of cases

HR (95% CI)

Interaction p value

Age at survey (years) 0·006

669

1·34 (1·21–1·48)

899

1·22 (1·13–1·31)

1082

1·08 (0·99–1·18)

1649

1·20 (1·05–1·37)

1012

1·33 (1·21–1·46)

669

1·45 (1·30–1·60)

899

1·29 (1·20–1·40)

1082

1·10 (1·03–1·18)

1649

1·27 (1·12–1·43)

1012

1·32 (1·22–1·42)

0·051

Sex 0·643

0·030

Waist circumference Age at survey (years) <0·0001

0·001

Sex 0·056

0·429

BMI (kg/m2) 0·965

734

1·29 (1·12–1·47)

868

1·23 (1·09–1·40)

1059

1·21 (1·10–1·32)

669

1·39 (1·28–1·50)

899

1·27 (1·18–1·37)

1082

1·13 (1·05–1·20)

1649

1·27 (1·12–1·45)

1012

1·29 (1·20–1·39)

0·136

Waist-to-hip ratio Age at survey (years) <0·0001

0·0001

Sex 0·092

0·754

BMI (kg/m2)

1·0

1·2

1·4

1·6

0·995

734

1·21 (1·12–1·31)

868

1·18 (1·09–1·29)

1059

1·21 (1·13–1·30) 1·0

1·2

1·4

0·967

1·6

Figure 2: HRs for coronary heart disease (A) and ischaemic stroke (B) per 1 SD higher baseline values of BMI, waist circumference, and waist-to-hip ratio, according to age, sex, and BMI at baseline Analyses for coronary heart disease were based on up to 203 338 participants from 51 studies, and analyses for ischaemic stroke were based on up to 122 914 participants from 25 studies. HRs are presented per 4·56 kg/m² higher BMI, 12·6 cm higher waist circumference, and 0·083 higher waist-to-hip ratio (ie, 1 SD higher baseline values). Study-specific HRs were adjusted for age at baseline and smoking status, and stratified, where appropriate, by sex. Analyses were restricted to participants with BMI of 20 kg/m² or higher. X-axes are shown on a log scale. p values for interaction were calculated by use of continuous values of variables, when appropriate. HRs for coronary heart disease, initially adjusted for age, sex, and smoking status, and then additionally adjusted for BMI, were 1·31 (1·24–1·37) and 1·23 (1·15–1·32), respectively, with waist circumference, and 1·29 (1·23–1·35) and 1·21 (1·16–1·26), respectively, with waist-to-hip ratio. HRs for ischaemic stroke, initially adjusted for age, sex, and smoking status, and then additionally adjusted for BMI, were 1·26 (1·19–1·33) and 1·26 (1·16–1·36), respectively, with waist circumference, and 1·25 (1·19–1·32) and 1·18 (1·13–1·24), respectively, with waist-to-hip ratio. BMI=body-mass index. HR=hazard ratio.

associated with one another across the range of values (webappendix p 13). Correlation coefficients adjusted for age and sex were 0·85 (95% CI 0·84–0·86) between BMI and waist circumference, 0·43 (0·40–0·45) 1088

between BMI and waist-to-hip ratio, and 0·70 (0·68–0·72) between waist circumference and waist-to-hip ratio. Mean adiposity values generally increased with the age of participants until about 55–75 years, then flattened or www.thelancet.com Vol 377 March 26, 2011

Articles

A

Bottom third of BMI (<24·5)

Middle third of BMI (24·5 to <28·0)

Top third of BMI (≥28·0)

2·5

HR (95% CI)

2·0

1·5

1·0 80 48 400 1138

Number of participants Number of events

90 100 Waist circumference (cm) 17 622 797

110 1682 112

B Bottom third of BMI (<24·5)

80 18 414 410

90 100 Waist circumference (cm) 32 770 1259

110

16 563 1002

80 3108 67

Middle third of BMI (24·5 to <28·0)

90 100 Waist circumference (cm) 15 396 498

110

49 433 2467

Top third of BMI (≥28·0)

2·5

HR (95% CI)

2·0

1·5

1·0 0·8 Number of participants Number of events

37 007 701

0·85 0·9 0·95 Waist-to-hip ratio 20 880 791

1·0 9817 555

0·8 21 504 446

0·85 0·9 0·95 Waist-to-hip ratio 22 823 840

23 420 1385

1·0

0·8 15 987 408

0·85 0·9 0·95 Waist-to-hip ratio 18 981 647

1·0

32 969 1977

Figure 3: HRs for coronary heart disease across thirds of waist circumference (A) and waist-to-hip ratio (B) by baseline values of BMI Analyses were based on 203 338 participants (7750 cases) from 51 studies. Analyses were restricted to participants with BMI of 20 kg/m² or higher. Regression analyses were adjusted for age at baseline and smoking status, and stratified, where appropriate, by sex. Adjusted study-specific log HRs were combined by multivariate random-effects meta-analysis. Y-axes are shown on a log scale. Reference groups are the lowest third of waist circumference or waist-to-hip ratio in the bottom third of BMI. Results were similar for the full BMI range. BMI=body-mass index. HR=hazard ratio.

declined in participants older than 75 years (webappendix p 13). Adiposity measures were higher in people with a history of diabetes, were positively associated with non-HDL cholesterol, apolipoprotein B, triglyceride, fibrinogen, C-reactive protein, and systolic blood pressure, and were inversely associated with HDL cholesterol and apolipoprotein AI (webappendix pp 8–9 and 14–15). Systolic blood pressure, non-HDL cholesterol, HDL cholesterol, triglyceride, and C-reactive protein had similar associations with waist circumference or waistto-hip ratio across BMI values (webappendix p 16). People with a history of diabetes had higher mean waist www.thelancet.com Vol 377 March 26, 2011

circumference and waist-to-hip ratio than did people without diabetes across the range of BMI values (webappendix p 17). Serial adiposity assessments done in up to 63 821 people (mean interval 5·7 years [SD 3·9]) were used to calculate regression dilution ratios adjusted for age and sex: 0·95 (95% CI 0·93–0·97) for BMI, 0·86 (0·83–0·89) for waist circumference, 0·88 (0·85–0·91) for hip circumference, 0·63 (0·57–0·70) for waist-to-hip ratio, 0·63 (0·59–0·67) for non-HDL cholesterol, and 0·57 (0·52–0·61) for systolic blood pressure. On the basis of these regression dilution ratios, reproducibility of BMI was higher than that for waist circumference or 1089

Articles

C-index change (95% CI) Model including age and sex*

Reference

Plus BMI only

0·0051 (0·0031 to 0·0072)

Plus waist circumference only

0·0077 (0·0053 to 0·0100)

Plus waist-to-hip ratio only

0·0102 (0·0080 to 0·0125)

Plus BMI and waist circumference only

0·0077 (0·0054 to 0·0101)†

Plus BMI and waist-to-hip ratio only

0·0108 (0·0083 to 0·0134)‡

Plus non-lipid variables only§

0·0497 (0·0455 to 0·0538)

Plus Framingham variables only¶

0·0584 (0·0539 to 0·0628) 0

0·01

0·02

0·03

0·04

0·05

0·06

Figure 4: Changes in C-index for cardiovascular disease risk prediction from addition of adiposity measures or conventional risk factors to a model containing age and sex only Analyses were based on 144 795 participants (8347 cardiovascular events) in 39 studies. Analyses were restricted to participants with BMI of 20 kg/m² or higher. BMI=body-mass index. *Reference C-index was 0·6741 (95% CI 0·6685 to 0·6798) for the model including age and stratified by sex. †p=0·0001 for change in C-index after addition of waist circumference into the model with age, sex, and BMI. ‡p<0·0001 for change in C-index after addition of waist-to-hip ratio into the model with age, sex, and BMI. §Smoking status, systolic blood pressure, and history of diabetes. ¶Smoking status, systolic blood pressure, history of diabetes, and total and HDL cholesterol. C-index change (95% CI) Non-lipid-based model*

Reference

Plus BMI only

0·0006 (–0·0002 to 0·0013)

Plus waist circumference only

0·0010 (0·0001 to 0·0019)

Plus waist-to-hip ratio only

0·0023 (0·0013 to 0·0033)

Plus BMI and waist circumference only

0·0010 (0·0000 to 0·0020)†

Plus BMI and waist-to-hip ratio only

0·0022 (0·0011 to 0·0033)‡

Plus total cholesterol only

0·0030 (0·0018 to 0·0041)

Plus HDL cholesterol only

0·0051 (0·0037 to 0·0066)

Plus both total and HDL cholesterol only

0·0087 (0·0068 to 0·0106) 0

0·002

0·004

0·006

0·008

0·010

0·012

Figure 5: Changes in C-index for cardiovascular disease risk prediction from addition of adiposity measures or lipid markers to a non-lipid-based model Analyses were based on 144 795 participants (8347 cardiovascular events) in 39 studies. Analyses were restricted to participants with BMI of 20 kg/m² or higher. BMI=body-mass index. *Reference C-index was 0·7238 (95% CI 0·7186 to 0·7291) for the model including age, smoking status, systolic blood pressure, and history of diabetes, and stratified by sex. †p=0·175 for change in C-index after addition of waist circumference into the reference model plus BMI. ‡p<0·0001 for change in C-index after addition of waist-to-hip ratio into the reference model plus BMI.

waist-to-hip ratio. Between-study heterogeneity for the reproducibility in BMI and waist circumference was modest, whereas the corresponding heterogeneity for waist-to-hip ratio was substantial (I²=99%; webappendix p 18). During 1·87 million person-years at risk (median 5·7 years to first outcome, IQR 3·0–9·0), we recorded 14 297 first-ever outcomes: 8290 coronary heart disease outcomes (4982 non-fatal myocardial infarctions and 3308 coronary heart disease deaths), 2906 incident ischaemic stroke outcomes (2763 non-fatal and 143 fatal outcomes), 596 haemorrhagic stroke outcomes, 2070 unclassified stroke outcomes, and 435 other cerebrovascular outcomes (webappendix pp 3–4). In analyses adjusted for age, sex, and smoking status only, we recorded nearly log-linear associations of BMI, waist circumference, and waist-to-hip ratio with risk of coronary heart disease and ischaemic stroke, except at low BMI values (figure 1). Associations with coronary heart disease risk were similar when clinically defined categories of BMI and waist circumference were combined (webappendix p 19). 1090

To account for non-linear associations at low BMI values, further analyses excluded 9355 participants (4%) with BMI of less than 20 kg/m². After adjustment for age, sex, and smoking status, HRs for coronary heart disease and ischaemic stroke were broadly similar for BMI, waist circumference, and waist-to-hip ratio (table 1). After further adjustment for systolic blood pressure, history of diabetes, and total and HDL cholesterol, these HRs reduced substantially. HRs reduced even more after additional adjustment for C-reactive protein or fibrinogen (webappendix p 10). The excess risk of cardiovascular disease associated with BMI, waist circumference, and waist-to-hip ratio was three-to-four times stronger at age of 40–59 years than at older than 70 years, but HRs were similar in men and women (figure 2, webappendix p 20). HRs for waist circumference and waist-to-hip ratio were generally similar at different BMI values (figures 2 and 3) and triglyceride values (webappendix p 21), and HRs were slightly reduced after adjustment for BMI (figure 2). HRs were possibly higher at below-average systolic blood pressure, but otherwise did not vary substantially by clinical and demographic characteristics recorded www.thelancet.com Vol 377 March 26, 2011

Articles

BMI

Waist circumference

Waist-to-hip ratio

BMI and waist circumference

BMI and waist-to-hip ratio

–0·0001 (–0·0005 to 0·0002)

–0·0001 (–0·0006 to 0·0005)

0·0008 (0·0001 to 0·0014)

0·0000 (–0·0005 to 0·0006)

0·0006 (0·0000 to 0·0013)

Discrimination (39 studies, 144 795 participants, 8347 cases) C-index change (95% CI)* p value† p value‡

0·430 Reference

0·816

0·027

0·933

0·068

0·627

0·006

0·454

0·009

Reclassification (20 studies, 43 944 non-cases, 4777 cases) Participants who developed cardiovascular disease at 10 years Appropriately reclassified

68 (1·42%)

111 (2·32%)

132 (2·76%)

106 (2·22%)

141 (2·95%)

Inappropriately reclassified

73 (1·53%)

110 (2·30%)

136 (2·85%)

116 (2·43%)

142 (2·97%)

4636 (97·05%)

4556 (95·37%)

4509 (94·39%)

4555 (95·35%)

4494 (94·08%)

Appropriately reclassified

507 (1·15%)

806 (1·83%)

1091 (2·48%)

856 (1·95%)

1111 (2·53%)

Inappropriately reclassified

545 (1·24%)

839 (1·91%)

1078 (2·45%)

847 (1·93%)

1116 (2·54%)

42 892 (97·61%)

42 299 (96·26%)

41 775 (95·06%)

42 241 (96·12%)

41 717 (94·93%)

–0·19% (–0·70 to 0·32)

–0·05% (–0·69 to 0·58)

–0·05% (–0·76 to 0·65)

–0·19% (–0·83 to 0·45)

–0·03% (–0·75 to 0·69)

No change Participants who were event free at 10 years

No change Net reclassification improvement (95% CI) p value Integrated discrimination improvement (95% CI) p value

0·461 0·0001 (–0·0002 to 0·0003) 0·654

0·867 0·0004 (0·0000 to 0·0007) 0·043

0·880 0·0010 (0·0004 to 0·0015) 0·0003

0·562 0·0005 (0·0001 to 0·0008) 0·016

0·930 0·0009 (0·0004 to 0·0015) 0·001

Framingham risk score covariates include age, smoking status, systolic blood pressure, history of diabetes, and total and HDL cholesterol, and model was stratified by sex. Analyses are restricted to cohort studies and participants with BMI of 20 kg/m2 or higher. Cohort studies recording both fatal and non-fatal cardiovascular outcomes and more than 10 years of follow-up contributed to reclassification analyses. BMI=body-mass index. *Reference C-index was 0·7325 (95% CI 0·7274 to 0·7376) for the model including Framingham risk score covariates. †p value is for changes in C-index as compared with a model including Framingham risk score covariates. ‡p value is for changes in C-index as compared with addition of BMI alone.

Table 2: Reclassification of 10-year predicted risk and changes in risk discrimination for cardiovascular disease after addition of adiposity measures to a model including Framingham risk score covariates

(webappendix p 22). Heterogeneity in HRs for BMI, waist circumference, and waist-to-hip ratio (table 1) was only partly explained by the characteristics recorded (webappendix p 22). HRs did not appreciably alter after omission of the first 5 or 10 years of follow-up, but HRs decreased somewhat after correction for longterm average (usual) values of systolic blood pressure, history of diabetes, and total and HDL cholesterol (data not shown). HRs for BMI, waist circumference, and waist-to-hip ratio recorded in the full study population were qualitatively similar to those in analyses that excluded subgroups of participants or studies: current smokers (data not shown); participants who were not of European descent (data not shown); the 29 905 participants who had only self-reported adiposity measures (data not shown); the few studies with the most discrepant findings (webappendix p 23); or the 21 139 participants known to be receiving lipid-lowering, blood-pressure-lowering, or other cardiovascular drugs at baseline (data not shown). HRs were also broadly similar in fixed-effect models to those in random-effects models (webappendix p 23), and after additional adjustment for cigarette pack-years (in addition to smoking status), alcohol consumption, or measures of socioeconomic status (data not shown). We recorded no evidence of bias due to small studies (webappendix p 24). www.thelancet.com Vol 377 March 26, 2011

When information on BMI, waist circumference, or waist-to-hip ratio at baseline was added to a risk prediction model with age and sex only, the incremental gain in predictive ability was modest (figure 4). When information on adiposity measures was added to a risk prediction model without lipid information but containing information on other Framingham risk factors (age, sex, smoking status, systolic blood pressure, history of diabetes, and HDL and total cholesterol), the incremental gain in predictive value provided by a combination of BMI and waist-to-hip ratio was about a quarter of the predictive gain provided by total and HDL cholesterol (figure 5). When information on lipids and other conventional risk factors was available, additional information on BMI, waist circumference, or waist-to-hip ratio, either singly or in combination, did not appreciably change cardiovascular disease risk discrimination, according to C-index changes, nor improve reclassification of participants into predicted 10-year risk categories (table 2). To assess any incremental gain in predictive ability provided by adiposity measures, irrespective of the sequence of their addition to a risk model, we assessed the effect of omission of one risk factor at a time from a full model. Omission of BMI and waist circumference had nearly zero effect, and omission of waist-to-hip ratio 1091

Articles

C-index change (95% CI)

Variable omitted in turn

BMI Model including Framingham risk score covariates* plus BMI†

Reference

Smoking status

–0·0110 (–0·0130 to –0·0090)

Systolic blood pressure

–0·0147 (–0·0169 to –0·0125)

History of diabetes

–0·0078 (–0·0094 to –0·0062)

Total cholesterol

–0·0035 (–0·0047 to –0·0023)

HDL cholesterol

–0·0051 (–0·0065 to –0·0037) 0·0001 (–0·0002 to 0·0005)

BMI

Variable omitted in turn

Waist circumference Model including Framingham risk score covariates* plus waist circumference‡

Reference

Smoking status

–0·0111 (–0·0130 to –0·0091)

Systolic blood pressure

–0·0145 (–0·0167 to –0·0123)

History of diabetes

–0·0075 (–0·0091 to –0·0060)

Total cholesterol

–0·0036 (–0·0048 to –0·0024)

HDL cholesterol

–0·0048 (–0·0061 to –0·0034) 0·0001 (–0·0005 to 0·0006)

Waist circumference

Variable omitted in turn

Waist-to-hip ratio Model including Framingham risk score covariates* plus waist-to-hip ratio§

Reference

Smoking status

–0·0107 (–0·0126 to –0·0088)

Systolic blood pressure

–0·0143 (–0·0166 to –0·0121)

History of diabetes

–0·0073 (–0·0089 to –0·0058)

Total cholesterol

–0·0033 (–0·0045 to –0·0022)

HDL cholesterol

–0·0046 (–0·0059 to –0·0033)

Waist-to-hip ratio

–0·0008 (–0·0014 to –0·0001) –0·020

–0·0150

–0·010

–0·005

0

0·005

Figure 6: Changes in C-index for cardiovascular disease risk prediction from omission of individual risk factors from a full model containing Framingham risk score covariates plus BMI, waist circumference, or waist-to-hip ratio Analyses were based on 144 795 participants (8347 cardiovascular events) in 39 studies. Analyses were restricted to participants with BMI of 20 kg/m² or higher. BMI=body-mass index. *Framingham risk score covariates include age, smoking status, systolic blood pressure, history of diabetes, and total and HDL cholesterol, and model was stratified by sex. †Reference C-index of 0·7324 (95% CI 0·7272 to 0·7375). ‡Reference C-index of 0·7324 (95% CI 0·7273 to 0·7376). §Reference C-index of 0·7333 (95% CI 0·7281 to 0·7384).

had a small effect (figure 6). This result applied in analyses that either included or omitted people with history of diabetes at baseline (webappendix p 25), and for a wide range of other circumstances (webappendix p 26). Qualitatively similar results to those noted above were recorded in analyses assessing integrated discrimination improvement (table 2, webappendix p 11). We could not assess risk prediction at different ages because age distributions differed substantially between studies.

Discussion The results from our analysis of individual data from 221 934 people without initial cardiovascular disease have shown that BMI, waist circumference, or waist-tohip ratio, assessed singly or in combination, do not importantly improve prediction of cardiovascular disease risk when additional information is available on blood pressure, history of diabetes, and cholesterol measures (panel). This finding applies to a wide range of circumstances and clinically relevant subgroups. The main finding of this study does not, of course, diminish the importance of adiposity as a major 1092

modifiable determinant of cardiovascular disease. Rather, because excess adiposity is a major determinant of the intermediate risk factors noted above,3 our findings underscore the importance of controlling adiposity to help prevent cardiovascular disease. These data also suggest the relevance of controlling intermediate risk factors to combat the detrimental vascular effects of overweight and obesity.25 However, the findings of our study indicate that for populationwide assessment of cardiovascular disease risk, when additional information is available on factors used in standard risk scores, simple adiposity measures provide little or no additional information on cardiovascular risk. Even so, other reasons could support inclusion of adiposity measures in risk assessment, such as promotion of behaviour change26 or improvement of risk communication. Findings of previous smaller studies (and WHO guidelines) have suggested that for situations in which no information is available on lipids for cardiovascular disease risk prediction (such as in resource-limited settings), assessment of simple adiposity measures can be used instead, with only a modest loss of predictive ability.27,28 However, data from www.thelancet.com Vol 377 March 26, 2011

Articles

our study indicate that a combination of BMI and waistto-hip ratio provides only about a quarter of the extra predictive information provided by total and HDL cholesterol. The findings of our study emphasise, therefore, the desirability of supporting the development of lipid assessment in resource-poor settings. Our findings reliably refute previous recommendations to adopt baseline waist-to-hip ratio instead of BMI as the principal clinical measure of adiposity.5 Contrary to a report from INTERHEART (a large retrospective casecontrol study of acute myocardial infarction) that waistto-hip ratio is three times more strongly related to myocardial infarction than is BMI,5 we have shown that BMI, waist circumference, and waist-to-hip ratio each have a similar strength of association with cardiovascular disease risk. Indeed, the odds ratio for myocardial infarction was only 1·12 per 5 kg/m² higher baseline BMI in INTERHEART,5 whereas the HR for coronary heart disease per 5 kg/m² higher baseline BMI was 1·32 in our study. This discrepancy might be due to the greater susceptibility of retrospective studies of acute myocardial infarction to some biases (eg, selection biases, reverse causality) than are long-term prospective studies of people without a history of cardiovascular disease, as in our analysis. Furthermore, results from our study indicate that the long-term reproducibility of BMI is superior to that of waist-to-hip ratio (or waist circumference). Findings from our study could provide insight into cardiovascular disease mechanisms. Measures of abdominal adiposity, such as waist circumference and waist-to-hip ratio (especially in combination with high triglyceride concentration29), might be better indicators of visceral fat accumulation and an adverse metabolic profile than is raised BMI alone, since BMI is a measure of general adiposity.30,31 Yet, our results indicate that BMI, waist circumference, and waist-to-hip ratio each have a similar strength of association with cardiovascular disease risk, including similar associations with coronary heart disease and ischaemic stroke. The similarity of effect of adiposity on coronary heart disease and ischaemic stroke contrasts with results previously reported for proatherogenic lipids (which are more strongly associated with coronary heart disease than ischaemic stroke32) and systolic blood pressure (which is more strongly related to ischaemic stroke than coronary heart disease33). Our data, therefore, underscore the potential importance of reduction of adiposity for both coronary heart disease and ischaemic stroke. We did not record any modification of the effect of abdominal adiposity on cardiovascular disease risk by BMI. By contrast, age was a strong effect-modifier, with about three-to-four times higher excess risk in early middle age than in individuals older than 70 years.34 Our study importantly extends previous analyses of prospective studies. By contrast with analyses from the www.thelancet.com Vol 377 March 26, 2011

Panel: Research in context Systematic review Findings of a previous systematic review of 27 guideline statements showed substantial variation in recommendations about the value of inclusion of clinical measures of adiposity in risk scores for the primary prevention of cardiovascular disease in developed countries.1 Recommendations range from omission of adiposity measures, to inclusion of such measures as additional screening tests, to formal inclusion of such measures as risk factors in prediction models. This lack of consensus is an indication, in part, of uncertainty in the epidemiological evidence about the predictive value of adiposity measures in the context of information on other conventional risk factors. 5–12 In our investigation, individual participant data (ie, primary records) were collated and harmonised from all available relevant long-term prospective studies of cardiovascular disease, and analyses were restricted to individuals in these studies without a history of cardiovascular disease at the initial examination. These studies were identified by computerised searches of electronic databases (eg, PubMed), inspection of reference lists of relevant studies, manual searches, and discussion with investigators. The main analyses included people with concomitant information on body-mass index, waist circumference, waist-to-hip ratio, and several other conventional risk factors (ie, age, sex, blood pressure, history of diabetes, and total and HDL cholesterol). Interpretation The results from this study show that, whether assessed singly or in combination, body-mass index, waist circumference, and waist-to-hip ratio do not importantly improve prediction of first-onset cardiovascular disease when additional information exists on blood pressure, history of diabetes, and cholesterol measures.

Prospective Studies Collaboration (PSC)6 and the National Cancer Institute Cohort Consortium (NCICC),7 and a study of BMI in 1·1 millian Asians,35 which lacked information on waist circumference and waist-to-hip ratio, our study had concomitant data for each participant on BMI, waist circumference, and waist-to-hip ratio. By contrast with analyses from NCICC and the European Prospective Investigation into Cancer8 (EPIC), which lacked information on lipids, our study had concomitant information on lipids, blood pressure, and other conventional risk factors. The PSC, NCICC, and EPIC analyses all lacked non-fatal cardiovascular disease outcomes, whereas our study comprised 14 297 fatal or first-onset non-fatal outcomes of cardiovascular disease recorded during 1·87 million person-years at risk. Our study has twice as many cardiovascular disease outcomes than in the NCICC analysis and four times more than in the EPIC analysis, with less than 10% of data in our study overlapping with these previously published analyses. A unique feature of our study is inclusion of long-term serial assessments of BMI, waist circumference, and waist-to-hip ratio for up to 63 821 people. The generalisability of our findings, at least to populations in developed countries, is supported by broadly consistent results across 58 studies in 17 countries. Because 90% of the participants in this study were of European descent, additional studies are needed in people of non-European descent.36–38 Nevertheless, our finding that adiposity measures, 1093

Articles

whether assessed singly or in combination, do not importantly improve prediction of cardiovascular risk when additional information is available on blood pressure, history of diabetes, and lipids, should help to clarify divergent recommendations in cardiovascular disease guidelines about clinical measures of adiposity in assessment of risk. Contributors JD and DW drafted the report. DW did the analyses, with input from SKa, EDA, and AMW. All members of the writing committee provided critical revisions. All investigators shared individual data and had opportunities to contribute to the interpretation of the results and critical revision of the report. The data management team collated and harmonised the data. All members of the coordinating centre contributed to the data collection, harmonisation, analysis, and interpretation.

For the list of funding sources see http://www.phpc.cam.ac. uk/CEU

1094

The Emerging Risk Factors Collaboration Writing Committee: David Wormser, Stephen Kaptoge*, Emanuele Di Angelantonio*, Angela M Wood*, Lisa Pennells, Alex Thompson, and Nadeem Sarwar (University of Cambridge, Cambridge, UK); Jorge R Kizer (Weill Cornell Medical College, New York, NY, USA); Debbie A Lawlor (University of Bristol, Bristol, UK); Børge G Nordestgaard (University of Copenhagen, Copenhagen, Denmark); Paul Ridker (Brigham and Women’s Hospital, Boston, MA, USA); Veikko Salomaa (National Institute of Health and Welfare, Helsinki, Finland); June Stevens (University of North Carolina, Chapel Hill, NC, USA); Mark Woodward (University of Sydney, Sydney, NSW, Australia); Naveed Sattar (University of Glasgow, Glasgow, UK); Rory Collins (University of Oxford, Oxford, UK); Simon G Thompson (Medical Research Council Biostatistics Unit, Cambridge, UK); Gary Whitlock (University of Oxford, Oxford, UK); John Danesh (University of Cambridge, Cambridge, UK). *Denotes equal contribution. Investigators: ARIC A R Folsom, L E Chambless, J Stevens; ATTICA D B Panagiotakos, C Pitsavos, C Chrysohoou, C Stefanadis; AUSDIAB R Atkins, E L M Barr, J E Shaw, P Z Zimmet; BRHS P H Whincup, S G Wannamethee, R W Morris; BRUN S Kiechl, J Willeit, F Oberhollenzer, A Mayr; BWHHS S Ebrahim, D A Lawlor; CAPS J Yarnell, J Gallacher; CHARL P J Nietert, S E Sutherland, D L Bachman, J E Keil; CHS K J Mukamal, J R Kizer, I H de Boer (see http://www.chs-nhlbi.org for acknowledgments); COPEN A Tybjærg-Hansen, B G Nordestgaard, R Frikke-Schmidt; CUORE S Giampaoli, L Palmieri, S Panico, D Vanuzzo, L Pilotto; DRECE A Gómez de la Cámara, M A Rubio; EPESENCA D G Blazer, J M Guralnik, C L Phillips; EPICNOR K-T Khaw; FINRISK-92, FINRISK-97 V Salomaa, K Harald, P Jousilahti, E Vartiainen; FRAMOFF R B D’Agostino Sr, R S Vasan, C S Fox, M J Pencina; GOH R Dankner, A Chetrit, F Lubin; GOTO13 L Wilhelmsen, H Eriksson, K Svärdsudd, L Welin; GOTO33, GOTO43 A Rosengren, L Wilhelmsen, G Lappas, H Eriksson; GOTOW C Björkelund, L Lissner, C Bengtsson; HBS T E Strandberg, V Salomaa, R S Tilvis, T A Miettinen; HISAYAMA Y Kiyohara, H Arima, Y Doi, T Ninomiya; HOORN J M Dekker, G Nijpels, C D A Stehouwer; HPFS E B Rimm, J K Pai; IKNS H Iso, A Kitamura, K Yamagishi, H Noda; LASA D Deeg, J L Poppelaars; MESA A R Folsom, B M Psaty, S Shea; MOGERAUG2, MOGERAUG3 A Döring, W Koenig, C Meisinger; MORGEN W M M Verschuren, A Blokstra, H B Bueno-de-Mesquita; MOSWEGOT L Wilhelmsen, A Rosengren, G Lappas; MRCOLD A Fletcher; NHANES III R F Gillum, M Mussolino; NHS E B Rimm, S E Hankinson, J E Manson, J K Pai; NSHS K W Davidson, S Kirkland, J A Shaffer, M R Korin; OSAKA A Kitamura, H Iso, S Sato; PREVEND S J L Bakker, R T Gansevoort, H L Hillege; PRIME P Amouyel, D Arveiler, A Evans, J Ferrières; RANCHO E Barrett-Connor, D L Wingard, R Bettencourt; ROTT J Witteman, I Kardys, H Tiemeier, A Hofman; SHHEC H Tunstall-Pedoe, R Tavendale, G D O Lowe, M Woodward; SHS B V Howard, Y Zhang, L Best, J Umans; TARFS A Onat, G Hergenç, G Can; TOYAMA H Nakagawa, M Sakurai, K Nakamura, Y Morikawa; TROMSØ I Njølstad, E B Mathiesen, M L Løchen, T Wilsgaard; ULSAM J Ärnlöv, J Sundström, U Risérus, E Ingelsson; WHI-HaBPS S Wassertheil-Smoller, J E Manson; WHITE II E Brunner, M Shipley; WHS P Ridker, J Buring.

Data Management Team: M Walker, S Watson. Coordinating Centre: M Alexander, A S Butterworth, R Collins, E Di Angelantonio, O H Franco, P Gao, R Gobin, P Haycock, S Kaptoge, S R Kondapally Seshasai, S Lewington, L Pennells, N Sarwar, A Thompson, S G Thompson, M Walker, S Watson, I R White, A M Wood, D Wormser, J Danesh (principal investigator). Conflicts of interest AT has been a consultant for GlaxoSmithKline, and is now an employee of Roche. JRK and his institution have received research grants from the National Heart, Lung and Blood Institute, and diaDexus; and JRK has received payment for lectures from Merck. DAL’s institution has received funding from Department of Health Policy Research Programme and the British Heart Foundation. PR’s institution has received research grants from the National Institutes of Health, the National Cancer Institute, AstraZeneca, Novartis, and Merck. PR is listed as a co-inventor on patents relating to the use of inflammatory biomarkers in cardiovascular disease that are held by the Brigham and Women’s Hospital and have been licensed to Siemens and AstraZeneca; and has served as a consultant to Genzyme, Vascular Biogenics, Boerringher Ingelheim, and Amylin. NSat’s institution has received research grants from Allergan. RC and GW’s institution has received research funding from Merck, AstraZeneca, Bristol-Myers Squibb, Schering-Plough, Novartis, Bayer, and Solvay. RC’s institution has also received travel and accommodation expenses from Merck. JD has been a board member for Merck and Novartis; and has been a consultant for Novartis, Merck, GlaxoSmithKline, and Pfizer. JD’s institution has received research grants from the British Heart Foundation, BUPA Foundation, Denka, diaDexus, European Union, Evelyn Trust, Fogarty International Centre, GlaxoSmithKline, Merck, National Heart, Lung and Blood Institute, National Institute of Neurological Disorders and Stroke, Novartis, Pfizer, Roche, UK Biobank, UK Medical Research Council, and Wellcome Trust. All other members of the writing committee declare that they have no conflicts of interest. Acknowledgments The Emerging Risk Factors Collaboration Coordinating Centre is underpinned by a programme grant from the British Heart Foundation (RG/08/014) and grants from the UK Medical Research Council. Various sources have supported recruitment, follow-up, and laboratory measurements in the 122 cohorts contributing to the Emerging Risk Factors Collaboration. Investigators of several of these studies have contributed to a list naming some of these funding sources. References 1 Ferket BS, Colkesen EB, Visser JJ, et al. Systematic review of guidelines on cardiovascular risk assessment: which recommendations should clinicians follow for a cardiovascular health check? Arch Intern Med 2010; 170: 27–40. 2 WHO Consultation of Obesity. Obesity: preventing and managing the global epidemic. Geneva, Switzerland: World Health Organization, 2000. 3 National Institutes of Health. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults—the evidence report. Obes Res 1998; 6 (suppl 2): 51–209S. 4 Cooney MT, Dudina AL, Graham IM. Value and limitations of existing scores for the assessment of cardiovascular risk: a review for clinicians. J Am Coll Cardiol 2009; 54: 1209–27. 5 Yusuf S, Hawken S, Ôunpuu S, et al, on behalf of the INTERHEART Study Investigators. Obesity and the risk of myocardial infarction in 27 000 participants from 52 countries: a case-control study. Lancet 2005; 366: 1640–49. 6 Prospective Studies Collaboration. Body-mass index and cause-specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies. Lancet 2009; 373: 1083–96. 7 Berrington de Gonzalez A, Hartge P, Cerhan JR, et al. Body-mass index and mortality among 1·46 million white adults. N Engl J Med 2010; 363: 2211–19. 8 Pischon T, Boeing H, Hoffmann K, et al. General and abdominal adiposity and risk of death in Europe. N Engl J Med 2008; 359: 2105–20.

www.thelancet.com Vol 377 March 26, 2011

Articles

9

10

11 12

13 14 15 16

17

18

19 20

21 22 23

Canoy D, Boekholdt SM, Wareham N, et al. Body fat distribution and risk of coronary heart disease in men and women in the European Prospective Investigation into Cancer and Nutrition in Norfolk cohort: a population-based prospective study. Circulation 2007; 116: 2933–43. Taylor AE, Ebrahim S, Ben-Shlomo Y, et al. Comparison of the associations of body mass index and measures of central adiposity and fat mass with coronary heart disease, diabetes, and all-cause mortality: a study using data from 4 UK cohorts. Am J Clin Nutr 2010; 91: 547–56. Kizer JR, Biggs ML, Ix JH, et al. Measures of adiposity and future risk of ischemic stroke and coronary heart disease in older men and women. Am J Epidemiol 2011; 173: 10–25. Huxley R, Mendis S, Zheleznyakov E, Reddy S, Chan J. Body mass index, waist circumference and waist:hip ratio as predictors of cardiovascular risk—a review of the literature. Eur J Clin Nutr 2010; 64: 16–22. Jacobs EJ, Newton CC, Wang Y, et al. Waist circumference and all-cause mortality in a large US cohort. Arch Intern Med 2010; 170: 1293–301. Cook NR. Use and misuse of the receiver operating characteristic curve in risk prediction. Circulation 2007; 115: 928–35. Lloyd-Jones DM. Cardiovascular risk prediction: basic concepts, current status, and future directions. Circulation 2010; 121: 1768–77. The Emerging Risk Factors Collaboration. The Emerging Risk Factors Collaboration: analysis of individual data on lipid, inflammatory and other markers in over 1·1 million participants in 104 prospective studies of cardiovascular diseases. Eur J Epidemiol 2007; 22: 839–69. The Emerging Risk Factors Collaboration. C-reactive protein concentration and risk of coronary heart disease, stroke, and mortality: an individual participant meta-analysis. Lancet 2010; 375: 132–40. The Emerging Risk Factors Collaboration. Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: a collaborative meta-analysis of 102 prospective studies. Lancet 2010; 375: 2215–22. Emerging Risk Factors Collaboration. Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med 2011; 364: 829–41. Thompson S, Kaptoge S, White I, Wood A, Perry P, Danesh J, and the Emerging Risk Factors Collaboration. Statistical methods for the time-to-event analysis of individual participant data from multiple epidemiological studies. Int J Epidemiol 2010; 39: 1345–59. Easton D, Peto J, Babiker A. Floating absolute risk: an alternative to relative risk in survival and case-control analysis avoiding an arbitrary reference group. Stat Med 1991; 10: 1025–35. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003; 327: 557–60. Fibrinogen Studies Collaboration. Measures to assess the prognostic ability of the stratified Cox proportional hazards model. Stat Med 2009; 28: 389–411.

www.thelancet.com Vol 377 March 26, 2011

24

25 26

27 28 29 30 31 32 33

34 35 36 37 38

Pencina MJ, D’Agostino RB Sr, D’Agostino RB Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med 2008; 27: 157–72. Logue J, Thompson L, Romanes F, Wilson DC, Thompson J, Sattar N. Management of obesity: summary of SIGN guideline. BMJ 2010; 340: c154. Steptoe A, Doherty S, Rink E, Kerry S, Kendrick T, Hilton S. Behavioural counselling in general practice for the promotion of healthy behaviour among adults at increased risk of coronary heart disease: randomised trial. BMJ 1999; 319: 943–47. Lim SS, Gaziano TA, Gakidou E, et al. Prevention of cardiovascular disease in high-risk individuals in low-income and middle-income countries: health effects and costs. Lancet 2007; 370: 2054–62. WHO. Prevention of cardiovascular disease. Guidelines for the assessment and management of total cardiovascular risk. Geneva, Switzerland: World Health Organization, 2007. Lemieux I, Poirier P, Bergeron J, et al. Hypertriglyceridemic waist: a useful screening phenotype in preventive cardiology? Can J Cardiol 2007; 23 (suppl B): 23–31B. Van Gaal LF, Mertens IL, De Block CE. Mechanisms linking obesity with cardiovascular disease. Nature 2006; 444: 875–80. Grundy SM. Obesity, metabolic syndrome, and cardiovascular disease. J Clin Endocrinol Metab 2004; 89: 2595–600. The Emerging Risk Factors Collaboration. Major lipids, apolipoproteins, and risk of vascular disease. JAMA 2009; 302: 1993–2000. Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet 2002; 360: 1903–13. Stevens J, Cai J, Pamuk ER, Williamson DF, Thun MJ, Wood JL. The effect of age on the association between body-mass index and mortality. N Engl J Med 1998; 338: 1–7. Zheng W, McLerran DF, Rolland B, et al. Association between body-mass index and risk of death in more than 1 million Asians. N Engl J Med 2011; 364: 719–29. Asia Pacific Cohort Studies Collaboration. Central obesity and risk of cardiovascular disease in the Asia Pacific Region. Asia Pac J Clin Nutr 2006; 15: 287–92. Whincup PH, Gilg JA, Papacosta O, et al. Early evidence of ethnic differences in cardiovascular risk: cross sectional comparison of British South Asian and white children. BMJ 2002; 324: 635. Colin A, Adair LS, Popkin BM. Ethnic differences in the association between body mass index and hypertension. Am J Epidemiol 2002; 155: 346–53.

1095

Articles

The importance of early treatment with tranexamic acid in bleeding trauma patients: an exploratory analysis of the CRASH-2 randomised controlled trial The CRASH- collaborators*

Summary Lancet 2011; 377: 1096–101 Published Online March 24, 2011 DOI:10.1016/S01406736(11)60278-X See Online/Comment DOI:10.1016/S01406736(11)60396-6 See Online/Correspondence DOI:10.1016/S01406736(11)60317-6 *Members listed at end of Article and in online version of this paper Correspondence to: Clinical Trials Unit, London School of Hygiene and Tropical Medicine, London WC1E 7HT, UK [email protected]

Background The aim of the CRASH-2 trial was to assess the effects of early administration of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage. Tranexamic acid significantly reduced all-cause mortality. Because tranexamic acid is thought to exert its effect through inhibition of fibrinolysis, we undertook exploratory analyses of its effect on death due to bleeding. Methods The CRASH-2 trial was undertaken in 274 hospitals in 40 countries. 20 211 adult trauma patients with, or at risk of, significant bleeding were randomly assigned within 8 h of injury to either tranexamic acid (loading dose 1 g over 10 min followed by infusion of 1 g over 8 h) or placebo. Patients were randomly assigned by selection of the lowest numbered treatment pack from a box containing eight numbered packs that were identical apart from the pack number. Both participants and study staff (site investigators and trial coordinating centre staff) were masked to treatment allocation. We examined the effect of tranexamic acid on death due to bleeding according to time to treatment, severity of haemorrhage as assessed by systolic blood pressure, Glasgow coma score (GCS), and type of injury. All analyses were by intention to treat. The trial is registered as ISRCTN86750102, ClinicalTrials.gov NCT00375258, and South African Clinical Trial Register/Department of Health DOH-27-0607-1919. Findings 10 096 patients were allocated to tranexamic acid and 10 115 to placebo, of whom 10 060 and 10 067, respectively, were analysed. 1063 deaths (35%) were due to bleeding. We recorded strong evidence that the effect of tranexamic acid on death due to bleeding varied according to the time from injury to treatment (test for interaction p<0·0001). Early treatment (≤1 h from injury) significantly reduced the risk of death due to bleeding (198/3747 [5·3%] events in tranexamic acid group vs 286/3704 [7·7%] in placebo group; relative risk [RR] 0·68, 95% CI 0·57–0·82; p<0·0001). Treatment given between 1 and 3 h also reduced the risk of death due to bleeding (147/3037 [4·8%] vs 184/2996 [6·1%]; RR 0·79, 0·64–0·97; p=0·03). Treatment given after 3 h seemed to increase the risk of death due to bleeding (144/3272 [4·4%] vs 103/3362 [3·1%]; RR 1·44, 1·12–1·84; p=0·004). We recorded no evidence that the effect of tranexamic acid on death due to bleeding varied by systolic blood pressure, Glasgow coma score, or type of injury. Interpretation Tranexamic acid should be given as early as possible to bleeding trauma patients. For trauma patients admitted late after injury, tranexamic acid is less effective and could be harmful. Funding UK NIHR Health Technology Assessment programme, Pfizer, BUPA Foundation, and J P Moulton Charitable Foundation.

Introduction The CRASH-2 trial showed that administration of tranexamic acid to adult trauma patients with, or at risk of, significant haemorrhage, within 8 h of injury, significantly reduces all-cause mortality (relative risk [RR] 0·91, 95% CI 0·85–0·97; p=0·0035) with no apparent increase in vascular occlusive events.1 As a consequence of this trial, tranexamic acid has been incorporated into trauma treatment protocols worldwide. Results from the CRASH-2 trial raise some important questions. The trial was motivated by the evidence that tranexamic acid reduces bleeding in patients undergoing elective surgery, and the hypothesised mechanism was inhibition of fibrinolysis leading to improved effectiveness of haemostasis.2 However, no significant 1096

difference was recorded in transfusion requirements between the tranexamic acid and placebo groups, and the CRASH-2 trial did not measure the effect of this drug on fibrinolytic assays. Thus an alternative hypothesis is that tranexamic acid might act by reducing the pro-inflammatory effects of plasmin, rather than by improving haemostasis.3 There has also been discussion about which trauma patients should be treated with tranexamic acid. The CRASH-2 trial1 reported the few subgroup analyses that were prespecified in the statistical analysis plan. These analyses assessed the effect of tranexamic acid on the primary endpoint of all-cause mortality, according to time since injury, systolic blood pressure, Glasgow coma score, and type of injury. No strong evidence of www.thelancet.com Vol 377 March 26, 2011

Articles

heterogeneity was recorded for any of these analyses, suggesting that tranexamic acid is likely to be equally effective in all the subgroups examined. The focus on all-cause mortality was appropriate because it is an outcome that matters to patients and one that is not affected by the methodological problem of competing risks.4 However, the effect of the trial treatment on the biologically relevant outcome could have been diluted by outcomes on which tranexamic acid might have little or no effect. In response to these concerns, we undertook exploratory analyses of the effect of tranexamic acid on mortality due to bleeding. We report the same prespecified subgroup analyses but for the outcome that we hypothesise would be most affected by this drug, specifically mortality due to bleeding.

Heterogeneity in treatment effects across subgroups was assessed by a χ² test. We had prespecified that unless there was strong evidence against the null hypothesis of homogeneity of effects (ie, p<0·001), the overall RR would be considered the most reliable guide to the approximate RRs in all subgroups. To test the N

All causes of death

Bleeding death

Non-bleeding death

20 127

0·91 (0·85–0·97); p=0·0035

0·85 (0·76–0·96); p=0·0077

0·94 (0·86–1·02); p=0·13

≤1

7451

0·87 (0·76–0·97)

0·68 (0·57–0·82)

1·04 (0·89–1·21)

>1–3

6033

0·87 (0·77–0·97)

0·79 (0·64–0·97)

0·91 (0·78–1·05)

>3

6634

1·00 (0·90–1·13)

1·44 (1·12–1·84)

0·89 (0·78–1·02)

Overall Time to treatment (h)

χ² test of homogeneity

Methods

Study design and patients The background to the trial, methods, and baseline characteristics of the randomised patients have been previously reported.1 Briefly, we randomly allocated 20 211 adult trauma patients with, or at risk of, significant bleeding who were within 8 h of injury to either tranexamic acid (loading dose 1 g over 10 min followed by infusion of 1 g over 8 h) or matching placebo, with 99·6% follow-up. In most hospitals we used a local pack system for randomisation. After eligibility had been confirmed and the locally approved consent procedures had been completed, patients were randomly assigned by selection of the lowest numbered treatment pack from a box containing eight numbered packs. Apart from the pack number, the treatment packs were identical. The pack number was recorded on the entry form, which was sent to the Trial Coordinating Centre in London, UK. Hospitals with telephone access used a telephone randomisation service. Both participants and study staff (site investigators and trial coordinating centre staff) were masked to treatment allocation.

Statistical analysis The primary outcome was death in hospital within 4 weeks of injury, with cause of death described with the following categories: bleeding, vascular occlusion (myocardial infarction, stroke, and pulmonary embolism), multiorgan failure, head injury, and other. All analyses were by intention to treat. We examined the effect of the trial treatment on death due to bleeding subdivided by four baseline characteristics: (1) time from injury to treatment (≤1, >1–3, >3 h); (2) severity of haemorrhage as assessed by systolic blood pressure (≤75, 76–89, >89 mm Hg); (3) Glasgow coma score (severe 3–8, moderate 9–12, mild 13–15); and (4) type of injury (penetrating only, blunt plus blunt and penetrating). These were the same subgroup analyses that were reported previously, but for the outcome of death due to bleeding rather than for all-cause mortality. www.thelancet.com Vol 377 March 26, 2011

··

4·411 (p=0·11)

23·516 (p=0·0000)

2·537 (p=0·28)

Table : Relative risk (95% CI) of death with tranexamic acid, overall and by time to treatment

≤1 h (n=7451)

>1–3 h (n=6033)

>3 h (n=6634)

Age (years) Mean (SD)

33·4 (13·9)

35·0 (14·0)

35·5 (14·8)

<25

2283 (30·6%)

1557 (25·8%)

1773 (26·7%)

25–34

2360 (31·7%)

1832 (30·4%)

1882 (28·4%)

35–44

1356 (18·2%)

1177 (19·5%)

1262 (19·0%)

>44

1452 (19·5%)

1467 (24·3%)

1716 (25·9%)

Systolic blood pressure (mm Hg) ≤75

1380 (18·5%)

1012 (16·8%)

768 (11·6%)

76–89

1203 (16·1%)

1064 (17·6%)

1029 (15·5%)

>89

4857 (65·2%)

3955 (65·6%)

4821 (72·7%)

Heart rate (beats per min) <77

681 (9·1%)

450 (7·5%)

603 (9·1%)

77–91

1189 (16·0%)

971 (16·1%)

1326 (20·0%)

92–107

1888 (25·3%)

1562 (25·9%)

1625 (24·5%)

>107

3637 (48·8%)

2990 (49·6%)

3059 (46·1%)

Respiratory rate (breaths per min) <10

149 (2·0%)

82 (1·4%)

77 (1·2%)

10–29

6144 (82·5%)

4992 (82·7%)

5590 (84·3%)

>29

1077 (14·5%)

901 (14·9%)

923 (13·9%)

≤2

2450 (32·9%)

2140 (35·5%)

2217 (33·4%)

3–4

3472 (46·6%)

2773 (46·0%)

3110 (46·9%)

>4

1131 (15·2%)

963 (16·0%)

1257 (19·0%) 1494 (22·5%)

Capillary refill time (s)

Glasgow coma score Severe (3–8)

1000 (13·4%)

1124 (18·6%)

Moderate (9–12)

868 (11·7%)

915 (15·2%)

909 (13·7%)

Mild (13–15)

5577 (74·9%)

3994 (66·2%)

4214 (63·5%)

Asia

1213 (16·3%)

2475 (41·0%)

3656 (55·1%)

Africa

2490 (33·4%)

1437 (23·8%)

872 (13·1%)

Central and South America

2453 (32·9%)

1456 (24·1%)

1355 (20·4%)

North America, Europe, and Oceania

1295 (17·4 %)

665 (11·0%)

751 (11·3%)

Continents

Data are number (%), unless otherwise stated.

Table : Patient characteristics by time to treatment

1097

Articles

Tranexamic acid allocated

Placebo allocated

Risk ratio (95% CI)

≤1

198/3747 (5·3%)

286/3704 (7·7%)

0·68 (0·57–0·82)

>1–3

147/3037 (4·8%)

184/2996 (6·1%)

0·79 (0·64–0·97)

>3

144/3272 (4·4%)

103/3362 (3·1%)

1·44 (1·12–1·84)

Time to treatment (h)

⇢=23·516; p<0·0000 Systolic blood pressure (mm Hg) >89

146/6878 (2·1%)

163/6761 (2·4%)

0·88 (0·71–1·10)

76–89

110/1609 (6·8%)

114/1689 (6·7%)

1·01 (0·79–1·30)

≤75

233/1562 (14·9%)

295/1599 (18·4%)

0·81 (0·69–0·95)

168/1789 (9·4%)

186/1830 (10·2%)

0·92 (0·76–1·13)

93/1349 (6·9%)

121/1344 (9·0%)

0·77 (0·59–0·99)

228/6915 (3·3%)

265/6877 (3·8%)

0·86 (0·72–1·02)

Blunt

308/6788 (4·5%)

347/6817 (5·1%)

0·89 (0·77–1·04)

Penetrating

181/3272 (5·5%)

227/3250 (7·0%)

0·79 (0·66–0·96)

489/10 060 (4·9%)

574/10 067 (5·7%)

0·85 (0·76–0·96)

⇢=2·235; p=0·33 Glasgow coma score Severe (3–8) Moderate (9–12) Mild (13–15) ⇢=1·275; p=0·53 Type of injury

⇢=0·923; p=0·34 All deaths Two-sided p=0·0077 0·6

0·8 Tranexamic acid better

1

1·2 1·4 Tranexamic acid worse

1·6

Figure : Mortality due to bleeding by subgroups

independence of any observed treatment interactions we ran a logistic model including all possible interactions in the four prespecified baseline characteristics and treatment subgroups. A logistic regression was estimated with death due to bleeding as the dependent variable and treatment group and time to treatment as explanatory factors. We included an interaction parameter to allow for a proportional change in the odds ratio (OR) as time to treatment increases. ORs and 95% CIs were estimated for different times to treatment. CIs were calculated with a logistic model with time as a continuous term and an interaction term between time and tranexamic acid. We also ran a model with an interaction term for time to treatment squared to allow for a non-constant proportional change in the OR. The trial is registered as ISRCTN86750102, ClinicalTrials.gov NCT00375258, and South African Clinical Trial Register/Department of Health DOH-270607-1919.

Role of the funding source The sponsor of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author (IR) had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Results Of the 3076 deaths from all causes, death due to bleeding accounted for 1063 (35%). The risk of death due to 1098

bleeding was significantly reduced with tranexamic acid. 489 of 10 060 (4·9%) patients died because of bleeding in the tranexamic acid group versus 574 of 10 067 (5·7%) in the placebo group (RR 0·85, 95% CI 0·76–0·96; p=0·0077). We noted no significant effect on the risk of death for all other (non-bleeding) causes combined (table table 1). Table 2 shows the baseline characteristics of patients according to time to treatment. Figure 1 shows the results of the subgroup analyses for death due to bleeding. Time to treatment was unknown in nine participants. Treatment given 1 h or less from injury significantly reduced the risk of death due to bleeding (198/3747 [5·3%] in tranexamic acid group vs 286/3704 [7·7%] in placebo group; RR 0·68, 95% CI 0·57–0·82; p<0·0001). Treatment given between 1 and 3 h also reduced the risk of death due to bleeding (147/3037 [4·8%] vs 184/2996 [6·1%]; RR 0·79, 0·64–0·97; p=0·03). Treatment given more than 3 h after injury significantly increased the risk of death due to bleeding (144/3272 [4·4%] vs 103/3362 [3·1%]; RR 1·44, 1·12–1·84; p=0·004). We recorded strong evidence that the effect of tranexamic acid on death due to bleeding varied according to time from injury to treatment (p<0·0001). The evidence for interaction remained strong even after adjustment for interactions between the other prespecified baseline characteristics and treatment (p<0·0001; data not shown). The estimated OR of tranexamic acid on death due to bleeding when given immediately after injury was 0·61 www.thelancet.com Vol 377 March 26, 2011

Articles

Discussion The effect of tranexamic acid on death due to bleeding depends on the time between injury and onset of treatment. Early treatment with this drug seems to be much more effective than does late treatment. These results also raise the possibility that late treatment with tranexamic acid might increase the risk of death due to bleeding, although there was no evidence of any increase in all-cause mortality in patients treated after 3 h (table 1). This finding might indicate that patients treated with tranexamic acid beyond 3 h who died from bleeding might otherwise have died from some other non-bleeding cause (competing risks). If late administration does cause harm, this finding would be important since many bleeding trauma patients in lowincome and middle-income countries have long prehospital times. Indeed, about a third of trauma patients in the CRASH-2 trial were treated more than 3 h after the injury. The inclusion criteria in the CRASH-2 trial were entirely clinical, and reflect the situation that doctors are faced with in clinical practice. Patients were enrolled if the treating physician judged them to have ongoing significant haemorrhage, as evidenced by hypotension or tachycardia, or if they were considered to be at risk of significant haemorrhage. Some of the included patients might not have been actively bleeding. Any such misdiagnosis would have reduced the power of the trial to show an effect of tranexamic acid on mortality from bleeding, in which case the large and highly significant reduction in bleeding mortality in patients treated with this drug within 1 h of injury is particularly noteworthy. Because patients were randomly assigned soon after hospital admission, before the precise anatomical location of bleeding and other injury was known, we were unable to do a stratified analysis based on an anatomical assessment of injury severity. We acknowledge that this omission is a methodological weakness, since such an analysis might provide insight into the mechanism of action of tranexamic acid. However, since this information would not normally be available to treating clinicians, especially in view of the importance of early treatment, the clinical value of a stratified analysis based on anatomical injury severity is small. www.thelancet.com Vol 377 March 26, 2011

0 1 2 Time to treatment (h)

(95% CI 0·50–0·74). We estimated that this OR is multiplied by 1·15 (95% CI 1·08–1·23) for every hour that passes since the injury. Figure 2 shows how the OR and 95% CIs vary with time to treatment. The interaction term for time to treatment squared was not significant (OR=0·99; p=0·38). We recorded no evidence of heterogeneity for the subgroup analyses according to systolic blood pressure, Glasgow coma score at randomisation, or type of injury (figure 1). We detected no evidence of heterogeneity in the effect of tranexamic acid on the risk of non-bleeding deaths (table 1).

3 4 5 6 7 8

0·5

1·0

1·5 2·0 OR (95% CI) of tranexamic acid

2·5

3·0

Figure : E ect of tranexamic acid on death due to bleeding by time to treatment Shaded area shows 95% CI. OR=odds ratio.

Data for the time between injury and treatment were available for all but nine trial participants. Because in some cases the injury would not have been witnessed, this interval sometimes had to be estimated and might therefore be inaccurate. However, any inaccuracy would be independent of the trial treatment and therefore should not bias the results. The ascertainment of a death as a bleeding death might also have been inaccurate, but similarly any inaccuracy should be independent of the trial treatment. In clinical trials, a treatment is not often beneficial in one subgroup but harmful in another (qualitative interaction), and some trialists recommend that qualitative interactions should generally be disbelieved.5 The results of our analysis of the effect of tranexamic acid on death due to bleeding do, however, satisfy most of the criteria against which the credibility of subgroup results should be judged:6 time from injury was measured at baseline; the hypothesis that early treatment with tranexamic acid might be more effective was prespecified in the trial protocol; the interaction suggests a very low likelihood that chance explains the findings; the interaction remained significant after controlling for the non-significant interactions between treatment and the other prespecified baseline prognostic factors; the subgroup effect is large; and a biological rationale supports the interaction. Although this clinical trial was not powered to examine subgroup effects, the interaction recorded is large and highly significant.7 Nevertheless, we prespecified in our trial protocol that the main subgroup analyses would be undertaken for allcause mortality, and not for mortality due to bleeding. Even though we postulated that tranexamic acid would act by reducing bleeding, we focused on all-cause mortality because overall survival is most important to patients. However, in view of the significant reduction in all-cause mortality, most of which was attributable to the effect of tranexamic acid on death due to bleeding, and the biological rationale that this drug would act by 1099

Articles

Panel: Research in context Systematic review A 2011 Cochrane systematic review12 of antifibrinolytic drugs for acute traumatic injury identified two randomised trials of tranexamic acid in bleeding trauma patients, involving 20 451 patients. The review concluded that tranexamic acid safely reduces mortality in bleeding trauma patients without increasing the risk of adverse events. Interpretation Our results emphasise the importance of early administration of tranexamic acid and the need for caution in patients presenting several hours after the injury.

improving haemostasis, our analyses, although not prespecified, would seem justified. Acute severe trauma is associated with increased fibrinolysis that contributes to an early coagulopathy and increased mortality.8,9 Fibrinolysis can be assessed by measurement of fibrin degradation products, which include small protein fragments called D-dimers. Brohi and colleagues8 showed that D-dimer concentrations are raised in trauma patients at the time of hospital admission (median prehospital time 28 min), with the highest concentrations measured in the most severely injured patients.8 Similar results were recorded in a 2009 study from Japan that measured fibrin degradation product and D-dimers in 314 severe trauma patients.10 If this early increased fibrinolysis exacerbates bleeding and increases the risk of death, then we might expect that an antifibrinolytic drug such as tranexamic acid would be most effective in this period. Although we had anticipated that early treatment with tranexamic acid might be most effective, the apparent increase in the risk of death due to bleeding in patients treated more than 3 h after the injury is unexpected and cannot readily be explained. It could be a chance finding and there might be no real biological effect. However, patients in the late phase of trauma can develop thrombotic disseminated intravascular coagulation, and antifibrinolytics could be contraindicated in this period.10,11 Although disseminated intravascular coagulation is characterised by fibrin formation and coagulation, the rapid consumption of coagulation proteins can lead to their exhaustion, resulting in uncontrolled bleeding. The need to avoid giving an antifibrinolytic in this late phase was why we restricted trial inclusion to patients who were within 8 h of injury. The possibility that the change to a prothrombotic state might occur sooner than was previously expected is open to debate and needs further research. We should also bear in mind that patients who arrive at hospital many hours after injury are likely to differ from those who arrive early. For example, there could be an increased prevalence of hypothermia and 1100

acidosis. These or other differences could explain the decreased efficacy of tranexamic acid administration when given late. A 2011 systematic review of randomised controlled trials concluded that tranexamic acid safely reduces mortality in bleeding trauma patients.12 Our results strongly endorse the importance of early administration of tranexamic acid in bleeding trauma patients and suggest that trauma systems should be configured to facilitate this recommendation (panel panel). In patients presenting late (several hours after injury) the clinician should be more cautious and make an assessment of the individual benefits and risks of this treatment, since the drug is likely to be much less effective and possibly even harmful. To the extent that our subgroup analyses are consistent with the results of studies showing an early increased fibrinolytic coagulopathy, they support the hypothesis that tranexamic acid acts through the inhibition of fibrinolysis with improved haemostasis. Future research using the CRASH-2 trial data will develop a prognostic model to predict death due to bleeding.13 This model will facilitate further analysis of the effect of tranexamic acid according to baseline risk of haemorrhage death. Contributors All members of the Writing Committee, apart from AA and GG, attended a 2-day meeting in London, UK, at which the subgroup analyses were presented and discussed and the report was drafted. Both AA and GG contributed to the discussions and drafting by phone and in correspondence. CRASH-2 trial coordination Writing Committee: Ian Roberts (UK) (chair), Haleema Shakur (UK), Adefemi Afolabi (Nigeria), Karim Brohi (UK), Tim Coats (UK), Yashbir Dewan (India), Satoshi Gando (Japan), Gordon Guyatt (Canada), B J Hunt (UK), Carlos Morales (Colombia), Pablo Perel (UK), David Prieto-Merino (UK), Tom Woolley (UK). National coordinators: Jonathan Dakubo (Ghana), Tamar Gogichaishvili (Georgia), Nyoman Golden (Indonesia), Mario Izurieta (Ecuador), Hussein Khamis (Egypt), Edward Komolafe (Nigeria), Jorge Loría-Castellanos (Mexico), Jorge Mejía-Mantilla (Colombia), Jaime Miranda (Peru), Ángeles Muñoz (Spain), Vincent Mutiso (Kenya), Patrick Okwen (Cameroon), Zulma Ortiz (Argentina), María Pascual, CENCEC (Cuba), R Ravi (India), April Roslani (Malaysia), Stefan Trenkler (Slovakia), Annalisa Volpi (Italy), Surakrant Yutthakasemsunt (Thailand). Trial Coordinating Centre team: Ian Roberts (clinical coordinator, chief investigator), Haleema Shakur (trial manager), Pablo Perel (regional coordinator), Lin Barnetson (data manager), Maria Ramos (trial administrator), Lisa Cook (assistant trial manager, regional coordinator from 2007), Taemi Kawahara (assistant trial manager, regional coordinator from 2007), Eni Balogun (regional coordinator from 2006), Matthew Berle (trial assistant from 2007), Collette Barrow (assistant administrator from 2008), Tony Brady (programmer to 2006), Chris Rubery (data assistant from 2009), Jackie Wayte (UK nurse coordinator from 2008), Cynthia To (data assistant 2007–09). Steering Committee: Ian Franklin (chair), Brigitte Chaudhry, Tim Coats, Charles Deakin, Steve Goodacre, Beverley Hunt, David Meddings, Richard Peto, Ian Roberts, Peter Sandercock. Management Group: Ian Roberts (chair), Haleema Shakur, Tim Coats, Phil Edwards, Beverley Hunt, Maria Ramos. Data Monitoring and Ethics Committee: Rory Collins (chair), Adrian Grant, John Myburgh, Alex Baxter (independent statistician). For the full list of collaborators please see the online version of this Article and Lancet 2010; 376: 23–32.

www.thelancet.com Vol 377 March 26, 2011

Articles

Confl icts of interest Members of the Writing Committee declare that they have no conflicts of interest. Acknowledgments The London School of Hygiene and Tropical Medicine supported the core trial coordinating staff during the first year of the trial set-up. Funding for the run-in stage was provided by J P Moulton Charitable Foundation and the BUPA Foundation. A grant-in-aid for purchasing the tranexamic acid and placebo was provided by Pfizer. The main phase of this trial was funded by the UK NIHR Health Technology Assessment programme and will be published in full in the Health Technology Assessment journal series. The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the Department of Health. References 1 The CRASH-2 Collaborators. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet 2010; 376: 23–32. 2 Henry DA, Carless PA, Moxey AJ, et al. Anti-fibrinolytic use for minimising perioperative allogeneic blood transfusion. Cochrane Database Syst Rev 2007; 4: CD001886. 3 Levy JH. Antifibrinolytic therapy: new data and new concepts. Lancet 2010; 376: 3–4. 4 Lauer MS, Blackstone EH, Young JB, Topol EJ. Cause of death in clinical research: time for a reassessment? J Am Coll Cardiol 1999; 34: 618–20.

www.thelancet.com Vol 377 March 26, 2011

5 6 7 8 9 10 11 12 13

Yusuf S, Wittes J, Probstfield J, Tyroler HA. Analysis and interpretation of treatment effects in subgroups of patients in randomized clinical trials. JAMA 1991; 266: 93–98. Sun X, Briel M, Walter SD, et al. Is a subgroup effect believable? Updating criteria to evaluate the credibility of subgroup analyses. BMJ 2010; 340: c117. Assmann SF, Pocock SJ, Enos LE, Kasten LE. Subgroup analysis and other (mis)uses of baseline data in clinical trials. Lancet 2000; 355: 1064–69. Brohi K, Cohen MJ, Ganter MT, et al. Acute coagulopathy of trauma: hypoperfusion induces systemic anticoagulation and hyperfibrinolysis. J Trauma 2008; 64: 1211–17. Hess JR, Brohi K, Dutton RP, Hauser CJ, Holcomb JB, Kluger Y. The coagulopathy of trauma: a review of mechanisms. J Trauma 2008; 65: 748–54. Sawamura A, Hayakawa M, Gando S, et al. Disseminated intravascular coagulation with a fibrinolytic phenotype at an early phase of trauma predicts mortality. Thromb Res 2009; 1214: 608–13. Prentice CR. Basis of antifibrinolytic therapy. J Clin Pathol 1980; 33: 35–40. Roberts I, Shakur H, Ker K, Coats T, for the CRASH-2 Trial collaborators. Antifibrinolytic drugs for acute traumatic injury. Cochrane Database Syst Rev 2011; 1: CD004896. NIHR Health Technology Assessment Programme. Development and validation of a risk score for trauma patients with haemorrhage. August 2010. http://www.hta.ac.uk/protocols/200900220165.pdf (accessed March 6, 2011).

1101

Articles

CRASH-2 trial collaborators by country Albania (115)—National Trauma Centre Hospital: Fatos Olldashi, Mihal Kerçi, Tefik Zhurda, Klotilda Ruçi; Spitali Civil Durres: Arben Banushi. Argentina (51)—Hospital Ángel Cruz Padilla: Mario Sardón Traverso, Juan Jiménez; Hospital Regional Rio Grande: Jorge Balbi; Hospital "4 de Junio" Dr Ramon Carrillo: Christian Dellera; Hospital Castro Rendón: Silvana Svampa; Hospital San Martín de La Plata: Gustavo Quintana; Hospital Municipal de Agudos "Dr Leonídas Lucero": Gustavo Piñero; Hospital Interzonal General de Agudos “Dr Oscar Alende”: Jorge Teves. Australia (17)—Nepean Hospital: Ian Seppelt; Sir Charles Gairdner Hospital: David Mountain; John Hunter Hospital: Zsolt Balogh. Bangladesh (12)—United Hospital Limited: Maniruz Zaman. Belgium (51)—Sint-Vincentius Hospital: Patrick Druwé, Robert Rutsaert; Centre Hospitalier Regional de Namur: Guy Mazairac. Cameroon (124)—Tombel District Hospital: Fogang Pascal, Zognou Yvette, Djeuchon Chancellin; St Theresa's Catholic Medical Centre: Patrick Okwen; Bamenda Provincial Hospital: Jules Djokam-Liapoe; Bali District Hospital: Ernest Jangwa; Bafut District Hospital: Lawrence Mbuagbaw; Fundong District Hospital: Ninying Fointama; St John of God Medical Centre: Fogang Pascal. Canada (2)—Hamilton General Hospital: Frank Baillie. China (51)—Renji Hospital: Ji-yao Jiang, Guo-yi Gao, Yin-hui Bao. Colombia (2940)—University of Antioquia, Hospital San Vicente de Paul: Carlos Morales, Juan Sierra, Santiago Naranjo, Camilo Correa, Carolina Gómez; Hospital Universitário San Jose de Popayan: Jorge Herrera, Liliana Caicedo, Alexei Rojas, Henry Pastas, Hugo Miranda; Hospital Pablo Tobon Uribe: Alfredo Constaín, Mayla Perdomo, Diego Muñoz, Álvaro Duarte, Edwin Vásquez; Hospital San Andrés de Tumaco: Camilo Ortiz, Bernardo Ayala, Hernán Delgado, Gloria Benavides, Lorena Rosero; Fundación Clínica Valle del Lili: Jorge Mejía-Mantilla, Ana Varela, Maríaisabel Calle, José Castillo, Alberto García; Clínica las Americas: Juan Ciro, Clara Villa, Roberto Panesso; Hospital General de Medellin: Luz Flórez, Argemiro Gallego; Hospital San Felix ESE: Fabián Puentes-Manosalva, Leonor Medina, Kelly Márquez; Hospital Universitário del Caribe: Adalgiza Reyes Romero, Ricardo Hernández, Julio Martínez; Hospital Universitário San Jorge: Wilson Gualteros; Hospital San Rafael Tunja: Zulma Urbina, Julio Velandia; Clínica La Estancia SA: Federico Benítez, Adolfo Trochez; Fundación Hospital San José de Buga: Andrés Villarreal, Pamela Pabón; Hospital Civil de Ipiales: Hernán Delgado; Hospital Universitário Departamental Nariño: Héctor López; Hospital Universitário del Valle: Laureano Quintero; Hospital Universitário de Neiva: Andrés Rubiano; Hospital Manuel Uribe Ángel: Juan Tamayo. Cuba (575)—Hospital Clínico-Quirúrgico Docente "Saturnino Lora": Marjoris Piñera; Hospital General Universitário "Carlos Manuel de Céspedes": Daimi Martínez, Heriberto Martínez; Hospital Provincial Docente "Manuel Ascunce Domenech": Eugenio Casola; Hospital Universitário "Arnaldo Milián Castro": Mario Domínguez, Carlos Herrera; Hospital Universitário "Dr Gustavo Aldereguía Lima": Marcos Iraola, Omar Rojas; Hospital Abel Santamaría Cuadrado: Irene Pastrana, Daniel Rodríguez, Sergio Álvarez de la Campa; Hospital Miguel Enríquez: Thorvald Fortún; Hospital General Calixto García: Martha Larrea; Hospital Antonio Luaces Iraola: Lensky Aragón; Hospital Provincial Docente VI Lenin: Aida Madrazo. Czech Republic (17)—Research Institute for Special Surgery and Trauma: Petr Svoboda. Ecuador (1198)—Hospital Luis Vernaza: Mario Izurieta, Alberto Daccach, Mónica Altamirano, Antonio Ortega, Bolívar Cárdenas, Luis González; Hospital José Carrasco Arteaga: Marcelo Ochoa, Fernando Ortega, Fausto Quichimbo, Jenny Guiñanzaca; Hospital de Niños Dr Roberto Gilbert Elizalde: Ines Zavala, Sayra Segura; Hospital Naval Guayaquil: Johnny Jerez; Hospital Alcivar: Daniel Acosta; Hospital "Dr Rafael Rodríguez Zambrano": Fabián Yánez; Clínica De Especialidades Medicas "San Gregorio": Rubén Camacho. Egypt (2234)—Mataria Teaching Hospital: Hussein Khamis, Hossam Shafei, Ali Kheidr, Hani Nasr, Moetaz Mosaad, Safwat Rizk; Suez Canal University: Hesham El Sayed, Taha Moati, Emad Hokkam; Aswan Teaching Hospital: Mamdouh Amin, Hany Lowis, Medhat Fawzy, Nabil Bedir, Mohamed Aldars. El Salvador (84)—Hospital Nacional Rosales: Virginia Rodríguez, Juan Tobar, Jorge Alvarenga. Georgia (1783)—Tbilisi State University Clinical Hospital

1101.e1

'I Javakhishvili': Budu Shalamberidze, Elza Demuria, Nikoloz Rtveliashvili, Gocha Chutkerashvili, David Dotiashvili; Tbilisi First Hospital, University Clinic, Neurosurgery Center: Tamar Gogichaishvili, George Ingorokva, David Kazaishvili, Besik Melikidze, Natia Iashvili; Tbilisi City Hospital #1: Gia Tomadze, Manana Chkhikvadze, Leri Khurtsidze, Zviad Lomidze, Diana Dzagania; Tbilisi State Medical University ER Department: Nikoloz Kvachadze, Giorgi Gotsadze, Vakhtang Kaloiani; Institute of Critical Care Medicine: Nino Kajaia. Ghana (136)—Korle Bu Teaching Hospital: Jonathan Dakubo, Simon Naaeder, Priscilla Sowah; Nyinahin Government Hospital: Adamu Yusuf, Alhaji Ishak; Sogakope District Hospital: Paul Selasi-Sefenu; Methodist Hospital Wenchi: Ballu Sibiri; Effia Nkwanta Regional Hospital: Sampson Sarpong-Peprah; Saint Theresa's Hospital: Theodore Boro. India (4768)—Medical Trust Hospital Kochi: Kanjithanda Bopaiah, Kishore Shetty, Raja Subbiah, Lukman Mulla, Anand Doshi; Christian Medical College Ludhiana: Yashbir Dewan, Sarvpreet Grewal, Pradipta Tripathy, Jacob Mathew, Bharat Gupta; Aditya Neuroscience Centre: Anil Lal, Majulie Choudhury; Sri Sai Hospital: Sanjay Gupta, Smita Gupta, Arun Chug; Care Hospital: Venkataramana Pamidimukkala, Palaniappan Jagannath, Mohan Maharaj, Ramaraju Vommi, Naresh Gudipati; North Bengal Neuro Research Centre: W H Chhang; Sheth VS General Hospital and NHL Municipal College: Pankaj Patel, Nilay Suthar, Deepa Banker, Jyotish Patel; LTM Medical College and General Hospital: Satish Dharap, Ranjeet Kamble, Shraddha Patkar, Sushil Lohiya; Government Medical College and Associated Hospitals Jammu: Rakesh Saraf, Dinesh Kumar, Satish Parihar, Rahul Gupta; MKCG Medical College: Rasananda Mangual, Alagumuthu, Don Kooper, Chinmaya Mohapatra; Christian Medical College Hospital Vellore: Suresh David, Wesley Rajaleelan, Appas; KLE Hospital and Medical Research Centre: Ashok Pangi, Vivek Saraf, Santhosh Chikareddy; NKP Salve Institute of Medical Sciences and Lata Mangeshkar Hospital: Sushil Mankar, Anil Golhar, Rahul Sakhare, Nilesh Wagh; Sanjivani Diagnostics and Hospital: Anil Lal, Dhiman Hazarika; Parkar Hospital: Pratyush Chaudhuri; Jeevan Jyoti Hospital and Research Centre: Prakash Ketan; Mansarovar Hospital: Govindbhai Purohit, Yogesh Purohit, Mandakini Pandya; Postgraduate Institute of Medical Science Rohtak: Rakesh Gupta, Shashi Kiran, Saurab Walia; Goyal Hospital Jalna: Sonam Goyal, Sidhant Goyal, Satish Goyal; Government Medical College Chandigarh: Sanjay Gupta, Ashok Attri, Rajeev Sharma; Oberai Hospital: Ashok Oberai, Mahesh Oberai, Supriya Oberoi; Rajeev Gandhi Memorial Hospital and Research Centre: Gajendra Kant Tripathi; Calicut Medical College Hospital: Vijayan Peettakkandy, Premkumar Karuthillath, Pavithran Vadakammuriyil; Krishnamai Medical and Research Foundation's NIKOP Hospital: Jalindar Pol, Sunita Pol, Manisha Saste; St Stephen's Hospital: Subrat Raul, Shashi Tiwari, Neileino Nelly; Government Rajaji Hospital: M Chidambaram; Medical College Trivandrum: Viswanathan Kollengode, Sam Thampan; Sanjeevani Hospital: Sunder Rajan, Sushrut Rajan; Kamineni Hospital: Subodh Raju, Renuka Sharma; Sri Sakthi Hospital: Subbiah Venkatesh Babu, Chellappa Sumathi; Bhattacharya Orthopaedic and Related Research Centre: Protyush Chatterjee, Alok Agarwal; Sushrut Hospital: Hemant Magar, Meera Magar; All India Institute of Medical Sciences: Manmohan Singh, Deepak Gupta; GM Hospital (P): LTD: Anil Lal, Kamal Haloi; Government Medical College and Superspeciality Hospital Nagpur: Varsha Sagdeo, Pramod Giri; Government Medical College New Civil Hospital: Nimesh Verma, Ravi Jariwala, Ashish Goti; Chikitsa Hospital: Aman Prabhu-Gaonkar, Sagar Utagi; Apollo Health City: Mahesh Joshi, Ruchit Agrawal; Apex Neurotrauma and Superspeciality Hospital: Gopal Sharma, Gurvinder Saini; Neuro Center Gola Ghat: Vinod Tewari; NSCB Medical College: Yad Yadav, Vijay Parihar; BGS Global Hospital: Neelam Venkataramana, Shailesh Rao; Chettinad Hospital and Research Institute: Narayana Reddy; Sir Sayajirao General Hospital and Medical College Baroda: Virsing Hathila; Goyal Hospital and Research Centre Jodhpur: Vithal Das; Krishna Surgical Hospital and Trauma Care Centre: Kantibhai Agaja; Nizam's Institute of Medical Sciences: Aniruddh Purohit; Niramay Hospital: Akilesh Lahari; Apex Hospital Bhopal: Rajesh Bhagchandani; Dr Jeyasekharan Medical Trust: Bala Vidyasagar; Himalayan Institute of Medical Sciences: P K Sachan; Apollo Gleneagles Hospitals: Tanmoy Das; Civil Hospital Gandhinagar:

www.thelancet.com Vol 377 March 26, 2011

Articles

Sharad Vyas; Sukhdev Raj Soin Hospital: Sujoy Bhattacharjee; Sancheti Institute for Orthopaedics and Rehabilitation: Parag Sancheti; St James Hospital: T Manoj; Al Shifa Hospital: Mubarak Moideen; Anant Institute of Medical Sciences: Kailash Pansey; Vinayaka Mission Hospital: V P Chandrasekaran; Gauhati Medical College and Hospital: Kabul Saikia; Krishna Hospital and Medical Research Centre: Hoshedar Tata; Ruby Hall Clinic: Sanjay Vhora; Shreejee Hospital: Aniket Shah; Nazareth Hospital: Gordon Rangad; Ganga Hospital: S Rajasekaran; Vadamalayan Hospitals: S T Shankarlal; Devadoss Multispeciality Hospital: Sathish Devadoss; KIOT Hospital: M Saleem; Baby Memorial Hospital: Haroon Pillay; Bethany Hospital: Zulfiquer Hazarika; Suretech Hospital and Research Centre: Parikshit Deshmukh; Surya Hospital: S P Murugappan; Apollo Clinic Varanasi: Amit Jaiswal; Fortis Escorts Hospital: Deepak Vangani; Gokul Hospital and Trauma Centre: Prakash Modha; International Hospital Assam: Chawngrolien Chonzik; Lifeline Multispeciality Hospital: Megha Praveen; Meenakshi Mission Hospital and Research Centre: Vijaya Sethurayar; MOSC Medical College Hospital: Sojan Ipe; MS Ramaiah Memorial Hospital: Naresh Shetty; Saykhedkar Hospital and Research Centre: Aniket Shah; Shanti Mukand Hospital: R P Gupta; Shri KM Memorial Jain Heart and General Hospital: Vinod Jain; Usha Hospital: Ketan Shah. Indonesia (706)—Soebandi Hospital Jember: Moch Dwikoryanto; Sanglah General Hospital: Nyoman Golden, Kuning Atmadjaya, Ketut Wiargitha, Ketut Sudiasa, Gede Suwedagatha; Saiful Anwar General Hospital: Farhad Bal'afif; Dr Soetomo General Hospital: Vicky Budipramana, Tabrani, Agung Lemuel; Cipto Mangunkusumo Hospital: Susilo Chandra; Muhammadiyah Lamongan Hospital: Faisol Ama. Iran (134)—Nemazi Hospital: Ehsan Sherafatkazemzadeh, Ehsan Moradi, Alireza Sheikhi; Erfan Hospital: Ali Ziaee, Ahmad Fanaei; Loqman Medical Center: Esmaeil Hajinasrollah; Imam Hosain Hospital: Kamran Heidari. Iraq (392)—Diwaniyah College of Medicine: Bassim Mohammad, Najah Hadi. Italy (57)—Spedali Civili di Brescia: Giovanna Perone, Elena de Peri; Azienda Ospedaliera Di Parma: Annalisa Volpi. Jamaica (11)—University Hospital of the West Indies: Jean Johnson. Japan (9)—Fukuoka University Hospital: Masayoshi Abe. Kenya (31)—Kenyatta National Hospital: Vincent Mutiso, Benjamin Okanga; Kapenguria District Hospital: Daniel Ojuka. Malaysia (216)—Hospital University Science Malaysia: Baharudin Abdullah, Hishamuddin Rahman, Yazid Noh; Sungai Buloh Hospital: Sabariah Jamaluddin, Hasnah Dawal; University of Malaya Medical Centre: April Roslani, Chee-Wei Law, P Devashanti; Hospital Tengku Ampuan Rahimah: Yusof Wahab, Shanta Velaiutham; Ampang Hospital: Ridzuan Dato. Mexico (479)—Hospital General Regional 25: Jorge Loría, Erandy Montes, Eduardo Gómez, Víctor Cazales, Paúl Bautista; Hospital Gustavo Rovirosa: Raúl Bautista, David Ahumada, Edwin Hernández, Germahin Velásquez; Hospital General de Uruapan "Dr Pedro Daniel Martínez": Patricia Ortega, Gabriel Lira, Francisco Estrada; Hospital General Ecatepec Las Américas: Jorge Martínez; Hospital General La Perla: Juan Martínez; Hospital General de Ecatepec "Dr José María Rodríguez": Julio Casasola. Nigeria (2053)—National Hospital Abuja: Oluwole Olaomi, Yari Abubakar, Kanati Apollo, Olawale Badejo, Ogemdi Ihekire; University of Benin Teaching Hospital: Pius Iribhogbe, Olugbenga Oludiran, Emmanuel Obeta, Collins Okojie, Ernest Udefiagbon; Obafemi Awolowo University Teaching Hospitals: Edward Komolafe, Patrick Olaleye, Tochukwu Uzochukwu, Uvie Onakpoya; Irrua Specialist Teaching Hospital: Andrew Dongo; Olabisi Onabanjo University Teaching Hospital: Lateef Thanni; University College Hospital Ibadan: Adefemi Afolabi, Titilade Akinola, Adeyinka Ademola, Olusola Akute; Ahmadu Bello University Teaching Hospital: Lawal Khalid, Lawal Abubakar, Muhammad Aminu, Mike Ogirima; Baptist Medical Centre: Ambrose Attansey, Durodola Michael, Olaniran Aremu; University of Ilorin Teaching Hospital: Odebode Olugbenga, Udoffa Ukpong, Yusuf Salman; Enugu State University Teaching Hospital: Nene Obianyo, Chinenye Ani, Roderick Ezeadawi; LAUTECH Teaching Hospital: Oluwadiya Kehinde,

www.thelancet.com Vol 377 March 26, 2011

Agodirin Olaide; Federal Medical Centre Makurdi: Andrea Jogo, Terna Bitto; Nnamdi Azikiwe University Teaching Hospital: Stanley Anyanwu, Okechukwu Mbonu; Lagos State University Teaching Hospital: Mobolaji Oludara, Michael Somoye; Usmanu Danfodiyo University Teaching Hospital: Bello Shehu, Nasir Ismail; National Orthopaedic Hospital Enugu: Amechi Katchy; University of Calabar Teaching Hospital: Rowland Ndoma-Egba, Ngim Grace-Inah; University of Abuja Teaching Hospital: Zumnan Songden, Abdulrahman Abdulraheem; University of Uyo Teaching Hospital: Akpan Otu, Timothy Nottidge; Federal Medical Centre, Yenagoa: Domingo Inyang, David Idiapho; Seventh Day Adventist Hospital: Herb Giebel; Federal Medical Centre Birnin-Kebbi: Ramatu Hassan; Abia State University Teaching Hospital: Adeyinka Adisa; Wesley Guild Hospital: Akinbolaji Akinkuolie; Federal Medical Centre, Umuahia: Kalu Okam; University of Maiduguri Teaching Hospital: Abubakar Musa; National Orthopaedic Hospital, Igbobi: Ignatius Falope; University of Nigeria Teaching Hospital Enugu: John Eze. Peru (452)—Hospital Regional Docente de Trujillo: José Caballero, Wenceslao Azabache, Oscar Salirrosas; Hospital Nacional Hipolito Unanue: Alonso Soto, Elfi Torres, Gloria Ramírez, Mónica Pérez; Clinica Santa Ana: Cesar Malca; Hospital La Caleta: Juan Velez; Hospital Nacional Sergio E Bernales: Raul Yepez; Hospital de Apoyo de Sullana: Hernan Yupanqui; Hospital IV Essalud Huancayo: Pedro Lagos; Hospital Nacional Arzobispo Loayza: Diana Rodriguez; Hospital Municipal Los Olivos: Jorge Flores; Hospital Jose Cayetano Heredia: Anselmo Moya; Hospital Nacional Carlos Alberto Seguin Escobedo: Alejandro Barrionuevo; Hospital Nacional Dos De Mayo: Marco Gonzales-Portillo; Hospital Nacional Cayetano Heredia: Edgar Núñez. Saudi Arabia (70)—King Khalid University Hospital: Abdelazeem Eldawlatly, Mohammed Al Naami, Bilal Delvi; King Khalid National Guard Hospital: Walid Alyafi. Serbia (1)—Klinicki Centar Srbije: Branko Djurovic. Singapore (2)—National Neuroscience Institute: Ivan Ng. Slovakia (38)—FNsP Ružinov: Aktham Yaghi; NsP Poprad: Anton Laincz; NsP JA Reiman Hospital: Stefan Trenkler; Faculty Hospital F D Roosevelta: Jozef Valky. South Africa (76)—Dr George Mukhari Hospital: Mphako Modiba, Peter Legodi, Thomas Rangaka; George Provincial Hospital: Lee Wallis. Spain (23)—Hospital Universitário Virgen del Roció: Ángeles Muñoz; Hospital Ramón y Cajal de Madrid: Ana Serrano; Hospital Universitário Germans Trias i Pujol: Pilar Marcos; Hospital Torrecardenas: Martin Rubi; Hospital Universitário Virgen de la Victoria: Victoria de la Torre. Sri Lanka (103)—National Hospital of Sri Lanka: Ranjith Ellawala, Samitha Wijeratna, Lukshrini Gunaratna, Crishantha Wijayanayaka. Tanzania (64)—Muhimbili Orthopaedic Institute: Kitugi Nungu, Billy Haonga, Grenda Mtapa. Thailand (903)—Khon Kaen Regional Hospital: Surakrant Yutthakasemsunt, Warawut Kittiwattanagul, Parnumas Piyavechvirat, Tawatcahi Impool, Santipong Thummaraj; Pattani Hospital: Rusta Salaeh; Suratthani Hospital: Sakchai Tangchitvittaya; Bhumibol Adulyadej Hospital: Kamol Wattanakrai, Teerasak Jiravongbunrod, Chatchai Soonthornthum; Lampang Hospital: Surasak Meephant; Rayong Hospital: Pusit Subsompon; Roi-Et Hospital: Phaiboon Pensuwan; Phrae Hospital: Wicheanrat Chamnongwit. Tunisia (36)—Hospital Habib Thameur: Zouheir Jerbi, Abderraouef Cherif. UK (135)—University Hospital of North Staffordshire: Mark Nash; Royal London Hospital: Tim Harris; Leicester Royal Infirmary: Jay Banerjee; Nottingham University Hospitals NHS Trust: Ramzi Freij; Frenchay Hospital: Jason Kendall; Countess of Chester Hospital: Stephen Moore; Hull Royal Infirmary: William Townend; Royal Sussex County Hospital: Rowland Cottingham; Derby Hospitals NHS Trust: Dan Becker; Bedford Hospital NHS Trust: Stuart Lloyd; Royal Liverpool University Hospital: Peter Burdett-Smith; Colchester General Hospital: Kazim Mirza; Royal Lancaster Infirmary: Andrew Webster; Worthing Hospital: Suzanne Brady, Amanda Grocutt; Darent Valley Hospital: John Thurston; Hope Hospital: Fiona Lecky; Northern General Hospital: Steve Goodacre. Zambia (62)—University Teaching Hospital, Lusaka: Yakub Mulla, Dennis Sakala; Nchanga North General Hospital: Charles Chengo.

1101.e2

Clinical Picture

Placental vasa previa Kara B Markham, Richard W O’Shaughnessy Lancet 2011; 377: 1102 The Ohio State University, Columbus, OH, USA (K B Markham MD, R W O’Shaughnessy MD) Correspondence to: Dr Kara B Markham, The Ohio State University, 395 West 12th Avenue, 5th Floor, Columbus, OH 43210, USA [email protected]

In April, 2010, a 25-year-old woman at 33 weeks’ gestation in her first pregnancy presented to us for examination of a previously identified velamentous umbilical cord insertion into the placenta. Ultrasonography showed a fundal placenta, and the cord insertion site was not visible inserting into the placental surface. Transvaginal ultrasonography was performed because of our patient’s reports of vaginal spotting. Doppler ultrasonography showed fetal blood vessels coursing over the internal cervical os, consistent with a diagnosis of vasa previa (figure A). Uterine tocometry A

demonstrated frequent contractions. After administration of betamethasone for enhancement of fetal lung maturity, a caesarean section was done because of the risk of impingement or laceration of the vulnerable cord vasculature. She delivered a vigorous male neonate weighing 2115 g. Examination of the placenta was consistent with our diagnosis of a velamentous cord insertion site, with fetal blood vessels present over the cervix indicating vasa previa (figure B). At last follow-up in June, 2010, our patient and her infant were doing well. B

Figure: Placental vasa previa (A) Doppler ultrasonography showing fetal blood vessels coursing over the internal cervical os; and (B) placenta showing fetal blood vessels within the ammotic membrane, consistent with the diagnosis of a velamentous cord insertion site.

1102

www.thelancet.com Vol 377 March 26, 2011

Seminar

Ventricular septal defect Daniel J Penny, G Wesley Vick III

Ventricular septal defects account for up to 40% of all congenital cardiac malformations. The diagnosis encompasses a broad range of anomalies, including isolated defects and those associated with other congenital cardiac malformations. Presentation, symptoms, natural history, and management of ventricular septal defects depend on size and anatomical associations of the anomaly, patient’s age, and local diagnostic and interventional expertise. In this Seminar, we describe the anatomical range of ventricular septal defects and discuss present management of these malformations. Genetic determinants, diagnostic techniques, physiological considerations, and management challenges are examined in detail. Unfortunately, in many circumstances, evidence on which to guide optimum management is scarce. We present some longer term considerations of ventricular septal defects in adolescents and adults, with particular emphasis on patients with raised pulmonary vascular resistance and Eisenmenger’s syndrome.

Introduction Ventricular septal defect is one of the commonest congenital malformations of the heart, accounting for up to 40% of all cardiac anomalies.1 Frequency of this defect varies with age at examination, since many small malformations present at birth close shortly afterwards; it is also dependent on sensitivity of the examination technique. Prevalence in newborn babies of up to 5% has been reported from screening with highly sensitive colour doppler echocardiography.2 Most are tiny muscular defects that disappear during the first year of life. Since many patients can be asymptomatic, and many anomalies close with time, the precise prevalence of ventricular septal defect within populations varies between studies, depending on mode of diagnosis and age of the population. In reports in which echocardiography was used in the diagnostic algorithm, a prevalence of up to 3·94 per 1000 patients has been recorded, which is greater than in previous work that relied on either clinical examination or post-mortem investigations.3,4 Ventricular septal defect is not only a common isolated cardiac malformation but also an intrinsic component of several complex malformations, including tetralogy of Fallot or univentricular atrioventricular connection. It might also be associated with lesions, including transposition of the great arteries, congenitally corrected transposition, and aortic coarctation or interruption. However, in this Seminar we will concentrate on patients for whom ventricular septal defect is the predominant malformation.

Genetics and cause Our understanding of the origins of ventricular septal defect is limited by our knowledge of mechanisms that lead to normal cardiac septation. At present, information suggests that the septum has both mesenchymal and muscular components.5 The mesenchymal element originates mainly from fusion of the conotruncal and atrioventricular endocardial cushions. Mechanisms that initiate development of the muscular septum are less well defined, and at least two processes have been proposed. Some researchers postulate that the muscular septum forms from coalescence of the part of the ventricular wall www.thelancet.com Vol 377 March 26, 2011

that is interposed between the enlarging free walls of the developing right and left ventricles, therefore, as the ventricular cavities become deeper the septum grows passively inwards.6 An alternative hypothesis suggests that the muscular septum originates from a cluster of cells, the so-called primitive interventricular septum, which expands actively towards the cushions of the atrioventricular canal.7 Several factors probably lead to development of ventricular septal defects. Failure of complete formation of the primitive interventricular septum could contribute to trabecular defects, although many muscular defects in the trabecular septum probably result from excessive undermining beneath and between trabeculae, during formation of the trabecular part of the septum. Failure of fusion of the atrioventricular cushions—with each other or with the primary septum—could result in an inlet defect, whereas malalignment or poor development of outlet cushions might add to outlet defects. Finally, failure of complete closure of the area that forms the membranous septum, in association with incomplete development of components of the muscular septum, could contribute to a perimembranous defect. Most forms of congenital heart disease, including ventricular septal defect, have multifactorial origins.8,9 An underlying inherited genetic predisposition could act synergistically with epigenetic factors, direct and indirect environmental causes, and purely stochastic effects to produce cardiac anomalies. Monogenic defects are, in some cases, clearly causative.8 Such defects have attracted much interest

Lancet 2011; 377: 1103–12 Published Online February 23, 2011 DOI:10.1016/S01406736(10)61339-6 Department of Cardiology, Royal Children’s Hospital, and Murdoch Children’s Research Institute, Melbourne, Vic, Australia (D J Penny MD); and Baylor College of Medicine, and Texas Children’s Hospital, Houston, TX, USA (G W Vick III MD) Correspondence to: Dr D J Penny, Texas Children’s Hospital, 6621 Fannin, Houston, TX 77303, USA [email protected]

Search strategy and selection criteria We searched PubMed with the term “ventricular septal defect”. We mainly selected publications from the past decade but did not exclude commonly cited references and highly regarded older publications. In view of recent advances in management of patients with Eisenmenger’s syndrome, we undertook a separate search of PubMed with the term “Eisenmenger syndrome”. Further, we reviewed reference lists of articles identified from these searches and selected those we judged especially relevant.

1103

Seminar

because their molecular characterisation has facilitated identification of important constituents of signalling pathways that govern cardiac development.10–12 Mutations in the transcription factors TBX5 and GATA4 have received particular attention. These factors are coexpressed in the heart and their interaction is vital for normal cardiac septation.13 TBX5 is expressed not only in the heart but also in the upper limb buds and eyes. The mutation reported most often in this transcription factor is associated with the autosomal dominant Holt-Oram syndrome,14,15 characterised by abnormalities of forelimbs and several cardiac malformations, including ventricular septal defect. A TBX5 polymorphism is also associated with ventricular septal defect (without limb abnormalities) in the Chinese Han population.16 Researchers have identified GATA4 sequence variants in familial cases of septal defects (particularly atrial) and in some patients with sporadic ventricular septal defect.17–20 Environmental factors such as teratogens, maternal infections, and untreated maternal metabolic illnesses (eg, phenylketonuria and pregestational diabetes) have been associated with ventricular septal defect.21 Purely stochastic events could also have an important role. Cardiac development is very elaborate, requiring precise operations for successful completion,22,23 which are likely to malfunction occasionally.

Anatomy Ventricular septal defect, in many respects, can be deemed one of the simpler forms of congenital malformation of the heart. However, no universal consensus exists for its classification.24–27 To be brief, we will present one system to describe the anatomy of ventricular septal defects because we believe the controversy surrounding these descriptors is beyond the scope of this Seminar. Broadly speaking, defects can be classified according to their location, either within the muscular septum (muscular defects) or at its margins. Ventricular septal defects at the margins of the muscular septum can be related to hinge-points of the leaflets of the atrioventricular valves (perimembranous), those of the arterial valves (juxta-arterial or subarterial), or both (figure 1).28 Muscular defects are located within the muscular septum. They are surrounded exclusively by muscular rims and, when viewed from the cavity of the right ventricle, can open into the right-ventricular inlet, outlet, or apex. Perimembranous defects open into the right ventricle where the subpulmonary outflow tract turns superiorly relative to the atrioventricular junction. Such malformations are characterised by presence of fibrous continuity between leaflets of the tricuspid and aortic valves. They can extend to open into either the inlet or outlet of the right ventricle (resulting in deviation of the

Pulmonary valve

Doubly committed and juxta-arterial

Aortic valve

Perimembranous Muscular rim

Pulmonary-aortic-tricuspid continuity

Muscular

Figure 1: Location of various types of ventricular septal defect (Left) Location of defects viewed from the right ventricle. (Upper right) Typical doubly committed and juxta-arterial defect. (Lower right) Doubly committed, juxta-arterial, and perimembranous defect. Modified from reference 28 with permission of Elsevier.

1104

www.thelancet.com Vol 377 March 26, 2011

Seminar

outlet septum) or they might be large enough to open to all parts of the ventricle, the so-called confluent defect. Doubly committed and juxta-arterial defects (also referred to as subarterial or supracristal defects) are found in an area that, in the normal heart, constitutes a freestanding tube of muscular tissue—the muscular infundibulum, which supports the pulmonary valve. An anomaly in this region will result in characteristic continuity between aortic and pulmonary valves. Most usually, these malformations have a posteroinferior rim of muscle, although they can extend into the perimembranous zone; thus, fibrous continuity also arises with the tricuspid valve, the so-called doubly committed and juxta-arterial and perimembranous defect (figure 1).

Pathophysiology Several key components determine the pathophysiological response to a ventricular septal defect. Primary factors are the amount and direction of interventricular shunting and the degree of volume loading to the cardiac chambers. Secondary effects include prolapse of the aortic valve and obstruction to the pulmonary or systemic outflow tract. The amount of interventricular flow is determined by the size of the defect and relative resistances of pulmonary and systemic vascular beds. Small malformations themselves, so-called restrictive defects, provide intrinsic resistance to flow. The size of flow through larger nonrestrictive defects is determined by relative resistances of pulmonary and systemic vascular beds. No agreed precise criteria exist for definition of a non-restrictive defect, although various cutoffs have been proposed, according to cross-sectional area of the defect versus area of the aortic orifice, diameter relative to body surface area, or velocity of flow across the malformation. When a defect is non-restrictive, major determinants of the resultant interventricular flow and symptoms are relative resistances of the pulmonary and systemic vascular beds. Importantly, this relation can be very variable and dependent, in particular, on age of the patient. Left-to-right shunting might initially be minimal in babies, with fairly large defects due to high pulmonary vascular resistance characteristic of the early neonatal period. As pulmonary vascular resistance falls, left-toright interventricular shunting rises and the patient becomes increasingly symptomatic due to excessive pulmonary blood flow.29 In some patients with ventricular septal defects, pulmonary vascular disease can develop in later childhood or in early adult life. In a few individuals, the typical postnatal decline in pulmonary vascular resistance could be delayed or arrested in the presence of a ventricular septal defect; therefore, they might never develop symptoms attributable to excessive leftto-right shunting and only present at a later stage with signs of pulmonary vascular disease. If a large lesion is left uncorrected then, over time, the amount of interventricular left-to-right shunting could decrease www.thelancet.com Vol 377 March 26, 2011

and, eventually, its direction might reverse, leading to cyanosis and Eisenmenger’s syndrome.30,31 Eisenmenger’s syndrome—resulting from chronic elevations of pressure and flow—is associated with functional and structural alterations within the pulmonary vasculature.32,33 Key functional modifications are increased pulmonary vasoreactivity and resistance and structural microvascular changes, which include medial hypertrophy, migration of smooth muscle distally into typically unmuscularised microvessels, and ultimately, formation of so-called plexiform lesions.34 Abnormalities within the endothelium contribute to all stages of this progression. Activation of endothelial-dependent vasoconstrictor pathways and aberrations of endogenous endotheliumdependent vasodilator processes play a part in functional and structural remodelling, both in animal models of shunt-related pulmonary vascular disease35,36 and in the clinical condition.37 In patients with large ventricular septal defects without pulmonary vascular disease, a rise in volume loading of the left atrium and ventricle (due to increased pulmonary blood flow and, in turn, augmented pulmonary venous return) results in left heart dilation throughout the cardiac cycle. In response to the amplification in wall stress, eccentric left-ventricular hypertrophy develops. Presence of relevant longstanding pulmonary hypertension could ultimately lead to right-ventricular hypertrophy and dilation. These features will predominate as a patient enters the terminal stages of severe Eisenmenger’s syndrome, which is characterised typically by pending or actual right heart failure. Secondary structural cardiac anomalies could contribute substantially to the clinical course of patients with ventricular septal defects. Continued surveillance of all affected individuals is essential to monitor development of these defects, because they can affect clinical management. Malformations located near the aortic valve (doubly committed, perimembranous, or muscular) can be complicated by aortic-valve prolapse and regurgitation, which result from generation of Venturi forces, in which the high-velocity jet sucks the leaflet of the aortic valve into the restrictive defect.38 Several additional mechanisms could contribute to this effect, including absence of structural support for leaflets and abnormal commissural suspension. Mid-cavity obstruction of the right ventricle due to hypertrophy of muscle bands creates the entity known as double-chambered right ventricle. This process results in formation of a proximal high-pressure chamber and a distal low-pressure chamber within the cavity of the right ventricle.39 In some patients, modest anterior deviation of the outlet septum can happen. The reported prevalence of double-chambered right ventricle in individuals with ventricular septal defect varies widely within published work. An association is well recognised between double-chambered right ventricle and discrete subaortic stenosis.40 1105

Seminar

RV

LA

Ao RV

LV

LV

LA

Figure 2: Diagnosis of muscular ventricular septal defect Echocardiograms show apical four-chamber (left) and parasternal long-axis view (right) of a large apical muscular defect (arrows). LA=left atrium. LV=left ventricle. RV=right ventricle. Ao=aortic valve.

RV Ao

TV RA

LA

Figure 3: Diagnosis of perimembranous defect Echocardiogram shows parasternal short-axis view of a perimembranous defect, with fibrous continuity between tricuspid (TV) and aortic (Ao) valves (arrows). RV=right ventricle. LA=left atrium. RA=right atrium.

Defects affecting the muscular outlet septum can be associated with its posterior deviation into the leftventricular outflow tract, resulting in muscular subaortic stenosis. This malformation usually presents in early infancy and might be associated additionally with aortic coarctation or interruption.

Diagnosis Clinical examination can show evidence of volume loading of the left ventricle from a large ventricular septal defect, with lateral displacement of the cardiac apex. A pansystolic murmur could be present, with intensity of the murmur indicating velocity of flow across the malformation, such that smaller defects are generally loudest and can be associated with a thrill. Large anomalies—leading to an increase in mitral inflow—could generate a diastolic rumble at the apex. Patients with Eisenmenger’s syndrome typically have cyanosis and clubbing, with a prominent right-ventricular heave, an accentuated pulmonary component of the second heart sound, and usually no murmur. 1106

The electrocardiogram can be normal in patients with small ventricular septal defects. Volume loading of the left ventricle might result in left-ventricular hypertrophy, whereas raised right-ventricular pressure due to either pulmonary hypertension or obstruction to the pulmonary outflow tract could lead to right-ventricular hypertrophy. Cross-sectional echocardiography is the mainstay of modern diagnosis of ventricular septal defect.41–43 The echocardiographer will aim not only to undertake a comprehensive study of the heart—based around a sequential approach—but also to provide several key pieces of data related to the malformation, including: size and location (figure 2); anatomical relations to tricuspid, aortic, and pulmonary valves (figure 3); associated obstruction to outflow from right or left ventricles; associated prolapse of aortic valve; assessment of rightventricular pressure; and assessment of the amount of loading of the right and left heart (left-ventricular dimension at end-systole and end-diastole should be measured and normalised for body surface area). Integration of spectral and colour doppler with twodimensional (2D) echocardiography greatly assists with identification and characterisation of ventricular septal defects.44–46 Reliable estimates of right-ventricular and pulmonary artery pressures, and of pressure differences between left and right ventricles, can usually be obtained with 2D-directed continuous-wave doppler.47 The need for cardiac catheterisation to obtain pressure data is thereby eliminated in most cases. The echocardiographer should also assess extracardiac vascular structures, since clinically important anomalies of the aorta—especially coarctation—and pulmonary arteries, pulmonary veins, and systemic veins can be seen. Transoesophageal echocardiography has assumed an important role in intraoperative assessment of ventricular septal defect because it greatly facilitates confirmation of repair and early identification and correction of any residual lesion.48,49 Three-dimensional echocardiography is becoming widely available and could provide important diagnostic assistance for assessment of unusually positioned ventricular septal defects and those associated with complex congenital heart malformations.50 Nowadays, cardiac catheterisation is undertaken rarely in patients with uncomplicated defects. This procedure is usually reserved either to measure pulmonary vascular resistance in individuals with suspected or actual pulmonary vascular disease or to close the malformation by a transcatheter approach. MRI is used increasingly to assess patients with many forms of congenital heart disease, both before and after surgery. Although, in most individuals with ventricular heart defects, adequate diagnostic information can be obtained from clinical examination and echocardiography, MRI might be of use, particularly in patients with poor echocardiographic images.51 MRI could provide additional useful information for definition of anatomy in individuals with complex defects—eg, those with www.thelancet.com Vol 377 March 26, 2011

Seminar

double-chambered right ventricle.52 This imaging technique is especially useful for measurement of pulmonary-to-systemic flow ratio. With MRI, accurate quantification of stroke volume of both the right and left ventricles can be made.53,54 Pulmonary-artery resistance can also be measured, provided simultaneous catheter or doppler determination of pulmonary-artery pressure is done.55 Associated extracardiac defects, such as coarctation of the aorta and pulmonary-artery branch stenoses, which are sometimes difficult to visualise by echocardiography in older patients, can be delineated clearly.56 Fetal diagnosis of ventricular septal defect is becoming increasingly frequent as imaging techniques improve. Interest has been shown in outcomes for fetuses with malformations seen by colour flow-mapping alone in the presence of apparently normal greyscale cross-sectional imaging. In a series of 146 such fetuses, 35 had extracardiac anomalies. Of 113 babies assessed a year after birth, the defect had closed in utero in 37 and during the first year of postnatal life in 50.57

Clinical scenarios

Symptomatic young infant with pulmonary hypertension A baby with such symptoms would typically become breathless with failure to thrive within the first few weeks of life. In this situation, we would usually recommend surgery within 3 months of birth. While awaiting surgery, medical treatment with low doses of diuretics with or without angiotensin-convertingenzyme inhibitors is typically used, although the evidence-base for these strategies is sparse. Monitoring of blood pressure and renal function should be done because renal failure and hypotension have been reported, particularly with angiotensin-convertingenzyme inhibitors.58 The early postoperative period can be complicated by pulmonary hypertension, which results from increased pulmonary vasoreactivity after cardiopulmonary bypass. Use of inhaled nitric oxide to treat postoperative pulmonary hypertension has become widespread,59 although in a Cochrane review, the paucity of data supporting its use early after cardiac surgery was highlighted.60

Asymptomatic patient without pulmonary hypertension but with volume overloaded left heart In this scenario, many centres would recommend closure of ventricular septal defects with the aim of avoiding potential late left-ventricular dysfunction secondary to ongoing dilation. In an observational study of 96 patients (mean follow-up almost 8 years), without any intervention, the left-ventricular end-diastolic dimension Z score fell in 29 of 33 patients and declined to less than 2 in 26 of these.61 Although this series was small, the findings suggest that the optimum approach for this group of patients, who would typically be most suitable for transcathether closure, might instead be conservative. www.thelancet.com Vol 377 March 26, 2011

Asymptomatic patient with small ventricular septal defect and no left-ventricular dilation Much information is available from natural history studies on the long-term outlook for individuals with ventricular septal defects. Gabriel and colleagues62 reported long-term outcomes in 229 patients with malformations judged “not to require surgical closure during childhood”. At a mean age of 30 years, mortality was zero, 95% were symptom-free, and left-ventricular size was normal in 89% and borderline in 10%. At the time of follow-up, four of 222 patients had experienced an episode of endocarditis. This study’s findings confirmed that with careful selection of patients, a conservative approach is warranted in this subgroup. In a regional cohort of 290 patients with ventricular septal defect, spontaneous closure was noted in 123 of 180 cases with completely muscular borders compared with only 31 of 107 with perimembranous defects. Furthermore, muscular malformations closed spontaneously in children aged up to 88 months, whereas no perimembranous defect closed in infants older than 62 months.63 Of 450 patients with perimembranous defects, at mean follow-up of 3 years, subaortic ridge developed in 26 (6%), aortic-valve prolapse in 53 (12%), and aortic regurgitation in 33 (7%).64

Asymptomatic patient with small defect and prolapse or regurgitation of aortic valve Best management for this population has historically been controversial, and up to now, no randomised trials have been published to define the optimum strategy. In an audit of patients who underwent surgery for aortic regurgitation in the setting of ventricular septal defect, those with severe preoperative regurgitation had less favourable long-term outcomes and a higher requirement for reoperation because of suboptimum valve repair.65 As a result, individuals with perimembranous ventricular septal defects and more than trivial aortic regurgitation should be referred for surgery.38 A reduced threshold for surgery might be justifiable in patients with juxta-arterial defects because of their high risk of aortic regurgitation and low rate of spontaneous closure.38

Patient with Eisenmenger’s syndrome Until recent times, treatment of individuals with Eisenmenger’s syndrome was only supportive.32,66 Dehydration and exposure to high altitudes should be avoided because these situations compound pre-existing hyperviscosity and arterial hypoxaemia. Historically, venesection to reduce the effects of polycythaemia was routine in many centres, although its benefits are questionable in asymptomatic patients.67 Indeed, venesection can worsen iron deficiency and exercise intolerance and can amplify risk of stroke.67,68 Anticoagulation has been used to manage Eisenmenger’s syndrome in the past, although supportive evidence is 1107

Seminar

JV

LV

FA

Figure 4: Transcatheter closure of muscular defect Patient with complex disease is paced for congenital complete heart block. (Left) A so-called railroad is created from the jugular vein (JV), across the defect, to the femoral artery (FA) to allow stable positioning of the delivery sheath. (Middle) The device is placed across the defect by advancing it through the delivery system, which has been introduced through the jugular vein. LV=left ventricle. (Right) The device is released and a ventriculogram shows it is well placed and there is no residual shunting.

LA Ao

RV

LV

Figure 5: Percutaneous closure of a perimembranous defect Echocardiogram after device closure of a perimembranous defect. The device is close to the aortic valve (Ao) and the conduction axis (see figure 1). LA=left atrium. LV=left ventricle. RV=right ventricle.

scarce and bleeding risk could be considerable.69 Female patients must be made aware that pregnancy is associated with substantial maternal and fetal risk (see Issues in adults with ventricular septal defects). Recognition of the role of disrupted endothelial messengers in pathogenesis of Eisenmenger’s syndrome has broadened pharmacotherapeutic options for this disorder.33 Endothelial-based treatments, particularly those aimed either at blockade of the potent vasoconstrictor endothelin 1 or at prevention of catabolism of the nitric oxide-dependent vasodilator cGMP, are used increasingly in this population of patients.70–72 In the BREATHE-5 study, bosentan—a dual endothelin receptor-antagonist—lengthened 6-min walk distances in a randomised placebo-controlled trial over 16 weeks73 and in an open-label extension.74 These improvements continued for an additional 24 weeks. Findings of an open-label study showed increases in functional capacity that were maintained, particularly in adults.75 Improvements in quality of life and functional capacity were reported in patients with Eisenmenger’s 1108

syndrome in response to oral sildenafil, a phosphodiesterase inhibitor that raises cGMP levels.76 Advanced endothelial-based treatments increase survival in individuals with Eisenmenger’s syndome.77 These advances, combined with the observation that many patients with Eisenmenger’s syndrome are responsive to vasodilators in the catheterisation laboratory,78 raise the possibility that aggressive endothelial-based treatments could restore operability in individuals who had previously been judged inoperable. Several published reports accord with this idea, although systematic studies are scarce.79

Endocarditis Traditionally, antibiotic prophylaxis was recommended routinely in patients with ventricular septal defects to prevent procedure-associated endocarditis. This guidance is based on recognition that such individuals are at increased risk of endocarditis, that this disorder could result from bacteraemia, that dental procedures might result in bacteraemia, and that treatment with antibiotics might reduce risk of bacteraemia and endocarditis. However, later evidence indicates that endocarditis is most likely to result from chronically poor dental hygiene and activities of daily living, which— coupled with the paucity of data in support of the effectiveness of antibiotic prophylaxis for prevention of endocarditis—has resulted in revised guidelines. These recommendations suggest that patients with uncomplicated ventricular septal defects do not need antibiotics, but they put strong emphasis on primary prevention of dental infections, with meticulous daily dental hygiene and regular dental review. However, antibiotic prophylaxis for dental and other procedures continues to be recommended for 6 months after complete surgical or transcatheter closure of a ventricular septal defect and indefinitely when a residual defect is present in relation to patch material, because this situation could inhibit endothelialisation.80 www.thelancet.com Vol 377 March 26, 2011

Seminar

Closure of ventricular septal defects Surgery

Patch closure of a ventricular septal defect through sternotomy, with cardiopulmonary bypass, has been done for more than 50 years. With enhanced selection of patients, early surgery, and advances in perioperative care, operative mortality is low and substantial postoperative morbidity is rare.81 Usually, access to the defect is obtained through either the atrioventricular or semilunar valves, thus avoiding ventriculotomy. In some individuals, intraoperative temporary detachment of anterior and septal leaflets of the tricuspid valve can increase exposure of the defect.82 Most patients after surgery report normal quality of life, compared with age-matched controls, although behavioural and school performance difficulties might be present in early childhood.83 With expansion of cardiac surgery to the developing world,84 interest has grown in techniques to close defects in patients who did not have access to surgery in infancy, who have subsequently developed raised pulmonary vascular resistance. Surgical closure in these high-risk individuals

can result in substantial morbidity and mortality, because increases in pulmonary-arterial and right-ventricular pressure result in right-ventricular failure. A modified surgical technique with creation of a so-called valved patch, which allows unidirectional right-to-left shunting across a deliberate residual defect, can be undertaken with low operative mortality,85 although the benefits of this procedure over conventional closure have been questioned.86 In the past, banding of the pulmonary artery was done frequently as an interim palliative procedure to reduce pulmonary blood flow, particularly in infants. This technique is now undertaken rarely, except in individuals with either many ventricular septal defects or apical malformations, in whom surgical access to the anomalies is anticipated to be especially difficult or impossible. Application of either absorbable or balloon-dilatable bands has been performed in patients with multiple muscular defects, which potentially could avoid the need for removal of the band if anomalies reduce in size over time.87 The main surgical challenges are with defects that are near the apex and are difficult to access through either the

Unrepaired defect

Repaired defect

Survival

Excellent survival for small defects; large defects can be associated with pulmonary vascular disease; might develop aortic regurgitation

Excellent survival; occasional residual shunt; might develop right-ventricular or left-ventricular outflow-tract obstruction; might develop aortic regurgitation

Haemodynamic issues

Left-to-right shunt; left-ventricular dilation and impaired function; aortic regurgitation; pulmonary vascular disease

Residual shunt; ventricular function; aortic regurgitation; right-ventricular or left-ventricular outflow-tract obstruction

Arrhythmia and pacing

Rare

Rare atrioventricular block; rare ventricular block

Chest radiography

Cardiomegaly

Cardiomegaly

ECG

Rhythm and conduction; right-ventricular hypertrophy

Rhythm and conduction; right-ventricular hypertrophy

Echocardiography

Number and size of defects; left-ventricular and right-ventricular function; aortic regurgitation

Residual shunt; ventricular function; aortic regurgitation

Transoesophageal echocardiography

If transthoracic image is inadequate

If transthoracic image is inadequate

Cardiac catheterisation

Pulmonary vascular disease; associated lesions

Rarely needed

Holter test

Only if symptomatic

Only if symptomatic

Investigations

If symptomatic; sports counselling

If symptomatic; sports counselling

Indications for intervention

Exercise test

Left-to-right shunt with heart overload; reversible pulmonary hypertension; aortic regurgitation; associated abnormalities (right-ventricular outflow tract, subaortic stenosis); previous endocarditis

Left-to-right shunt with heart overload; reversible pulmonary hypertension; aortic regurgitation; associated abnormalities (right-ventricular outflow tract, subaortic stenosis); previous endocarditis

Endocarditis prophylaxis

Not needed

If residual defect

Pregnancy

No contraindications with uncomplicated defects; contraindicated with pulmonary vascular disease

No contraindications with closed defects; contraindicated with pulmonary vascular disease

Physical exercise

No restriction for small defects

No restriction for closed defects

Follow-up care Aortic regurgitation

Exclusively specialist centre

Exclusively specialist centre

Complicated haemodynamics

Exclusively specialist centre

Exclusively specialist centre

Pulmonary vascular disease

Shared care between specialist centre and general adult centre

Shared care between specialist centre and general adult centre

Small defect

Managed in non-specialist centre with access to specialist centre if needed

Managed in non-specialist centre with access to specialist centre if needed

Modified from reference 92 with permission of Oxford University Press.

Table: Issues related to ongoing surveillance of patients with ventricular septal defects

www.thelancet.com Vol 377 March 26, 2011

1109

Seminar

atrioventricular or semilunar valves. In some situations, malformations can be accessed through a ventriculotomy, although transcatheter approaches have been used.

Transcatheter closure Over the past decade or so, transcatheter techniques for closure of ventricular septal defects have been developed. These methods have been especially useful for muscular defects, which can be the most difficult to access surgically (figure 4).88 Much interest has been generated in development of transcatheter approaches to close perimembranous defects (figure 5). At present, this technique is not undertaken in most units because of the unacceptably high rate of post-procedure heart block associated with currently available devices. Of particular concern is that this risk does not subside or fall with time, with late-onset heart block being fairly prevalent.89,90 As softer devices are developed, this method could potentially be reintroduced in the future.

Hybrid techniques In infants with muscular ventricular septal defects, in whom both transcatheter and standard surgical approaches are difficult, a hybrid technique for closure has been implemented, which brings together surgery and interventional methods. With this method, a sternotomy is done in the standard way and the device is placed in the right ventricle through its anterior wall, under transoesophageal and fluoroscopic guidance.91

Issues in adults with ventricular septal defects Exercise

Adults with small ventricular septal defects, normal pulmonary arterial pressure, normal ventricular function, and no associated lesions should have a normal tolerance for exercise and, therefore, no exercise restrictions should be imposed (table).92 Those with pulmonary arterial hypertension usually self-restrict their amount of exercise. In a large study of adults with congenital heart disease who underwent formal testing, patients with Eisenmenger’s syndrome achieved the lowest levels of peak oxygen consumption during exercise and, furthermore, the extent of the reduction in peak oxygen consumption was an important marker of prognosis.93

Pregnancy Women with small ventricular septal defects without pulmonary hypertension do not seem to be at increased cardiovascular risk during pregnancy (table). Those with moderate defects could have raised pulmonary blood flow during pregnancy, an indication of increased circulating volume, although this effect could—to a degree—be counterbalanced by reduction in systemic vascular resistance. By contrast, pregnancy in women with Eisenmenger’s syndrome is associated with a very high risk of maternal and fetal death and premature delivery.94 1110

In a study of 17 ongoing pregnancies in ten women with Eisenmenger’s syndrome,95 one maternal death was reported and another woman deteriorated greatly, requiring high-level intensive care. Four spontaneous abortions and one stillbirth happened. Of the 12 deliveries of live infants, ten were premature.95 Against this background, women with Eisenmenger’s syndrome should be counselled strongly against pregnancy and should be referred to a specialist for contraception advice.96 Although sterilisation might be deemed appropriate for some women, it should only be done at a specialist centre with careful periprocedural care of these high-risk patients. Early termination is often recommended for women with Eisenmenger’s syndrome who become pregnant. For those who choose to continue pregnancy, obstetric care should be undertaken at a specialist centre, with access to intensive care. Even after successful delivery, maternal risk continues beyond the time of birth; therefore, close monitoring must be maintained in the postpartum period. Data suggest that augmentation of endothelialbased treatments—eg, combination of intravenous epoprostenol and oral sildenafil—might improve outcome for pregnant women.97 Contributors Both authors wrote the report and approved the final version. Conflicts of interest We declare that we have no conflicts of interest. References 1 Hoffman JI. Incidence of congenital heart disease: I—postnatal incidence. Pediatr Cardiol 1995; 16: 103–13. 2 Roguin N, Du ZD, Barak M, Nasser N, Hershkowitz S, Milgram E. High prevalence of muscular ventricular septal defect in neonates. J Am Coll Cardiol 1995; 26: 1545–48. 3 Hoffman JIE, Kaplan S. The incidence of congenital heart disease. J Am Coll Cardiol 2002; 39: 1890–900. 4 Hoffman JIE, Kaplan S, Liberthson RR. Prevalence of congenital heart disease. Am Heart J 2004; 147: 425–39. 5 Contreras-Ramos A, Sánchez-Gómez C, García-Romero HL, Cimarosti LO. Normal development of the muscular region of the interventricular septum: I—significance of the ventricular trabeculations. Anat Histol Embryol 2008; 37: 344–51. 6 Goor AD, Edwards EJ, Lillehei W. The development of the interventricular septum of the human heart: correlative morphogenetic study. Chest 1970; 58: 453–67. 7 De La Cruz MV, Moreno-Rodriguez R. Embryological development of the apical trabeculated region of both ventricles: the contribution of the primitive interventricular septum in ventricular septation. In: De La Cruz MV, Markwald R, eds. Living morphogenesis of the heart. Basel: Birkhäuser, 1998: pp 120–30. 8 Pierpont ME, Basson CT, Benson DW, et al. Genetic basis for congenital heart defects: current knowledge—a scientific statement from the American Heart Association Congenital Cardiac Defects Committee, Council on Cardiovascular Disease in the Young, endorsed by the American Academy of Pediatrics. Circulation 2007; 115: 3015–38. 9 Nora JJ. Multifactorial inheritance hypothesis for the etiology of congenital heart diseases: the genetic-environmental interaction. Circulation 1968; 38: 604–17. 10 Bruneau BG. The developmental genetics of congenital heart disease. Nature 2008; 451: 943–48. 11 Srivastava D, Olson EN. A genetic blueprint for cardiac development. Nature 2000; 407: 221–26. 12 Benson DW. Genetic origins of pediatric heart disease. Pediatr Cardiol 2010; 31: 422–29.

www.thelancet.com Vol 377 March 26, 2011

Seminar

13 14 15 16 17 18 19 20 21

22 23 24 25 26 27 28

29 30 31 32 33 34 35

36 37

Maitra M, Schluterman MK, Nichols HA, et al. Interaction of Gata4 and Gata6 with Tbx5 is critical for normal cardiac development. Dev Biol 2009; 326: 368–77. Basson CT, Bachinsky DR, Lin RC, et al. Mutations in human TBX5 [corrected] cause limb and cardiac malformation in Holt-Oram syndrome. Nat Genet 1997; 15: 30–35. Li QY, Newbury-Ecob RA, Terrett JA, et al. Holt-Oram syndrome is caused by mutations in TBX5, a member of the Brachyury (T) gene family. Nat Genet 1997; 15: 21–29. Liu C-X, Shen A-D, Li X-F, et al. Association of TBX5 gene polymorphism with ventricular septal defect in the Chinese Han population. Chin Med J 2009; 122: 30–34. Zhang W-M, Li X-F, Ma Z-Y, et al. GATA4 and NKX2.5 gene analysis in Chinese Uygur patients with congenital heart disease. Chin Med J 2009; 122: 416–19. Zhang W, Li X, Shen A, Jiao W, Guan X, Li Z. GATA4 mutations in 486 Chinese patients with congenital heart disease. Eur J Med Genet 2008; 51: 527–35. Rajagopal SK, Ma Q, Obler D, et al. Spectrum of heart disease associated with murine and human GATA4 mutation. J Mol Cell Cardiol 2007; 43: 677–85. Garg V, Kathiriya IS, Barnes R, et al. GATA4 mutations cause human congenital heart defects and reveal an interaction with TBX5. Nature 2003; 424: 443–47. Jenkins KJ, Correa A, Feinstein JA, et al. Noninherited risk factors and congenital cardiovascular defects: current knowledge: a scientific statement from the American Heart Association Council on Cardiovascular Disease in the Young: endorsed by the American Academy of Pediatrics. Circulation 2007; 115: 2995–3014. Srivastava D. Genetic regulation of cardiogenesis and congenital heart disease. Ann Rev Pathol 2006; 1: 199–213. Bajolle F, Zaffran S, Bonnet D. Genetics and embryological mechanisms of congenital heart diseases. Arch Cardiovasc Dis 2009; 102: 59–63. Soto B, Becker AE, Moulaert AJ, Lie JT, Anderson RH. Classification of ventricular septal defects. Br Heart J 1980; 43: 332–43. Jacobs JP, Burke RP, Quintessenza JA, Mavroudis C. Congenital Heart Surgery Nomenclature and Database Project: ventricular septal defect. Ann Thorac Surg 2000; 69 (4 Suppl): S25–35. Soto B, Ceballos R, Kirklin JW. Ventricular septal defects: a surgical viewpoint. J Am Coll Cardiol 1989; 14: 1291–97. Van Praagh R, Geva T, Kreutzer J. Ventricular septal defects: how shall we describe, name and classify them? J Am Coll Cardiol 1989; 14: 1298–99. Benson LN, Yoo D-J, Al Habshan F, Anderson RH. Ventricular septal defects. In: Anderson RH, Baker EJ, Redington A, Rigby ML, Penny D, Wernovsky G, eds. Paediatric cardiology, 3rd edn. Philadelphia: Churchill Livingstone, 2009: pp 591–624. Rudolph AM. Circulatory adjustments after birth: effects on ventricular septal defect. Br Heart J 1971; 33 (suppl): 32–34. Duffels MGL, Engelfriet PM, Berger RMF, et al. Pulmonary arterial hypertension in congenital heart disease: an epidemiologic perspective from a Dutch registry. Int J Cardiol 2007; 120: 198–24. Van de Bruaene A, Delcroix M, Pasquet A, et al. The Belgian Eisenmenger syndrome registry: implications for treatment strategies. Acta Cardiol 2009; 64: 447–53. Diller GP, Gatzoulis MA. Pulmonary vascular disease in adults with congenital heart disease. Circulation 2007; 115: 1039–50. Beghetti M, Galiè N. Eisenmenger syndrome: a clinical perspective in a new therapeutic era of pulmonary arterial hypertension. J Am Coll Cardiol 2009; 53: 733–40. Rabinovitch M. Pulmonary hypertension: pathophysiology as a basis for clinical decision making. J Heart Lung Transplant 1999; 11: 1041–53. Oishi PE, Wiseman DA, Sharma S, et al. Progressive dysfunction of nitric oxide synthase in a lamb model of chronically increased pulmonary blood flow: a role for oxidative stress. Am J Physiol Lung Cell Mol Physiol 2008; 295: L756–66. Fasules JW, Tryka F, Chipman CW, Van Devanter SH. Pulmonary hypertension and arterial changes in calves with a systemic-to-left pulmonary artery connection. J Appl Physiol 1994; 77: 867–75. Diller GP, van Eijl S, Okonko DO, et al. Circulating endothelial progenitor cells in patients with Eisenmenger syndrome and idiopathic pulmonary arterial hypertension. Circulation 2008; 117: 3020–30.

www.thelancet.com Vol 377 March 26, 2011

38

39

40

41 42 43

44 45

46 47 48

49

50

51

52 53

54

55

56 57

Tweddell JS, Pelech AN, Frommelt PC. Ventricular septal defect and aortic valve regurgitation: pathophysiology and indications for surgery. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu 2006: 9: 147–52. Telagh R, Alexi-Meskishivili V, Hetzer R, Lange PE, Berger F, Abdul-Khaliq H. Initial clinical manifestations and mid- and long-term results after surgical repair of double-chambered right ventricle in children and adults. Cardiol Young 2008; 18: 268–74. Vogel M, Smallhorn JF, Freedom RM, Coles J, Williams WG, Trusler GA. An echocardiographic study of the association of ventricular septal defect and right ventricular muscle bundles with a fixed subaortic abnormality. Am J Cardiol 1988; 61: 857–60. Cheatham JP, Latson LA, Gutgesell HP. Ventricular septal defect in infancy: detection with two dimensional echocardiography. Am J Cardiol 1981; 47: 85–89. Minette MS, Sahn DJ. Ventricular septal defects. Circulation 2006; 114: 2190–97. Sutherland GR, Godman MJ, Smallhorn JF, Guiterras P, Anderson RH, Hunter S. Ventricular septal defects: two dimensional echocardiographic and morphological correlations. Br Heart J 1982; 47: 316–28. Ludomirsky A, Huhta JC, Vick GW, Murphy DJ, Danford DA, Morrow WR. Color Doppler detection of multiple ventricular septal defects. Circulation 1986; 74: 1317–22. Ortiz E, Robinson PJ, Deanfield JE, Franklin R, Macartney FJ, Wyse RK. Localisation of ventricular septal defects by simultaneous display of superimposed colour Doppler and cross sectional echocardiographic images. Br Heart J 1985; 54: 53–60. Sommer RJ, Golinko RJ, Ritter SB. Intracardiac shunting in children with ventricular septal defect: evaluation with Doppler color flow mapping. J Am Coll Cardiol 1990; 16: 1437–44. Houston AB, Lim MK, Doig WB, Reid JM, Coleman EN. Doppler assessment of the interventricular pressure drop in patients with ventricular septal defects. Br Heart J 1988; 60: 50–56. Ayres NA, Miller-Hance W, Fyfe DA, et al. Indications and guidelines for performance of transesophageal echocardiography in the patient with pediatric acquired or congenital heart disease: report from the task force of the Pediatric Council of the American Society of Echocardiography. J Am Soc Echocardiogr 2005; 18: 91–98. Bezold LI, Pignatelli R, Altman CA, et al. Intraoperative transesophageal echocardiography in congenital heart surgery: the Texas Children’s Hospital experience. Texas Heart Inst J 1996; 23: 108–15. Chen FL, Hsiung MC, Nanda N, Hsieh KS, Chou MC. Real time three-dimensional echocardiography in assessing ventricular septal defects: an echocardiographic-surgical correlative study. Echocardiography 2006; 23: 562–68. Kilner PJ, Geva T, Maemmerer H, Trindade PT, Schwitter J, Webb GD. Recommendations for cardiovascular magnetic resonance in adults with congenital heart disease from the respective working groups of the European Society of Cardiology. Eur Heart J 2010; 31: 794–805. Sato Y, Matsumoto N, Matsuo S, et al. Double chambered right ventricle: depiction at three-dimensional whole heart magnetic resonance imaging. Int J Cardiol 2007; 119: e14–16. Beerbaum P, Körperich H, Barth P, Esdorn H, Gieseke J, Meyer H. Noninvasive quantification of left-to-right shunt in pediatric patients: phase-contrast cine magnetic resonance imaging compared with invasive oximetry. Circulation 2001; 103: 2476–82. Körperich H, Gieseke J, Barth P, et al. Flow volume and shunt quantification in pediatric congenital heart disease by real-time magnetic resonance velocity mapping: a validation study. Circulation 2004; 109: 1987–93. Muthurangu V, Taylor A, Andriantsimiavona R, et al. Novel method of quantifying pulmonary vascular resistance by use of simultaneous invasive pressure monitoring and phase-contrast magnetic resonance flow. Circulation 2004; 110: 826–34. Vick GW. Recent advances in pediatric cardiovascular MRI. Curr Opin Pediatr 2003; 15: 454–62. Axt-Fliedner R, Schwarze A, Smrcek J, Germer U, Krapp M, Gembruch U. Isolated ventricular septal defects detected by color Doppler imaging: evolution during fetal and first year of postnatal life. Ultrasound Obstet Gynecol 2006; 27: 266–73.

1111

Seminar

58 59

60

61 62

63 64

65 66 67 68 69

70

71

72 73 74

75

76

77

78

1112

Gantenbein MH, Bauersfeld U, Baenziger O, et al. Side effects of angiotensin converting enzyme inhibitor (captopril) in newborns and young infants. J Perinat Med 2008; 46: 448–52. Namachivayam P, Theilen U, Butt WW, Cooper SM, Penny DJ, Shekerdemian LS. Sildenafil prevents rebound pulmonary hypertension after withdrawal of nitric oxide in children. Am J Respir Crit Care Med 2006; 174: 1042–47. Bizzarro M, Gross I. Inhaled nitric oxide for the postoperative management of pulmonary hypertension in infants and children with congenital heart disease. Cochrane Database Syst Rev 2005: 4: CD005055. Kleinman CS, Tabibian M, Starc TJ, Hsu DT, Gersony WM. Spontaneous regression of left ventricular dilation in children with restrictive ventricular septal defects. J Pediatr 2007; 150: 583–86. Gabriel HM, Heger M, Innerhofer P, et al. Long-term outcome of patients with ventricular septal defect considered not to require surgical closure during childhood. J Am Coll Cardiol 2002; 39: 1066–71. Turner SW, Hornung T, Hunter S. Closure of ventricular septal defects: a study of factors influencing spontaneous and surgical closure. Cardiol Young 2002; 12: 357–63. Eroğlu AG, Oztunç F, Saltik L, Bakari S, Dedeoğlu S, Ahunbay G. Evolution of ventricular septal defect with special reference to spontaneous closure rate, subaortic ridge and aortic valve prolapse. Pediatr Cardiol 2003; 24: 31–35. Kostolny M, Schreiber C, von Arnim V, Vogt M, Wottke M, Lange R. Timing of repair in ventricular septal defect with aortic insufficiency. Thorac Cardiovasc Surg 2006; 54: 512–15. Kumar RK, Sandoval J. Advanced pulmonary vascular disease: the Eisenmenger syndrome. Cardiol Young 2009; 19: 622–26. Spence MS, Balaratnam MS, Gatzoulis MA. Clinical update: cyanotic adult congenital heart disease. Lancet 2007; 370: 1530–32. Amnesh N, Warnes CA. Cerebrovacular events in adult patients with cyanotic congenital heart disease. J Am Coll Cardiol 1996; 28: 768–72. Galiè N, Manes A, Palazzini M, et al. Management of pulmonary arterial hypertension associated with congenital systemic-to-pumonary shunts and Eisenmenger’s syndrome. Drugs 2008; 68: 1049–66. Fine N, Dias B, Shoemaker G, Mehta S. Endothelin receptor antagonist therapy in congenital heart disease with shunt-associated pulmonary arterial hypertension: a qualitative systematic review. Can J Cardiol 2009; 25: e63–68. Mehta PK, Simpson L, Lee EK, Lyle TA, McConnell ME, Book WM. Endothelin receptor antagonists improve exercise tolerance and oxygen saturations in patients with Eisenmenger syndrome and congenital heart defects. Tex Heart Inst J 2008; 35: 256–61. Wort SJ. Sildenafil in Eisenmenger syndrome: safety first. Int J Cardiol 2007; 120: 314–16. Galiè N, Beghetti M, Gatzoulis MA, et al. Bosentan therapy in patients with Eisenmenger syndrome: a multicenter, double-blind, randomized, placebo-controlled study. Circulation 2006; 114: 48–54. Gatzoulis MA, Beghetti M, Galiè N, et al. Longer-term bosentan therapy improves functional capacity in Eisenmenger syndrome: results of the BREATHE-5 open-label extension study. Int J Cardiol 2008; 127: 27–32. van Loon RLE, Hoendermis ES, Duffels MGJ, et al. Long-term effect of bosentan in adults versus children with pulmonary arterial hypertension associated with systemic-to-pulmonary shunt: does the beneficial effect persist? Am Heart J 2007; 154: 776–82. Tay ELW, Papaphylactou M, Diller GP, et al. Quality of life and functional capacity can be improved in patients with Eisenmenger syndrome with oral sildenafil therapy. Int J Cardiol 2010; DOI:10.1016/j.ijcard.2010.02.020. Dimopoulos K, Inuzuka R, Goletto S, et al. Improved survival among patients with Eisenmenger syndrome receiving advanced therapy for pulmonary arterial hypertension. Circulation 2010; 121: 20–25. Limsuwan A, Khosithseth A, Wanichkul S, Khowsathit P. Aerosolized iloprost for pulmonary vasoreactivity testing in children with long-standing pulmonary hypertension related to congenital heart disease. Catheter Cardiovasc Interv 2009; 73: 98–104.

79

80

81 82

83

84 85

86

87

88

89 90 91

92

93 94 95

96 97

Dimopoulos K, Peset A, Gatzoulis MA. Evaluating operability in adults with congenital heart disease and the role of pretreatment with targeted pulmonary arterial hypertension therapy. Int J Cardiol 2008; 129: 163–71. Wilson W, Taubert KA, Gewitz M, et al. Prevention of infective endocarditis: guidelines from the American Heart Association: a guideline from the American Heart Association Rheumatic Fever, Endocarditis, and Kawasaki Disease Committee, Council on Cardiovascular Disease in the Young, and the Council on Clinical Cardiology, Council on Cardiovascular Surgery and Anesthesia, and the Quality of Care and Outcomes Research Interdisciplinary Working Group. Circulation 2007; 116: 1736–54. Scully BB, Morales DLS, Zafar F, McKenzie ED, Fraser CD, Heinle JS. Current expectations for surgical repair of isolated ventricular septal defects. Ann Thorac Surg 2010; 89: 544–49. Bol Raap G, Meijboom FJ, Kappetein AP, Galema TW, Yap S-C, Bogers AJJC. Long-term follow-up and quality of life after closure of ventricular septal defect in adults. Eur J Cardiothorac Surg 2007; 32: 215–19. Hövels-Gürich HH, Konrad K, Skorzenski D, et al. Long-term neurodevelopmental outcome and exercise capacity after corrective surgery for tetralogy of Fallot or ventricular septal defect in infancy. Ann Thorac Surg 2006; 81: 958–66. Jonas RA. Congenital heart surgery in developing countries. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu 2008; 11: 3–6. Novick WM, Sandoval N, Lazorhysynets VV, et al. Flap valve double patch closure of ventricular septal defects in children with increased pulmonary vascular resistance. Ann Thorac Surg 2005; 79: 21–28. Gan H-L, Zhang J-Q, Zhang Z-G, Luo Y, Zhou Q-W, Bo P. The unidirectional valve patch provides no benefits to early and long-term survival in patients with ventricular septal defect and severe pulmonary artery hypertension. J Thorac Cardiovasc Surg 2010; 139: 950–55. Brown S, Boshoff D, Rega F, Eyskens B, Meyns B, Gewillig M. Dilatable pulmonary artery banding in infants with low birth weight or complex congenital heart disease allows avoidance or postponement of subsequent surgery. Eur J Cardiothorac Surg 2010; 37: 296–301. Lim DS, Forbes TJ, Rothman A, Lock JE, Landzberg MJ. Transcatheter closure of high-risk muscular ventricular septal defects with the CardioSEAL occluder: initial report from the CardioSEAL VSD registry. Catheter Cardiovasc Interv 2007; 70: 740–44. Dumitrescu A, Lane GK, Wilkinson JL, Goh TH, Penny DJ, Davis AM. Trancatheter closure of perimembranous ventricular septal defect. Heart 2007; 93: 867. Collins NJ, Benson L, Horlick E. Late complete heart block in an adult patient undergoing percutaneous ventricular septal defect closure. J Invasive Cardiol 2008; 20: E200–03. Crossland DS, Wilkinson JL, Cochrane AD, D’udekem Y, Brizard CP, Lane GK. Initial results of primary device closure of large muscular ventricular septal defects in early infancy using perventricular access. Cathet Cardiovasc Interv 2008; 72: 386–91. Deanfield J, Thaulow E, Warnes C, et al, for the Task Force on the Management of Grown Up Congential Heart Disease of the European Society of Cardiology. Management of grown up congenital heart disease. Eur Heart J 2003; 24: 1035–84. Diller G-P, Dimopoulos K, Okonko D et al. Exercise intolerance in adult congenital heart disease: comparative severity, correlates, and prognostic implication. Circulation 2005; 112: 828–35. Bédard E, Dimopoulos K, Gatzoulis MA. Has there been any progress made on pregnancy outcomes among women with pulmonary arterial hypertension? Eur Heart J 2009; 30: 256–65. Dranenkiene A, Opitz CF, Gumbiené L, Doma B, Ewert R, Bühlmeyer K. Pregnancy in patients with Eisenmenger’s syndrome: experiences from Vilnius 1967–2003. Dtsch Med Wochenschr 2004; 129 (suppl 1): S35–39 [in German]. Silversides CK, Sermer M, Siu SC. Choosing the best contraceptive method for the adult with congenital heart disease. Curr Cardiol Rep 2009; 11: 298–305. Goland S, Tsai F, Habib M, Janmohamed M, Goodwin TM, Elkayam U. Favorable outcome of pregnancy with an elective use of epoprostenol and sildenafil in women with severe pulmonary hypertension. Cardiology 2010; 115: 205–08.

www.thelancet.com Vol 377 March 26, 2011

Health Policy

Medical schools in sub-Saharan Africa Fitzhugh Mullan, Seble Frehywot, Francis Omaswa, Eric Buch, Candice Chen, S Ryan Greysen, Travis Wassermann, Diaa ElDin ElGaili Abubakr, Magda Awases, Charles Boelen, Mohenou Jean-Marie Isidore Diomande, Delanyo Dovlo, Josefo Ferro, Abraham Haileamlak, Jehu Iputo, Marian Jacobs, Abdel Karim Koumaré, Mwapatsa Mipando, Gottleib Lobe Monekosso, Emiola Oluwabunmi Olapade-Olaopa, Paschalis Rugarabamu, Nelson K Sewankambo, Heather Ross, Huda Ayas, Selam Bedada Chale, Soeurette Cyprien, Jordan Cohen, Tenagne Haile-Mariam, Ellen Hamburger, Laura Jolley, Joseph C Kolars, Gilbert Kombe*, Andre-Jacques Neusy

Small numbers of graduates from few medical schools, and emigration of graduates to other countries, contribute to low physician presence in sub-Saharan Africa. The Sub-Saharan African Medical School Study examined the challenges, innovations, and emerging trends in medical education in the region. We identified 168 medical schools; of the 146 surveyed, 105 (72%) responded. Findings from the study showed that countries are prioritising medical education scale-up as part of health-system strengthening, and we identified many innovations in premedical preparation, teambased education, and creative use of scarce research support. The study also drew attention to ubiquitous faculty shortages in basic and clinical sciences, weak physical infrastructure, and little use of external accreditation. Patterns recorded include the growth of private medical schools, community-based education, and international partnerships, and the benefit of research for faculty development. Ten recommendations provide guidance for efforts to strengthen medical education in sub-Saharan Africa.

Introduction Health in Africa is important as an issue of human equity and as a precursor to poverty reduction and human development. Africa has 24% of the world’s burden of disease, but only 3% of the world’s health workforce.1 The Joint Learning Initiative2 and the 2006 World Health Report1 called attention to the particularly severe shortages of human resources for health in Africa. Early responses to the recognition of these shortages included calls for increased production of community health workers3 and non-physician clinicians,4 and task shifting to make effective use of available cadres.5 Attention has now focused on education and retention of medical doctors in Africa, not because doctors will solve the vast unmet health needs of the continent, but in the belief that no health system can function well without an adequate number of doctors to participate in clinical and public health work, management, education, and policy making.6 Sub-Saharan Africa has an estimated 145 000 physicians7 (5% of the 2 877 000 practising physicians in Europe) to serve a population of 821 million (more than the population in Europe).8 Overall, sub-Saharan Africa has a physician-to-population ratio of 18 per 100 000, compared with countries such as India (60 per 100 000), Brazil (170 per 100 000), and France (370 per 100 000).8 Africa’s poorest countries have even greater physician shortages. The very low physician-to-population ratios in countries in sub-Saharan Africa result from several factors, including a modest output of students by a small number of medical schools, and emigration of many graduates to other countries or continents. (The term medical school refers to medical schools and colleges of medicine.) Any effort intended to improve health-system functioning in these countries should consider options to increase both the productivity of medical schools and the retention of their graduates within their countries. National and international interest with respect to strategic investment in medical education in www.thelancet.com Vol 377 March 26, 2011

sub-Saharan Africa has been growing, but little is known about the status of medical schools or trends within medical education across the continent. For example, when we initially reviewed all available medical school databases (WHO, Institute for International Medical Education, Foundation for Advancement of International Medical Education and Research, and World Federation for Medical Education) in 2008, we identified 103 schools; however, our survey work identified 168 schools operating in sub-Saharan Africa. This absence of pan-African data and perspective is a major challenge for African governments and donor organisations seeking to address shortages in physician workforce. The Sub-Saharan African Medical School Study addressed this knowledge gap by developing an information base for the status of, trends in, and prospects for African medical education for educators, policy makers, and international organisations. Figure 1 outlines the structure, participants, and sequence of activities that comprised this study.

Findings

General results Of 168 medical schools, 146 were identified before the close of the study survey period in December, 2009. 105 (72%) schools responded. (All identified schools are shown in webappendix pp 7–10.) Countries with large populations (p<0·0001) and land masses (p<0·0001) had more medical schools than did smaller countries by multiple linear regression, but the recorded no significant correlation between a country’s gross domestic product (GDP) per head, region of Africa, or national language and its number of medical schools. Survey respondents and site visited institutions represented all regions of sub-Saharan Africa and all major language groups. The data collected in the study site visits and the survey are summarised in 13 categories.

Lancet 2011; 377: 1113–21 Published Online November 11, 2010 DOI:10.1016/S01406736(10)61961-7 This paper has been corrected. The corrected version first appeared at thelancet.com on November 19, 2010 See Editorial page 1047 See Online/Comment DOI:10.1016/S01406736(10)61920-4 The George Washington University, Washington, DC, USA (Prof F Mullan MD, S Frehywot MD, C Chen MD, T Wassermann MPH, H Ross MPH, H Ayas EdD, S B Chale BA, S Cyprien BA, Prof J Cohen MD, T Haile-Mariam MD, E Hamburger MD, L Jolley MPH); University of Pretoria, Pretoria, South Africa (Prof E Buch FFCH(CM)(SA)); Yale University, New Haven, CT, USA (S R Greysen MD); University of Gezira, Wad Madani, Sudan (Prof D E E Abubakr MD); University of Cocody, Abidjan, Côte d’Ivoire (Prof M J-M I Diomande MD); World Health Organization, Geneva, Switzerland (D Dovlo MWACP); Catholic University of Mozambique, Beira, Mozambique (Prof J Ferro MD); Jimma University, Jimma, Ethiopia (Prof A Haileamlak MD); Walter Sisulu University, Mthatha, South Africa (Prof J Iputo MBChB); University of Cape Town, Cape Town, South Africa (Prof M Jacobs MBChB); University of Mali, Bamako, Mali (Prof A K Koumaré MD); University of Malawi, Blantyre, Malawi (Prof M Mipando PhD); Global Health Dialogue Foundation, Yaounde, Cameroon (Prof G L Monekosso MD); University of Ibadan, Ibadan, Nigeria (E O Olapade-Olaopa MD); Hubert Kairuki Memorial

1113

Health Policy

University, Dar Es Salaam, Tanzania (Prof P Rugarabamu DDS); Makerere University, Kampala, Uganda (Prof N K Sewankambo MBChB); International Consultant in Health Systems and Personnel, Sciez-sur-Léman, France (C Boelen MD); African Center for Global Health and Social Transformation, Kampala, Uganda (F Omaswa FRCS); World Health Organization, Harare, Zimbabwe (M Awases PhD); University of Michigan Medical School, Ann Arbor, MI, USA (J C Kolars MD); Abt Associates, Bethesda, MD, USA (G Kombe MD); New York University School of Medicine, New York, NY, USA (A-J Neusy MD)

(1) Scaling up of medical education Many countries are scaling up medical education as part of health-sector strengthening. Several national governments are investing greatly in human resources for health, producing health-sector strategic plans that include increases in health-care workforce. Medical education is essential to the development of the health-care workforce and is an integral part of human resource plans. Seven medical schools that responded were founded before 1960, and another 29 during the independence decades

SAMSS Secretariat • 16 US-based medical school faculty and research sta (webappendix p 1)

Preparatory phase

*Dr Kombe died in November, 2009 Correspondence to: Prof Fitzhugh Mullan, George Washington University, 2121 K Street, NW Suite 210, Washington, DC 20037, USA [email protected]

(1960–79). Little growth occurred during the 1980s, but 58 responding schools have opened since 1990 (figure 2). Many medical schools are expanding enrolment of students. 59 of 78 (76%) schools that responded reported increases in the number of students in their first-year classes compared with 5 years ago. 56 of 105 (53%) reported plans to increase in the next 5 years, with 57 of 96 (59%) mandated to increase enrolment, generally from ministries of health or education. The present total enrolment of first-year students in 96 responding schools

See Online for webappendix

Literature review* • Databases: Medline, CINAHL, ERIC, Global Health, Embase, African Indicus Medicus, African Journals Online, and Biomed • Search terms: “Africa” and ”medical education” or “medical students” or “medical schools” • 642 abstracts reviewed (webappendix pp 2–3)

Key informant interviews • Semi-structured questionnaire gathered information about status and trends in SSA medical education, and country-level and school-level specifics • Key informants provided additional names that yielded further interview participants • 50 key informant interviews were undertaken (webappendix p 4)

Selection of site visit schools • 10 schools from di erent regions, linguistic groups, ages, ownership, and educational models (webappendix p 5)

Selection of SAMSS Advisory Committee • 16 members, including one from each site visit school plus six experts in SSA medical education • Membership included representatives from 13 African nations (webappendix p 6)

Analytical phase

Primary data collection phase

Partnering institution The University of Pretoria (UP)

For more on the Sub-Saharan African Medical School Study see http://SAMSS.org

1114

SAMSS site visits • Teams: two visitors from Secretariat and two from Advisory Committee • Semi-structured questionnaire guided visits with medical school dean, key faculty sta , student representatives, regulatory bodies, ministries • Visits made to hospital and community teaching sites • Themes: innovation, capacity building, retention • Topics included: school mission, social context of the school, curricular content, attitudes towards emigration and retention, programmes for underserved areas, admissions policies, institutional relationships, public health and leadership curricula, and PGME • 10 site visit reports produced†

SAMSS survey • Designed by Secetariat, partnering institution (UP-based team), and Advisory Committee; piloted at Advisory Committee schools; implemented by UP with assistance from Advisory Committee; analysed by UP and Secretariat • Sent to deans of all identified SSA medical schools, US$150 incentive for completion, aggressive follow-up for non-response • Questions about school demographics, students, graduates, faculty, curriculum, finances, infrastructure, relationships, and barriers • 72% response rate from 146 schools identified by December, 2009

SAMSS findings • The Advisory Committee and Secretariat drafted 13 overarching findings derived from the SAMSS site visits and SAMSS survey • The findings address context, challenges, and innovations in medical education in SSA

SAMSS recommendations • 10 recommendations agreed upon by Advisory Committee and Secretariat provide actionable steps for medical schools, donors, and governments • SAMSS recommendations form the basis of the conclusion of this report

Figure 1: Structure and methods of the Sub-Saharan Africa Medical School Study SAMSS=Sub-Saharan African Medical School Study. CINAHL=Cumulative Index to Nursing and Allied Health Literature. ERIC=Education Resources Information Center. SSA=sub-Saharan Africa. PGME=postgraduate medical education. *References available from authors on request. †Site visit reports available from the Sub-Saharan African Medical School Study website.

www.thelancet.com Vol 377 March 26, 2011

Health Policy

(2) Effect of the country’s health system When civil society is in disarray and governance is compromised, medical education and retention of physicians will be affected. Graduates in many countries decline to work in rural areas because of lack of clinical www.thelancet.com Vol 377 March 26, 2011

40 Number of medical schools (N=100)

is 18 349. The number of graduates in responding schools was 7861 in 2008 (figure 3). These graduates represent the output of the 105 responding medical schools. Many of the non-responding schools are private or new, and their characteristics imply fewer average graduates for the 63 non-responding schools than for the responding schools. These data suggest an estimated 10 000–11 000 graduates per year from medical schools in sub-Saharan Africa. Differences between enrolment and graduation figures are mainly attributable to the opening or expansion of schools. A few universities admit large numbers of students before numbers are reduced in the second year. 59 of 84 (70%) responding schools reported that at least 80% of first-year students graduate. The Ethiopian Government is investing greatly in a workforce scale-up plan based on a so-called flood and retain strategy, which involves a rapid, massive increase in the number of trained health workers and attendant retention measures. The Ministry of Education mandated that all medical schools expand their class sizes. Thus, Jimma University’s (Ethiopia) first-year enrolment for 2009 increased from 200 to 250, and is expected to reach 350 for 2011. The government supports this strategy by investing in physical infrastructure, including construction of a new teaching hospital at the university. The Hubert Kairuki Memorial University in Tanzania exemplifies private sector scale-up, expanding from an initial intake of 25 first-year medical students in 1998, to 70 per year in 2010. The Tanzanian Government has assisted by providing student loans and grants to private school students, enabling more students to afford tuition fees. For all medical schools in sub-Saharan Africa, including private schools, fees vary widely. Nine respondent schools (9%) offer free tuition, 47 (47%) charge US$1000 or less per year, and nine (9%) charge more than $5000. Private schools derive most of their income from tuition; public schools receive most of their operating budgets from the government (figure 4). Respondents were asked to identify the three greatest needs for scaling up the quality and quantity of their graduates in an open-ended question. Webappendix p 11 shows a summary of responses. Faculty-related issues were most commonly identified as key to improving the quality of graduates (35 of 94 respondents rated this issue as most important). Infrastructure issues were most frequently regarded as essential to improve the quantity of graduates (37 of 94 respondents rated this issue as most important). Curricular issues were viewed as affecting quality, whereas improvements in clinical sites were regarded as helping with quantity. Budgetary issues were referred to in response to both questions.

Public Private, non-profit, other Private, non-profit, faith-based Private, for profit

33

30 25 20

16 13

10

6 3

0

1

0

2

1

1910–19 1920–29 1930–39 1940–49 1950–59 1960–69 1970–79 1980–89 1990–99 2000–09

Figure 2: Founding dates of medical schools in sub-Saharan Africa Five schools did not respond.

support. Graduates from Ibadan University , for example, forgo employment in Nigeria’s large and crucial network of secondary hospitals because of poor pay and working conditions, and shortages of supplies, support personnel, and equipment. Assessment of retention strategies has been challenging because of the poor ability by most health systems to track medical school graduates. 47 of 58 (81%) survey respondents whose schools have graduated doctors reported no established tracking systems. Figure 5 shows the location of medical school graduates 5 years after graduation as estimated by the 62 schools responding to this survey question. The percentage of graduates estimated to be in rural general practice 5 years after graduation was positively correlated with the existence of a compulsory service programme (p=0·039), a moderate number of postgraduate medical education programmes (p=0·016), and French as a language of instruction (p=0·016); these data were analysed with ANCOVA. We recorded no significant correlation with GDP, the existence of a targeted recruitment programme for rural students, percentage of national population in rural areas, or use of community-based education. Many schools and nations are working to address emigration. National service is required from graduates in Mozambique, South Africa, Ethiopia, and Nigeria to obtain some clinical service from all graduates, although enforcement of these requirements is variable.

(3) Shortages within medical school faculties Shortages within medical school faculties are endemic, problematic, and made worse by emigration of health-care workers. Almost every school visited in this study had some degree of faculty shortage in both basic and clinical sciences. The number of teaching staff (salaried full-time or part-time, and volunteer) at 51 of 98 responding schools is fewer than 100; about half have between 52 (25th percentile) and 147 (75th percentile) teaching staff. Small salaries, limited career options, heavy teaching loads, growing enrolment, and absence of equipment and support staff are the main barriers to retaining faculty staff. 1115

Health Policy

20

20 20

First-year students (N=96) Graduates (N=80)

16 15

Number of medical schools

13

13 12 11 10

10

10

9 8 7

7

5

4

4 3

3 2

2

2

2 1

0

0

1–24

25–49

50–74

75–99

100–149

150–199

200–249

250–299

300–399

400–499

500–999 1000–1499 1500–1999

Figure 3: Number of first-year students and graduates in medical schools in sub-Saharan Africa (2008) *N varies because some medical schools have not yet graduated doctors. Private Public

25

Proportion of income from each of the following sources

24

Number of medical schools (N=100)

20

15

15

11 10

Private schools

Public schools

9

Government Student tuition and fees Local and international donors Research grants Other

9 9

8 6

5

4 2

1 0

0 1– 50 0 50 1– 10 00 10 01 –2 00 0 20 01 –3 00 30 0 01 –4 00 0 40 01 –5 00 0 50 01 –6 00 0 60 01 –7 00 0 70 01 –8 00 0 80 01 –9 00 0 90 01 –1 00 00 ≥1 00 00

0

10

1– 10

0

0

2

Yearly tuition fee (US$)

Figure 4: Tuition costs and sources of income in medical schools in sub-Saharan Africa

Shortages put extra pressure on faculty staff and promote emigration or relocation to private organisations and opportunities within non-governmental organisations. Faculty staff who are well trained and accredited are prime candidates to be recruited outside the country, resulting in the loss of both clinicians and teachers. Academic salaries severely restrict recruitment and retention of faculty staff. In many universities, clinical staff are paid on the same scale as are other university professors, 1116

which is lower than that of public sector doctors set by ministries of health. Research opportunities are often scarce, and teaching responsibilities are great. At Gezira University, Sudan, the absence of basic scientists means clinicians frequently have to teach basic science to medical students. Many schools rely on expatriate faculty staff. The founding faculty members of the Walter Sisulu University in South Africa were from Uganda, Cuba, and Nigeria. Some schools have initiated creative strategies to retain faculty staff, such as the Hubert Kairuki Memorial University in Tanzania, where incentives such as housing and communications allowances, free telephone air time, and seminar participation are provided. The Catholic University in Mozambique has made a targeted effort to train and promote the medical faculty. Nowadays more than half their faculty staff are Mozambicans, although the university remains dependent on expatriates as well. At Walter Sisulu University, the shortage of clinical faculty staff is relieved largely by partnerships with clinicians at local hospitals who are employed by the provincial Department of Health but obliged by their contracts to participate in teaching. Loss of faculty staff at surveyed schools was substantial, with a median 10% of staff from 5 years ago no longer with the schools, and half of schools losing between 6% (25th percentile) and 18% (75th percentile) of teaching staff in 5 years. The greatest reason given for faculty loss was emigration (webappendix p 12). The percentage of vacant faculty positions was lowest in countries with a high GDP per head (p=0·0084), and highest in public medical schools (p=0·0099), by multiple linear regression. Most respondents (80 of 100) believed that doctor retention in their www.thelancet.com Vol 377 March 26, 2011

Health Policy

country is a difficulty; however, only 51 listed any universitylevel steps taken to address the problem, most commonly salary increases or bonuses (20 respondents), strengthening programmes for postgraduate education (13), and community-based education (9).

(4) Weaknesses in medical education infrastructure Deficiencies in medical education infrastructure are ubiquitous and restricting. At Jimma University, power, water, and telecommunications are unreliable, jeopardising training and innovation. At Ibadan University, informants expressed concern about daily power outages. Departments have to purchase generators for clinical and teaching functions. At Catholic University, challenges include an insufficient number of computers, restrictions in internet connectivity, and the absence of student hostels. Inadequate student housing near clinical sites is also a difficulty at Walter Sisulu University and Mali University. The experience of the College of Medicine in Malawi is a good example of the role of partnerships to improve infrastructure. Assisted by funds from Sweden, Norway, and the Global Fund to Fight AIDS, Tuberculosis and Malaria, the school has constructed and improved lecture halls, libraries, hostels, computer facilities, offices, and recreational areas. These improvements accommodate larger class sizes and a growing faculty. The study survey considered both the quality and quantity of specific physical and communications resources. Multiple linear regression explored relations between six summary resource scores (scores for buildings, libraries, laboratories, clinical sites, internet, and advanced information communication technology [ICT]) and various national and institutional factors (webappendix pp 13–14). High GDP was associated with high scores for five of the six resources, older schools had high scores for four, and public schools rated their resources as worse in three indicators. Schools charging higher tuition fees reported more advanced ICT resources than did those charging lower fees.

(5) Coordination between ministries of education and health Insufficient coordination between ministries of education and health can be a barrier to medical schools’ ability to increase the capacity of the health workforce. Coordination between these two ministries was a problem in almost all countries visited. The ministry of education generally provides funds for medical schools, whereas the ministry of health is the main employer of school graduates. In many countries, coordinated planning for budgets, priorities, and outcomes between ministries of health and education is poor, which contributes to inappropriate curricula and the graduation of doctors who cannot find employment in the country. In Mali and Sudan, the yearly number of medical graduates substantially exceeds the in-country www.thelancet.com Vol 377 March 26, 2011

Rural private general practice 5%

Urban public general practice 21% Rural public general practice 9%

Urban private general practice 13% Other 1% Left the practice of medicine 2% Migrated to other African countries 6%

Urban specialist practice 19% Migrated outside of Africa 22%

Rural specialist practice 3%

Figure 5: Estimated location of graduates from medical schools in sub-Saharan Africa 5 years after graduation Data are mean reported percentage of graduates (N=62).

capacity to hire new physicians, despite the need for health services. Overall, ministries of education seem to be more active in setting of medical school priorities than are ministries of health. 69 of 100 survey respondents reported that ministries of education either contribute substantially or are the main drivers of priorities, compared with 49 of 101 reporting that ministries of health are the substantial or main contributors. One survey respondent mentioned that an important innovation had been the transfer of supervision of the school from the ministry of education to the ministry of health.

(6) Accreditation and quality measurement Accreditation and quality measurement are important developments to standardise medical education and physician capabilities. Various levels of accreditation and certification were noted in the countries visited. In Ethiopia, there is no official continuous accrediting body for medical schools. Accreditation is granted only when an institution is initially founded. However, many schools report progress in accreditation of institutions and assessment of graduates. In Mozambique, the newly formed Medical Council plans to develop accreditation standards for medical schools and external examinations for medical students. The Tanzanian Commission for Universities visits teaching institutions once before accreditation and then every 4 years. The Malawi Medical Council uses guidelines from the Southern African Development Community for accreditation and quality assurance.

(7) Educational planning focused on national health needs Education planning focused on national health needs is improving the ability of medical graduates to meet such needs. Schools are increasingly emphasising community oriented, relevant, or nationally focused medical 1117

Health Policy

40

Central Africa Southern Africa East Africa West Africa

Number of medical schools (N=84)

32 30

20 17 14 10

10

5 3 0

0 0%

1–9%

10–19%

20–29%

30–39% 40–49% Faculty involvement

50–59%

1 60–69%

2 100%

Figure 6: Faculty involvement in grant supported research in medical schools in sub-Saharan Africa, by region

education. Many of the schools are developing curricula around national priority health problems, and are using rural and community-based experiences to improve their programmes. Although some initiatives are undertaken by the schools alone, many are set in the context of government priorities and national service programmes. In Malawi, the curriculum is designed to immerse students in local health issues. The curriculum focuses on the most common diseases and health disorders in Malawi and the surrounding region. At Gezira University, community-based courses make up 25% of studies, and many courses are undertaken at field sites including district hospitals, community health centres, clinics, and patients’ homes. The Catholic University is incorporating management training for students, in recognition that some graduates will be serving in administrative positions as regional health officers or hospital chief medical officers after graduation.

(8) Importance of research Beyond the creation of new knowledge, research is important for development of medical school faculties, retention of staff, and infrastructure strengthening. Many schools reported that research promoted staff recruitment and retention and attracted external partners. Older schools with stronger research portfolios continue to develop research programmes with use of established sources of funding, providing some capability to train young faculty members. However, although well established schools experience success in garnering research support, newer and smaller schools face a challenge. These schools often have younger faculty staff who do not have the training and mentorship to bid successfully for research grants. Additionally, staff shortages at many schools increase the teaching load, restricting time available to pursue research. 1118

The University of Mali has purposefully built its research capacity over 30 years. Initial faculty members were sent abroad for graduate training, and returning graduates were guaranteed support. Present research faculty staff teach at the medical school, benefiting both staff and students. The University of Malawi collects 10% indirect costs from all research grants to create a financial base to support research. Their Research Support Centre assists the faculty in grant writing, research design, and grant administration, which further develops research capacity at the university increase salaries for research faculty staff. Despite these examples, at most medical schools in subSaharan Africa less than 10% of faculty members are involved in sponsored research (figure 6). We compared the percentage of faculty members involved in grantsupported research with a series of potential predictive variables by multivariable linear regression. Although many universities provide several types of research support (webappendix p 15), only two types (provision of strengthened institutional research instruments such as research or ethics committees [p=0·0002] and provision of funded research time [p=0·049]) were significantly correlated with an increased percentage of faculty members involved in grant-supported research. The presence of research training programmes (internal or external) was not a significant factor. We recorded linguistic variation, with schools using English as a language of instruction having faculties more likely to participate in research (p=0·027) and Arabic speaking faculties being less likely to do so (p=0·047).

(9) Curricular innovations Impressive curricular innovations are occurring in many schools. Survey respondents reported several nontraditional teaching methods used in medical schools in sub-Saharan Africa, including community-based education, problem-based learning, and multidisciplinary team-based learning (webappendix p 16). These three educational methods tended to be implemented together. Institutions using a higher degree of any of these instructional methods during the preclinical years were more likely to use the other methods during this period (p<0·0001 for each correlation), and institutions using more team-based learning during clinical years were likely to use more problem-based (p=0·0003) and communitybased (p<0·0001) learning during this period. Structured community exposure and communitybased education provide students with experiences working with underserved populations and improve graduates’ preparation to deal with national health problems. At Jimma University, community-based education is central to the educational mission, starting with a community-based training programme. Medical students begin work at community sites after successfully mastering tasks including data collection, community diagnosis, analysis, and preparation of a plan for an intervention for local problems. At the College of Health www.thelancet.com Vol 377 March 26, 2011

Health Policy

Sciences at Makerere University, Uganda, a new curriculum includes regular exposure to patients in rural communities throughout medical school. At Gezira University, the curriculum includes community orientation, providing an organising principal for medical students, faculty staff, and graduates. A sense of social accountability is present in all aspects of education, and the school’s community oriented mission is prominently displayed at the entrance to the medical school building. Faculty members in oncology, nephrology, and paediatrics commented to the site visit team about the continual effect of community oriented principles in their work.

(10) Significance of postgraduate medical education Postgraduate medical education is an important aspect of a national health-system development strategy. Of 96 survey respondents who reported about their postgraduate programmes, 58 reported that clinical postgraduate programmes were offered, and 47 reported the number of graduates per year from these programmes. These 47 programmes included 1909 places for clinical postgraduate programmes, representing available places for less than 25% of the 7861 total graduates reported to the study. The most commonly offered postgraduate programmes were in internal medicine (14% of reported postgraduate medical education places), obstetrics and gynaecology (13%), paediatrics (13%), surgical subspecialties (11%), and general surgery (11%). Makerere and Ibadan Universities have well developed systems of postgraduate medical education. Other schools offer fewer programmes for postgraduates, often paediatrics, surgery, and obstetrics/gynaecology, whereas a few have no postgraduate programmes. By increasing such programmes, schools have been able to retain more graduates and to hire some of the newly trained graduates as faculty members. Some schools implement so-called sandwich postgraduate programmes in which residents pursue the programme at home but spend time during training at a regional or international programme. This principle is intended to provide exposure to clinical work abroad while mitigating the tendency of doctors to remain abroad when all their training is done elsewhere.

(11) Quality of secondary schools Variability in quality of secondary schools creates challenges in medical school admissions. Sound secondary education systems are prerequisites to success in medical school, but the quality of secondary education is not strong in many countries. As a result, many schools have developed preparatory or recruiting programmes for disadvantaged students. The Catholic University, Hubert Kairuki Memorial University, and the College of Medicine in Malawi have implemented preparatory years to assist students in meeting the demands of formal medical education. The Walter Sisulu University has established quotas by race to reflect the demographics of www.thelancet.com Vol 377 March 26, 2011

South Africa, and assesses applicants with a scoring system that includes motivation and commitment to service. Additionally, the first-term curriculum provides language instruction, computer training, basic science remediation, and a focus on study skills. A peer mentoring programme helps students to adjust to campus life.

(12) Private medical schools Despite challenges, private medical schools represent an area of innovation and growth in medical education in sub-Saharan Africa. The first private medical schools opened in the 1990s, and private schools now constitute 21% (22 of 105) of all responding schools (figure 2). A search of databases suggests that a third (21 of 63) of non-responding schools are privately owned, which would mean that 26% (43 of 168) of currently operating medical schools are private schools. Supporters argue that the privatisation of medical education is in keeping with global trends in education, whereas others counter that this movement is largely commercial and inevitably inequitable. Two of the ten schools visited were private institutions. The Hubert Kairuki Memorial University is a not-forprofit institution founded in 1997, which has earned a reputation for graduating high-calibre physicians. This university partners with public district hospitals to provide students with additional clinical teaching sites. The Catholic University is a faith-based, not-for-profit school founded in 1995. The university is a model of successful collaboration with the Government of Mozambique, the Catholic Church, and several international organisations.

(13) International partnerships International partnerships are an important asset for many medical schools. Almost all medical schools in subSaharan Africa, including all schools visited, are engaged in collaborations locally and internationally, mainly with institutions in Europe, North America, and Africa (webappendix p 17). The University of Mali cooperates with the French Government, which has assisted them with programmes in community health practice and public health training. Mali has also developed research collaborations with European, North American, and other African universities. Makerere University has a long history of collaboration with foreign academic and nonprofit organisations. These linkages have contributed to the school’s research capabilities and training programmes. The University of Malawi exemplifies a so-called south–south collaboration, with its involvement in joint training programmes through the Southern Africa Human Capacity Development Coalition.

Discussion We have noted a remarkable growth in medical education in the region over the past two decades, which began well before the attention by the international community to the massive shortage of health workers in the region. The decision of many countries to invest in building new 1119

Health Policy

medical schools and expanding present ones, young people’s intense interest in the study of medicine, and the emergence of private medical education all provide evidence of this movement. The obstacles to scale-up, however, remain substantial. Without larger and more stable health workforces, there is little chance of stemming the AIDS epidemic or realising the Millennium Development Goals in Africa. Many international organisations (including the US President’s Emergency Plan for AIDS Relief, WHO, World Bank, the Global Fund, bilateral aid agencies, and philanthropies) have reached this conclusion.9 The present global attention being paid to scale-up of health workforce makes this a favourable time for medical education in sub-Saharan Africa. The universal insufficient numbers of faculty staff and the general disrepair of campus infrastructure are among the most prominent findings documented by this study. These findings are not surprising to anyone familiar with medical education in the region. Basic science and clinical faculty members are in short supply everywhere, severely restricting quality educational scale-up. Substantial deficiencies are often present in laboratories, libraries, classrooms, lecture halls, and hostels. The deficit in information technology and bandwidth is especially problematic, denying students the possibility of bypassing older learning methods to benefit from the rapidly advancing developments in internet-based learning. The study documented the consistency and magnitude of these problems, establishing a baseline from which to pursue scale-up. The convergence of global attention and clear specification of impediments facing the expansion of medical education create a moment of opportunity. National authorities and international partners working together with medical colleges have the possibility of creating new, collaborative investment strategies, developing long-term commitments for conjoint funding with use of mechanisms such as endowment funds for faculty enhancement, dedicated funds for new laboratories and computer needs, or umbrella funding to support north–south or south–south partnerships between universities. An unforeseen aspect of this study was the opportunity it afforded African medical education leaders to meet and work together. Twice previously (in the 1960s and the 1990s) an African Medical Schools Association was initiated but not sustained.10 Most medical education policy in Africa has centred on programmes and issues within countries, with little opportunity to discuss common problems and solutions between countries. The opportunity afforded by this study for educational leaders to visit elsewhere in Africa to compare programmes and policies proved very instructive and energising. A revitalised African Medical Schools Association would be in a position to have an important role in the expansion of medical education in sub-Saharan Africa, including data collection for existing programmes, documentation of best practices, promotion 1120

of accreditation and certification initiatives, and updates on present and developing medical schools. This study also noted an important change in the role of postgraduate medical education in sub-Saharan Africa. Positions in postgraduate medical education programmes are not plentiful in most African countries. Many African medical graduates seeking advanced training travel to Europe and North America, and many do not return. To train more graduates in basic specialties and to stop the emigration of young graduates, many countries are now opening and expanding postgraduate medical education programmes. This trend will be essential to build sustainable, quality physician cadres in all countries and to improve the ranks of medical school faculties. The study was not modelled on any previous work but succeeded in creation of a data gathering strategy that is unique and potentially applicable to studies of other geographical areas in which information about health professional education is not well developed. The key characteristics of the study structure were: (1) literature review and key informant interviews to establish baseline knowledge; (2) appointment of an advisory board of educators and policy leaders from the area of study to provide perspective, advice, findings analysis, and conclusion framing; (3) data gathering with qualitative (structured site visits) and quantitative (survey) instruments, both involving primary participation by the advisory board; and (4) inclusive (advisory board and secretariat) engagement in analysis, writing, editing, and publishing.

Limitations This study had several limitations. The published work about medical education in Africa was mainly in English. Articles written in French were fewer, and in Portuguese and Arabic rare. Although attempts were made to identify more non-English articles with the assistance of study Advisory Committee members, few were found. The preponderance of Anglophone Advisory Committee members and staff also restricted the study’s use of nonEnglish sources of information. Of the ten site visited schools, two (in Mali and Côte d’Ivoire) were French speaking (one less than population parity would warrant), six were Anglophone, one was Lusophone, and one was Arabic speaking. The under-representation of Francophone countries occurred despite several attempts to establish contacts with schools in the Democratic Republic of the Congo, Senegal, and Cameroon. Similarly, the survey responses showed some regional under-representation, with lower response rates in countries such as the Democratic Republic of the Congo, Angola, and Sudan. 22 schools were not surveyed because they were founded or identified after the study was completed. Survey limitations included the subjective nature of several questions, unanswered questions in surveys of some individual schools, and inconsistently answered questions from schools within the same countries. Questions such as the proportion of income from various www.thelancet.com Vol 377 March 26, 2011

Health Policy

Panel: Recommendations to promote and improve medical education and population health in sub-Saharan Africa 1 Launch campaigns to develop capacity of medical school faculties, including recruitment, training, and retention 2 Increase investment in medical education infrastructure 3 Build structures to promote interministerial collaboration for medical education 4 Fund research and research training at medical schools 5 Promote community oriented education based on principles of primary health care 6 Establish national and regional postgraduate medical education programmes to promote training, excellence, and retention 7 Establish national or regional bodies that are responsible for accreditation and quality assurance of medical education 8 Increase donor investment in medical education aligned with national health needs 9 Recognise and review the growing role of private institutions in medical education 10 Revitalise the African Medical Schools Association

sources, reasons for staff loss, and graduates’ emigration and practice choices were usually estimates by respondents rather than data-based answers. Neither the survey nor site visits assessed graduate competencies because of the absence of established regional standards of educational quality. Unanswered questions within returned surveys proved problematic. Some questions were understandably omitted by specific schools, such as questions about graduates from schools that had yet to graduate students. When questions were left unanswered without explanation, attempts were made to contact respondents to complete questionnaires. The number of responses to each relevant question is reported for each finding. In some cases, inconsistent answers pertaining to national requirements were found in schools in the same countries. For example, eight countries with multiple schools responding gave inconsistent answers about whether a compulsory service requirement exists in their country.

Conclusions and recommendations This study analysed the workings of African medical schools during 2 years of investigations. Advisory Committee members felt that the perspectives they had developed in this work warranted collective recommendations about actions that should be taken to promote and improve medical education and, in turn, population health in sub-Saharan Africa. A set of recommendations was drafted, circulated to all participants in the study, discussed, modified, and ratified at a study meeting in Dar es Salaam, Tanzania, on April 16, 2010. The panel shows these recommendations. We hope that the work of the Sub-Saharan African Medical School Study and its recommendations will www.thelancet.com Vol 377 March 26, 2011

catalyse further innovation and investment in medical education in sub-Saharan Africa. This action, in turn, should help to produce a more robust workforce with the aim of improved health in Africa. Contributors All authors contributed to the research for this article. FM, SF, FO, SRG, EB, and CC contributed to the design of the article; FM, SF, CC, SRG, TW, and HR to the writing of the article; and EB, CC, FM, SF, and TW to the data analysis. Conflicts of interest FO, PR, and NKS received grant funding from the Bill & Melinda Gates Foundation. HA, SC, EB, CC, JC, SBC, SF, SRG, TH-M, EH, LJ, FM, HR, and TW received salary indirectly from grant funding from the Bill & Melinda Gates Foundation. AH, EB, MJ-MID, DEEA, JF, JI, MJ, GLM, EOO-O, FO, and NKS received honoraria indirectly from the Bill & Melinda Gates Foundation. EH and A-JN received a consulting fee indirectly from the Bill & Melinda Gates Foundation. AH, HA, SBC, EB, CC, JC, SC, MJ-MID, DEEA, JF, SF, SRG, EH, JI, MJ, MM, GLM, FM, A-JN, HR, EOO-O, FO, PR, and NKS received support for travel to the African Medical Education Symposium for study purposes. EB, CC, JC, MJ-MID, TH-M, JI, MJ, LJ, MM, GLM, FM, A-JN, EOO-O, FO, PR, and NKS received support for travel to the initial study meeting in Kampala, Uganda, for study purposes. CC, JC, JI, MM, GlM, FM, AN, HR, EOO-O, FO, PR, and NKS received support for travel to site visit meetings for study purposes. AH, EB, MJ, A-JN, and NKS received fees for participation in review activities such as data monitoring boards, statistical analysis, endpoint committees, etc. AH, EB, and A-JN received payment for writing or reviewing the report. AH and EB were provided administrative support. The following authors state that they have received other grants or have grants pending from the Bill & Melinda Gates Foundation: CC will travel to a meeting in Kampala, Uganda, in November, 2010; NKS is part of a subcontract for a twinning grant for Makerere University to Johns Hopkins University; and MJ-MID and NKS received funds to organise site visits by the study team to their institutions. All other authors declare that they have no conflicts of interest. Acknowledgements This work was supported by the Bill & Melinda Gates Foundation. References 1 WHO. Working together for health; the World Health Report 2006. Geneva: World Health Organization, 2006. 2 Joint Learning Initiative. Human resources for health: overcoming the crisis. Cambridge: Harvard University Press, 2004. 3 Zuvekas A, Nolan L, Tumaylle C. Impact of community health workers on access, use of services and patient knowledge and behaviour. Washington: Bureau of Primary Health Care, US Public Health Services, 1998. http://aspe.hhs.gov/pic/reports/hrsa/6355. pdf (accessed June 24, 2010). 4 Mullan F, Frehywot S. Non-physician clinicians in 47 sub-Saharan African countries. Lancet 2007; 370: 2158–63. 5 WHO. Task shifting global recommendations and guidelines. Geneva: World Health Organization, 2008. http://data.unaids.org/ pub/Manual/2007/ttr_taskshifting_en.pdf (accessed June 24, 2010). 6 WHO. Report on the WHO/PEPFAR planning meeting on scaling up nursing and medical education. Geneva: World Health Organization, 2009. http://www.who.int/hrh/resources/scaling-up_ planning_report.pdf (accessed June 24, 2010). 7 WHO. World health statistics 2010. Geneva: World Health Organization, 2010. http://www.who.int/whosis/whostat/EN_ WHS10_Full.pdf (accessed June 24, 2010). 8 Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat. World population prospects: the 2008 revision. March, 2009. http://esa.un.org/unpp (accessed June 1, 2010). 9 WHO and Global Health Workforce Alliance. Scaling up, saving lives: Task Force for Scaling Up Education and Training for Health Workers. Geneva: World Health Organization and Global Health Workforce Alliance, 2008. http://www.who.int/workforcealliance/ documents/Global_Health%20FINAL%20REPORT.pdf (accessed June 10, 2009). 10 Bowers JZ. Association of African medical schools. J Med Educ 1962; 37: 220–21.

1121

Wakley Prize Essay

Health district development and the need to dig deeper Ellen M Einterz Lancet 2011; 377: 1122–23 Published Online December 16, 2010 DOI:10.1016/S01406736(10)62191-5 See Comment Lancet 2010; 376: 2053 See Wakley Prize Essay Lancet 2011; 377: 350 and Lancet 2010; 376: 2114 Kolofata District Hospital, Kolofata, BP 111, Mora, Extrême-Nord, Cameroon (E M Einterz MD) Correspondence to: Dr Ellen M Einterz, Kolofata District Hospital, Kolofata, BP 111, Mora, Extrême-Nord, Cameroon [email protected] To listen to the essay see Online for webaudio

1122

It is probably possible to fill a well by pouring water from above, but better in the long run to dig deeper and let water rise from below. Where people thirst for good health care, the same is true. As Africa struggles with fading hope of reaching its Millennium Development Goals for health, those goals for which so much unprecedented international aid has been allocated over the past decade, a visit to a typical health district might help to explain why. The district covers a dusty Sahelian plain dotted with baobabs, spiny shrubs, and thorn trees. In the middle of this plain is the district hospital—11 tin-roofed one-storey rectangular buildings situated on a large compound. Power to the district has been out for 28 of the past 30 days, but today there is hope: the electricity company says that tomorrow the lines will be fixed. True, they said the same thing yesterday, the day before yesterday, and the day before that. A clunky generator that devours diesel and breaks down every other day roars at one end of the compound. The hospital has installations for running water, but since water only runs when there is power in the town, the pipes have been dry for quite some time. A physician and a surgeon, the only doctors for the district population of 122 000, work here alongside 30 nurses and 23 auxiliary personnel. The medical, paediatric, obstetric, and surgical wards hold a total of 100 beds—cots, really—and not only are they all occupied, but some are laden with two or three patients apiece. It is the last month of the rainy season, so there is a lot of malaria about. In the paediatric ward half the children are being treated for malaria. They are feverish; some convulse intermittently; two are deeply comatose; two others are receiving transfusions of their mothers’ blood. There are children with pneumonia and bronchiolitis, others with AIDS. A nurse is starting a transfusion on a new patient, a floppy 40-day-old infant whose eyes keep rolling up into his head. Like most 40-day-old infants in these parts, he had his uvula amputated by the local barber, but unlike most, he haemorrhaged massively after the excision. There are malnourished children here, some just wasted and stunted, others with full-blown kwashiorkor—the bloated bellies, oedematous limbs, oozing sores, limp yellowish hair, and the implacable irritability or the abject lethargy. There is a 3-month-old girl who weighs 2 kg. Her mother bled to death on the dirt floor of their hut 2 hours after delivery, and the girl has been sucking at the withered breasts of her old granny since birth. In the medical wards are more cases of malaria, more respiratory tract infections. There is a young man, bitten by a carpet viper yesterday, with a hideously swollen right hand, arm, and neck. Blood seeps from his gums and

from the healer-incised nicks along the bitten arm. There are ten patients with late-stage AIDS. A 16-year-old girl whose left ear was removed by a co-wife brandishing an axe sleeps curled up in a bed in the corner. There are patients with heart failure, diabetes, renal failure. A middle-aged man with hepatitis-B-induced cirrhosis waits to have his dome-shaped abdomen tapped. Another man, injured when his motorcycle hit a craterous pothole and threw him against a boulder, is just awakening from a 5-day coma. The surgical ward is imbued with the faintest whiff of urine and has several older men recovering from prostatectomies. A couple of victims of motorcycle accidents have fractured limbs pinned in external fixation devices. Two women who have had caesarean sections occupy adjoining beds. One is still in coma with eclampsia. Nearby a 14-year-old girl awaits repair of a rectovesicovaginal fistula. She had been in labour at home for 3 days before the local midwife wrenched out a dead baby. A few days after that, the family decided it was time to seek medical care. So they got a wheelbarrow and put her in it and rolled her 12 km through the sand to reach the hospital. On initial examination, the tissues of her vagina had come out in globs of putrid brown-yellow rot and her lower urinary and gastrointestinal tracts leaked through huge holes. Her pelvis had been so yanked out of shape that to move her legs she had to take them in her hands and lift them one at a time. Over in the maternity ward, a pudgy boy born last night by vacuum extraction lies beside his exhausted mother. Two women in what appears to be normal labour pace the floor. Another woman, visibly at term, is carried into the delivery room draped in the arms of her husband. His face is slick with sweat and his wide eyes glisten with terror. A copious stream of bright red blood flows to the floor in their wake. Tucked behind the other buildings at the far end of the compound is the isolation ward. Here on one side of the building are five patients with tuberculosis. Two of them also have AIDS, and one has a form of tuberculosis that is resistant to the regimen of five available antituberculous drugs. On the other side of the building are a dozen patients with cholera, the latest of the hundreds that have been treated since the epidemic erupted 2 months ago. A new one is just arriving: a 3-year-old girl, watery fluid pouring from both ends as if from two spouts. Her father carries her upright at arm’s length, like a crucifix to an altar. She is so parched her eyeballs are lost in their sockets. She has no voice, no saliva, and no muscle tone. The father sets her on a plastic-covered mattress on the floor. A nurse tries to give her oral rehydration solution but the girl vomits it immediately. A nasogastric tube is inserted, but the vomiting www.thelancet.com Vol 377 March 26, 2011

Wakley Prize Essay

continues. The nurse searches painstakingly for an uncollapsed vein, finds one and starts an intravenous drip, but 10 minutes later the vein blows. The room reeks of cholera, a dead sea sort of smell, and the patient is slippery slimy with the stuff that just keeps pouring out. Another squiggler of a vein is found and the drip is placed once more, and now a desiccated 60-year-old man, hooked by the armpits and dragged by his two sons, arrives. In the outpatient department, some 100 patients are awaiting consultation and listening as a health promoter flips a chart and talks to them about AIDS. A nurse in a side room is receiving pregnant women for antenatal check-ups, while in the courtyard behind the building, on a scale hanging from a tree limb, children are being weighed. Others are receiving vitamin A supplements and vaccinations. In the pharmacy and the laboratory, frazzled staff scurry to serve the backed-up crowd. Beyond the district hospital is a ring of seven health centres set in villages 10–30 km away. Each health centre is staffed by a handful of nurses who deliver babies and provide curative care for up to 60 patients a day. The more problematic of these are referred to the district hospital. Supervised by the district service, each centre also conducts a wide range of routine activities that include antenatal care, vaccinations, disease surveillance, epidemic prevention and control, and health education. Most of these activities take place in the health centre and in the villages as well, and they are planned and executed in collaboration with community representatives. The district health system, with its grassroots network of hospitals and primary health-care centres, is the health system recommended by WHO and officially adopted by countries in Africa. It is capable of providing the whole population with high quality health care that is appropriate, comprehensive, community-based, and community-driven. It is a system that merits development and support. It is also the system more sabotaged than sustained by international aid organisations whose white-knuckle grip on disease-specific programmes prevents them from embracing broader possibilities. Subsidised tuberculosis treatment programmes are good, but their value is limited in the absence of reliable community health structures needed to provide early diagnosis. Malaria prevention is important, but nets will not save the thousands of children dying because there are not enough antimalaria drugs or blood bags available beyond the cities. Mobile clinics sent into towns for semi-annual 1-day-only HIV testing are one way of detecting new cases—the photo ops are excellent too—but better would be tests made available all year, every day, in every healthcare facility. National mass vaccination and treatment campaigns—blitz operations favoured by funding agencies for their easily measured and rapid results— may produce impressive numbers in the short term, but they do this only by disrupting routine health-care activities, emptying health centres of personnel, creating www.thelancet.com Vol 377 March 26, 2011

shortages of medicines and vaccines, and depriving vulnerable people of service. It is unhelpful for aid personnel to be blind to these shortcomings. Aid organisations—most based in Washington, Paris, Atlanta, Geneva, and other places like them—have experts who necessarily view the world from where they sit, thousands of miles away in steel-girded buildings with year-round electricity, computers connected to high-speed wireless internet, air conditioners they can turn on when the air gets above body temperature. They live in countries generally governed by the rule of law, where bribery is not a fact of daily existence for most people, famine is unknown, water—drinkable water at that—comes from a tap, tribalism is not a threat to the social order, watches are worn for the purpose of telling time, and women are rarely imprisoned or executed for being witches. There is nothing wrong with Americans living an American lifestyle or Europeans living a European lifestyle, but this detachment—in distance, culture, and technology—maintained by people trying to fix things in Africa is dangerous. Only by being physically present for extended stays at the operational level of the country they are helping—not half a world away, not even holed up in the capital—can staff of aid organisations watch those wells and witness up close what happens when the digging stops too soon and water is just poured from above. Only there, at the district hospital, in the village, can they see, really see, that the needs of the people are urgent and constant, life is precarious, and the provision of good health care is seamless, not partitioned into disease-specific programmes. The fight against maternal and childhood mortality, measles, AIDS, tuberculosis, malaria, tetanus, polio, and all those other diseases should indeed be waged vigorously and with all the arms in our armoury—but it should be waged every day from the bottom up, not in fits and starts from the top. Except in rare circumstances, vaccination and treatment campaigns should be abandoned and their aims and resources integrated into routine services offered by health-care facilities within the district health system. To strengthen that system, physicians, surgeons, and nurses should be enticed to work in rural areas, where the best of the best are needed. Aid should promote the development and provision of appropriate technology: better diagnostic tests, less expensive and more durable imaging equipment, new vaccines, new drugs, cheaper drugs, more robust drugdelivery systems, and improved computer and internet access. Investments in female education, water, sanitation, roads, and energy would do more than any campaign to develop health care in Africa and improve the health of the people. “Thunder is good, thunder is impressive”, Mark Twain wrote, “but it is lightning that does the work”. A functioning district health system can quietly illuminate an entire country, and international aid should do everything it can to brighten, broaden, and prolong that light. 1123

Case Report

Donor-derived adult T-cell leukaemia Akira Nakamizo, Yojiro Akagi, Toshiyuki Amano, Satoshi O Suzuki, Rie Otsuka, Yasunobu Abe, Koji Yoshimoto, Toru Iwaki, Tomio Sasaki Lancet 2011; 377: 1124 Department of Neurosurgery (A Nakamizo MD, Y Akagi MD, T Amano MD, K Yoshimoto MD, T Sasaki MD); Department of Neuropathology (S O Suzuki MD, T Iwaki MD), and Department of Medicine and Bioregulatory Science (R Otsuka MD, Y Abe MD), Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan Correspondence to: Akira Nakamizo, Department of Neurosurgery, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan [email protected]. kyushu-u.ac.jp See Online for webappendix

In July, 2008, a 55-year-old woman was referred to us with a 2-month history of progressive aphasia after a febrile episode. She had been diagnosed with chronic adult T-cell leukaemia in June, 1999, and had had a nonmyeloablative peripheral-blood stem-cell transplant with conditioning of pentostatin and 2 Gy total body irradiation in March, 2002. The stem cells had been harvested from her brother, an asymptomatic carrier of human T-cell lymphotropic virus type 1 (HTLV-1). After treatment, she had complete remission. Ciclosporin was discontinued 6 months after the transplantation because of graft-versus-host disease. Our patient had had no bacterial or viral infections. She had no skin lesions, lymphadenopathy, or splenomegaly. Her leucocyte count was 8·74×10⁹/L with 2·5% adult T-cell leukaemia-like cells. The leukaemia was therefore considered to be in remission. CD3-positive cells in the peripheral blood were confirmed as donor derived by XY-fluorescence insitu hybridisation (XY-FISH). Serum IgG was 13·7 g/L, IgA was 2·82 g/L, and IgM was 1·16 g/L. Our patient was immunocompetent; absolute number of circulating cells positive for CD4 was 1·74×10⁹/L, CD8 0·475×10⁹/L, CD19 0·522×10⁹/L, and CD56 0·326×10⁹/L. Brain MRI showed hyperintense, non-enhancing lesions in the left temporoparietal lobe and right frontal lobe (figure A), which suggested encephalitis, progressive multiple leukoencephalopathy, demyelinating disorder, or lymphoma. Serum concentrations of calcium, lactate dehydrogenase, soluble interleukin-2 receptor, and β-2 microglobulin were not raised. CSF examination, culture, and PCR, abdominal ultrasonography, endoscopy, and ¹⁸F-flurodeoxyglucose (¹⁸F-FDG) PET were normal. Craniotomy and biopsy of the left temporal lobe was done in July, 2008. Histopathology showed perivascular accumulation and cerebral infiltration by mononuclear tumour cells with large, indented nuclei (figure B). Tumour cells were positive for CD3, CD4, CD8, and CD45RO, but negative for B-cell markers (CD20 and CD79a), suggesting cerebral invasion by adult T-cell leukaemia cells A

(webappendix). XY-FISH showed an XY genotype of the tumour cells (figure C), suggesting that they had originated from the donated leucocytes. After whole brain irradiation and localised irradiation of the left temporoparietal lobe, our patient’s aphasia improved remarkably, and repeat MRI showed shrinkage of the brain lesions. In October, 2010, our patient was fully recovered. For patients with adult T-cell leukaemia, asymptomatic carriers of HTLV-1 have been widely considered acceptable as donors in allogeneic stem-cell transplantation.1 The immune system is thought to have an important role in leukaemia occurrence in HTLV-1-seropositive carriers.2 The only reported case of donor-derived adult T-cell leukaemia was in 2006, in an immunosuppressed recipient of a peripheral-blood stem-cell transplant.3 No cases of donorderived adult T-cell leukaemia have been reported in immunocompetent patients. The prognosis tends to be better for these patients than for those with CNS involvement subsequent to systemic adult T-cell leukaemia.4 Only two cases of isolated CNS involvement have been reported so far.4,5 The function of CD4-positive cells was not fully analysed, so there is a possibility of some functional deficit in immune reconstitution. Our patient’s case suggests that the consent practices in patients who receive HTLV-1-positive cells should be re-evaluated. Contributors All the authors looked after the patient and did the investigations. AN, YA, SOS, and TS wrote the report. Written consent to publish was obtained. References 1 Utsunomiya A, Miyazaki Y, Takatsuka Y, et al. Improved outcome of adult T cell leukemia/lymphoma with allogeneic hematopoietic stem cell transplantation. Bone Marrow Transplant 2001; 27: 15–20. 2 Hoshida Y, Li T, Dong Z, et al. Lymphoproliferative disorders in renal transplant patients in Japan. Int J Cancer 2001; 91: 869–75. 3 Tamaki H, Matsuoka M. Donor-derived T-cell leukemia after bone marrow transplantation. N Engl J Med 2006; 354: 1758–59. 4 Dungerwalla M, Osuji N, Waldman AD, et al. Isolated central nervous system involvement in adult T-cell lymphoma/leukaemia. Br J Haematol 2005; 130: 511–15. 5 Marshall AG, Pawson R, Thom M, Schulz TF, Scaravilli F, Rudge P. HTLV-I associated primary CNS T-cell lymphoma. J Neurol Sci 1998; 158: 226–31.

B

10 µm

C

30 µm

Figure: Donor derived adult T-cell leukemia (A) T2-weighted brain MRI showing hyperintense, non-enhancing lesion in the left temporoparietal lobe (arrow) and right frontal lobe; (B) Histopathology showing perivascular accumulation of mononuclear tumour cells with large, indented nuclei (haematoxylin and eosin stain). Inset on the bottom left shows indented nuclei (green arrows); and (C) XY-FISH analysis shows the tumour cells to have an XY genotype. Red chromosome X; green chromosome Y.

1124

www.thelancet.com Vol 377 March 26, 2011