Intended for healthcare professionals


Ethics and international research

BMJ 1997; 315 doi: (Published 18 October 1997) Cite this as: BMJ 1997;315:965

Research standards are the same throughout the world; medical care is not

  1. Neal A Halsey, Professora,
  2. Alfred Sommer, Deana,
  3. Donald A Henderson, Distinguished professora,
  4. Robert E Black, Chaira
  1. a Johns Hopkins University School of Hygiene and Public Health, Baltimore, MD 21205, USA

    A recent commentary in the New England Journal of Medicine by Lurie and Wolfe criticised placebo controlled trials designed to identify simple and effective interventions to prevent maternal-infant HIV transmission in developing countries.1 Their commentary reflects a lack of understanding of the realities of health care in developing countries and ethical principles of research. The commentary and an accompanying editorial by Angell2 take a position that would prevent developed countries from collaborating with developing countries to identify practical and affordable health interventions.

    Lurie and Wolfe propose that studies supported by the United States government should provide all participants with the same level of care that is available to Americans. This is a misinterpretation of the Council for International Organisation of Medical Sciences guidelines, which call for universal principles of ethical research, not universal standards of medical care.3 Under the guidelines, interventions must be appropriate for the country where the research is conducted and no research subjects may be denied care that would otherwise be available for them. The studies criticised adhere to these principles.

    Also, providing care routinely provided to women and their infants in America, at a cost of thousands of dollars per patient, would serve as a powerful incentive to participate in trials in countries where per capita health care expenditures are usually less than $10. This would violate the guideline to avoid undue inducements for participation in research and would make almost all research sponsored by US organisations in these countries totally impracticable. If these unsustainable services were provided on a temporary basis what would happen when the research project ended and local practitioners could no longer provide diagnostic tests, infant monitoring, and intensive care units necessary to support the regimen?

    In 1994 the ACTG 076 study conducted in the United States and France showed that a complex regimen of zidovudine administered orally from mid-pregnancy to delivery, intravenously during labour, and orally to the infant for six weeks reduced the rate of maternal-infant HIV transmission from 25.5% to 8.3%.4 This regimen has become the standard in America and some European countries but in most developing countries has not been implemented because the medical care infrastructure cannot identify HIV infected women early in pregnancy or deliver the regimen; most women seek prenatal care too late in pregnancy to be offered a regimen that begins early in the second trimester; many deliveries take place outside hospitals; intravenous infusions for all women are impractical (or unsafe); and the regimen costs 100–500 times the healthcare funds available per person per year in many countries. Moreover, many pregnant women are anaemic and anaemia is a common complication of zidovudine therapy, so the safety of the 076 regimen is unknown in these settings.

    Studies are in progress to evaluate short course zidovudine regimens that would be more affordable and practical than the 076 regimen. Since the effectiveness of these short regimens is not known, they are being compared with placebo. In these countries zidovudine and other antiretroviral drugs are not available for pregnant women. Women taking part in these studies are fully informed of their HIV infection status, the purpose of the study, the probability of receiving drug or placebo, and the possibility of not participating. The studies have been reviewed by ethics committees locally and at major universities and discussed by the National Institutes of Health, the Centers for Disease Control and Prevention, and the World Health Organisation. Also, data and safety monitoring boards monitor the studies. None of the participants are being denied the care they would otherwise have. Any analogy with the Tuskeegee study, where none of these safeguards were in place, is offensive.5

    Lurie and Wolfe argue that all women in these studies should receive some zidovudine. They propose comparing a short course regimen to the 076 regimen or comparing two different short course regimens. Both designs would probably yield uninterpretable results. If the 076 regimen was shown to be better than the short course no one would benefit as the 076 regimen could not be implemented and we would not know if the short course was efficacious in the absence of a control arm. Until we know that short course regimens are safe and efficacious, any study showing similar rates of transmission with two different short regimens would be uninterpretable, as it would be impossible to determine the benefit produced by either regimen. Exposing people to the potential risks of research that is unlikely to answer the study question is unethical. If a short course regimen is shown to be safe and effective in future then it will become the comparison arm for future studies instead of a placebo.

    Imposing the American standard of medical care on all participants in international trials funded by the United States would prevent developed nations from collaborating with developing countries to identify feasible and affordable means of preventing and treating many diseases. Not only is this considered “medical and ethical imperialism” by colleagues in developing countries; it would also have prevented the development of many interventions, such as oral rehydration, micronutrient supplementation, and low cost surgical procedures,6 7 8 that have dramatically improved health care throughout the world.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.