Intended for healthcare professionals


Standards of care in research

BMJ 2004; 329 doi: (Published 11 November 2004) Cite this as: BMJ 2004;329:1114
  1. Zulfiqar Bhutta, Husein Lalji Dewraj professor of paediatrics and child health (zulfiqar.bhutta{at}
  1. Aga Khan University, Karachi 74800, Pakistan

    Should reflect local conditions and not the best western standards

    The interpretation of “standard of care” in research has generated a lot of controversy among researchers. It has underpinned much of the debate on the use of placebos in randomised controlled trials such as the one around the zidovudine trials, which were proposed to prevent mother to child transmission of HIV infection in Africa. These trials entailed an evaluation of a short course treatment regimen in comparison with a control population receiving nothing.13 A high profile debate ensued that led to a reconsideration of guidelines on international research as well and spurred initiatives to make low cost antiretroviral treatment available in developing countries.4 5

    In October 2000, the World Medical Association modified the Helsinki declaration to state that “the benefits, risks, burdens and effectiveness of a new method should be tested against those of the best current prophylactic, diagnostic, and therapeutic methods.”6 The Council for International Organizations of Medical Sciences guidelines took the debate a stage further by using the term “established effective therapy” to indicate a degree of consensus and acceptability among health professionals about the nature of treatment 7 But crucially, these guidelines did not specify if established effective therapy applies to a local or global context. Given the paucity of relevant health research in developing countries, aspiring for best standards of care may make research in these countries irrelevant and unsustainable.8

    Total absence of care or health services cannot be considered a suitable control standard. Nor can harmful practices in a dysfunctional system, such as unsafe injections or female circumcision. Whether best standards of care should reflect the best available western care or an international standard of care is undecided. Recommendations and management protocols from the World Health Organization, the closest we have to international standards, do not cover all disorders and circumstances and assume a certain level of performance of health systems.

    Two aspects of the debate need further consideration. Firstly, should the standard be a prevalent local standard or one that is ideal for that setting? Pragmatists regard this debate as one between what can be done versus what ought to be done in a given situation. Secondly, standards of care often refer to specific interventions or drugs that are being used in a trial and not to the overall care in a health system where other factors, such as support of the health system, may be equally important.

    Standard of care therefore depends on the context. Furthermore, established interventions in the developed world may not withstand the scrutiny of cost effectiveness and sustainability in the developing world. For example, in the United States health expenditure per head is several times higher than in Sri Lanka, but does not yield proportionately greater health gains.9 The prevalent standards of care in the United States cannot qualify as gold standards elsewhere on the basis of cost effectiveness alone.

    This debate needs to be resolved in a manner that does not preclude further development of health systems through targeted research. It should permit pragmatic improvement rather than waiting for revolutionary changes in health systems that may never happen. The development of low cost alternative interventions is only possible through such a process.

    Not doing locally relevant research deprives poor populations of the benefits of incremental improvement of care, and this is unethical. If a rigid yardstick had been used for research on oral rehydration therapy—at a time when intravenous rehydration was considered the gold standard—one of the greatest advances of the past century would not have been discovered. Similarly, Kangaroo Mother Care of low birth-weight infants in Colombia and studies of domiciliary or community management of neonatal illness by community health workers in India evaluated innovative methods of caring for high risk infants, which fell below prevalent standards of care.10 11 12 They were designed to research the benefits of low cost feasible alternatives. The existing public health system was the standard for comparison, and this was less than what could be provided in affluent urban settings. Other projects, such as cluster randomised trials or trials of vaccine effectiveness, are designed specifically to provide the kind of evidence that can move programmes forward in local settings.

    The best way forward is to adopt a more flexible and pragmatic approach that allows existing guidelines to be interpreted in the context of the standards and quality of care available in local or comparable public health systems. Better still, the existing Helsinki and Council for International Organizations of Medical Sciences guidelines must suggest a contextual interpretation of the guidelines for standards of care in research.6 7 The current stalemate among ethicists is not acceptable as it could impede the development of low cost alternatives in developing countries.


    • Education and debate p 1179

    • Competing interests None declared.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    View Abstract