Using routinely collected data a posteriori for research purposes: How can MSF deal with the ethical issues? Doris Schopper, MD, DrPH Chairperson Ethics Review Board MSF UK Scientific Day 7 June 2007 Research proposals reviewed 2002-5/2007 67 proposals reviewed usually approved in 1-3 rounds 5 studies cancelled; 1 re-designed more expedited reviews recently Variable research activity 4 (2002) 14 (2003) 8 (2004) 6 (2005) 17 (2006) 18 (May 2007) Recently requests for ethical advice/review of research papers before publication 2 in 2006, 5 since Jan 2007 Main ethical issues • Insufficient informed consent procedure • Lack of explicit community involvement • Difficulty to get assurance that benefits will be made available to research participants and community in the medium term • Need to prevent potential abuse of tissue samples provided by research participants • Collaboration with for-profit entities Why worry about ethical issues when analyzing routinely collected data? If data used for programme monitoring, main issue is correct storage and use of data to protect patients and guarantee confidentiality (medical ethics) If data used to test research hypothesis: same ethical requirements as research If formal publication envisaged: ethics approval often requested HIV in prisons in Thailand: ethics review necessary? Diverging views between journal and MSF ERB • Descriptive use of routine data (FUCHIA) • Qualitative information collected purposefully • 3 case vignettes Queries by journal editor and advice by journal ethics committee – Is this research? (divergent views) – Potential exploitation of and harm to (research) participants MSF ERB position – It is not research, but purely descriptive – Routine data collection ethically okay – Potential harm for individuals described in case vignettes: written informed consent had been obtained! Kala-azar outbreak investigation in Somalia Request for ethical advice before submitting • No testing of research hypothesis • Routine monitoring and evaluation activity Question: are persons involved protected from exploitation and harm and is information gathered treated respectfully? Ethical issues to address explicitly: – Were patients informed why blood was taken and that it would be exported for analysis? – What is the fate of blood samples in foreign research laboratory? Interesting findings from evaluation of sleeping sickness programme in the Republic of Congo 2 papers submitted to WHO Bulletin Queries: – safeguards that data are secure, and that links to original records can not be reformed at later date – ownership of data and plans to communicate results to MoH – data are a necessary component of treatment programme and will improve patient care and scientific knowledge WHO Bulletin published after submitting papers to 2 WHO ethics committees: – public health interests outweigh any potential harm to research participants – after permission by MoH of RoC to use the data for this purpose Editorial in same WHO Bulletin puts onus on MSF to do better Conclusions (1): relevant ethical issues • Protection of patient confidentiality (initial collection and storage of information received from the patient) • Individual informed consent necessary or possible to waive • Collaboration and partnership with local/national researchers • Do public health benefits largely outweigh potential harms to the individual (autonomy of patients) and to the community (discrimination, stigmatization)? • Will study community receive benefits of the research? Avoid premeditated research disguised as retrospective analysis of routine data to circumvent ethical review! The example of the FUCHIA data base – 200 days booked with Epicentre to do pooled analysis of HIV cohort data in several MSF projects – Research questions not yet defined, will emerge during analysis What to do? – This goes way beyond improving patient management – As soon as research questions or hypothesis to verify are formulated, MSF should seek ethical review Conclusions (2): When and how to seek ethical approval 1. Analysing routinely collected data to directly improve patient / programme management no ethical review 2. Generalisable research hypothesis to be tested by a posteriori analysis of routinely collected data ethical review ideally at stage of data analysis or draft research paper 3. Grey area “knowledge of general interest” (between testing generalisable research hypothesis and routine monitoring): – Be prepared to answer ethics questions – Journals may request ethics review For the future…? Short “decision framework” to: • raise awareness about ethics issues for authors of papers based on an analysis of routinely collected data; • give direction as to when it is appropriate to submit a proposal to the ethics review board • present a simplified version of the full ERB framework.