The Declaration of Helsinki, periodically updated by the World Medical Association (WMA), contains specific guidelines for clinical trial design to protect enrolled subjects and to ensure that research yields meaningful data (1). The WMA policy clearly states that the risk associated with a trial must be low and there should be a potential prospective benefit (1). The policy further states that for safety monitoring and to show a potential benefit, the experimental approach has to be tested against a control group treated according to the best established clinical standard. In an amendment, the policy grants that in the absence of a proven care, placebo-control groups might be considered an option, but nowhere is it mentioned that interventional protocols are allowable without a control group (1). A suitable control group enables scientific conclusions to be appropriately drawn from the research. In the absence of a control group, findings that could be construed as beneficial may influence medical practice without justification and, therefore, be a potential hazard to the population. In addition, the lack of a control group is hazardous to enrolled subjects, as potential detrimental effects cannot be detected. Nowhere in the declaration is it mentioned that patients can serve as their own controls. It is therefore not clear why investigators design interventional research studies without a control group, thus disregarding study subject and medical population safety. Additionally, independent study evaluation committees (such as Institutional Review Boards) who supervise the implementation of clinical trials to ensure that patient safety and scientific integrity are maintained, should be aware that interventional trials need a control group comprised of a group of patients treated according to the best established clinical standard. The PROGIS study reported by Chan et al. in Transplantation (2) was an interventional trial in kidney transplantation without a control group. It could be argued that previous randomized prospective trials comparing enteric-coated mycophenolate sodium (EC-MPS) and mycophenolate mofetil (MMF) have shown therapeutic equivalence between the two formulations (3, 4) and therefore, a switch should be safe in this study. However, the safety of patients with the gastrointestinal (GI) symptoms being treated in this trial is a separate issue. The hypothesis that GI side effects in kidney transplant patients can be treated by switching to EC-MPS has never been proven, and only speculative data exist. The limited existent data might be sufficient to justify a randomized trial to test this hypothesis (1), but a trial without a control group places enrolled patients at risk because of the absence of safety monitoring and invalidates any conclusions derived (1). With potentially self-limiting conditions like GI symptoms, the absence of a suitable control group renders observations meaningless because the symptoms could have resolved as rapidly with standard treatment. The study gives some impression that there is a control group; the investigators followed another group of patients that did not have symptoms or require intervention. This is certainly not a control group for an interventional trial and creates confusion for the interpretation of the results. The appropriate control group would include patients with the same symptoms treated with the standard approach (1). In the absence of such a control group the concluding statement of the paper, “In conclusion, these preliminary results suggest that converting MMF-treated renal transplant patients with GI complaints to EC-MPS significantly reduces their symptom burden and improves their (health-related quality of life), functioning and psychological well-being,” cannot be justified. Herwig-Ulf Meier-Kriesche Jesse D. Schold University of Florida College of Medicine Gainesville, FL Henrik Ekberg Lund University Malmo, Sweden
Read full abstract
Jan 1, 2008
Markus Hohenwarter + 1
Open Access
