Ricerche di medicina complementare in Lombardia
Circular instead of hierarchical: Methodological basics in the evaluations of CAM therapy effects and other complex interventions
Dr Harald Walach
Director of research, Institute of Environmental Medicine, University Hospital
of Freiburg
Abstract
The randomised controlled trial (RCT) is considered the gold standard of evidence based medicine research. This implies a hierarchical model of methodology, with outcomes and observational studies at the base and RCTs at the top. This model is questioned here, especially for complementary and alternative medicine (CAM), but also for other complex interventions. It is argued that many presuppositions implicitly or explicitly made by RCTs are not met with CAM interventions, mostly the lack of preference and equipoise, but also the importance of specific effects. Moreover, RCTs have great problems securing external and model validity, and hence are liable to produce false negative results. The most important point, however, is that they run into an efficacy paradox: If interventions produce strong non-specific effects, as many CAM interventions do, but small specific effects, then such interventions may be discarded as not efficacious, although their absolute effectiveness might be higher than studies with large specific, but smaller overall effects. It is therefore argued that one should adopt a circular research strategy which focuses on appropriate methods for specific questions and understands that every method has a shortcoming that needs balancing by other methods. Thus the merits and problems of diverse research methods are complementing each other in a circular way rather than supporting one gold standard method which then becomes the arbiter. This is not only a situation which has to be respected for CAM research, but also for researching conventional interventions, if they are complex.
Key Words: Methodology, Evidence Based Medicine, Randomised Controlled Trial, Complementary and Alternative Medicine, Evaluation
Edited by Aldo Campana,