[Does the study "Complementary medicine in health insurance" provide a basis for scientifically valid conclusions?].

In Switzerland public attention was caught by a health economic analysis of health insurance's reimbursement of complementary medical procedures. Objectives of the study were 1) cost development, 2) health status development of insured citizens, 3) substitutional/additional use of complementary medical procedures. However, several conceptual, methodical and statistical flaws limit the conclusions that can be drawn from this study: The analysis claimed to be based on a large randomized trial. However, the attempt to achieve an unbiased equality in experimental and control group through stratified sampling grossly failed. Effects of complementary medical procedures on health status were inappropriately assessed and only based on 10 patients. The question whether complementary medicine is being used as a substitution or addition to conventional medicine was not even explored (the published result is based on a misleading interpretation of the health insurance's database). Finally, statements on cost developments were not based on estimations, but resulted from methodical and statistical artifacts. Furthermore, the study suffered from several other weaknesses, like a 50% dropout rate, an inadequate description of statistical procedures, an extremely reduced statistical power, and a lack of clear distinction between experimental and retrospective analyses. Therefore, no valid conclusions can be drawn from this study. Copyright 1999 S. Karger GmbH, Freiburg