Integrating complementary medicine?

on balance, the evidence for clinical effectiveness is still highly inconclusive (e.g. Patel et al, 1989; Ter Riet et al, 1990). Yet the burden of proof in CM should be no different than for mainstream medicine (Smith, 1995). And what is the evidence on safety? Even though promoted as safe, CM is certainly not free of adverse reactions (e.g. Ernst, 1994; Ernst, 1995a). At present we simply do not know how frequently adverse events and complications occur. Our own data suggest that up to 15 % of users of CM do experience side-effects (Abbot et al, 1996). Considering this lack of fundamental information on safety and effectiveness, a risk-benefit analysis of CM is impossible. And what about costs? CM may be cheap but cost-effectiveness, by definition, must remain uncertain in the absence of firm evidence for effectiveness. With some degree of variation, the above statements apply not only to acupuncture but to virtually all complementary therapies (Ernst, 1996a). By integrating largely unproven remedies into medical routine, a disfavour might be done to the science of medicine. Integration may also be against the interest of CM itself. More than once I have