Meta-Analysis Examples

Several examples of meta-analysis show very little difference between treatments. One example performed by Madsen et al. studied the effects of acupuncture on patients with various pain conditions. 13 trials were performed on 3,025 patients. There was not much difference between the sham and the real acupuncture, but a larger difference with the group that received no acupuncture. This is perhaps due to the fact that these clinical trials were not blind for the patients or the practitioners.

Dr. Colquhoun has a different take on the meta-analysis done by Vickers et al. The study shows that there is little difference in the outcomes of the treatments, but the way the study was published and the way the media reported the findings changed the perception of the results.

Another example of this bias occurred when a few statistically significant findings were picked out of a study called the CACTUS trial. The positive findings were minimal yet “the result was trumpeted as a success for acupuncture.”

Dr. Stephen Barrett, co-founder of the National Council against Health Fraud, is another professional that opposes acupuncture. This council has concluded that:

• Acupuncture is an unproven modality of treatment.

• It’s theory and practiced are based on primitive and fanciful concepts of health and disease

Dr. Andrew L. Avins, a research scientist for Kaiser Permanente, and epidemiologist at the University of California, San Francisco, believes the findings from the meta-analysis by Vickers et al, does not clearly rule out the placebo effect. He says, “The fact that the effectiveness rate was much higher than treatment as usual but only slightly higher than the false treatment suggests that most of the benefit associated with acupuncture is indeed attributable to the placebo effect.” Believing in a placebo may not necessarily be a bad thing if a patient is feeling