A systematic review of how homeopathy is represented in conventional and CAM peer reviewed journals

BACKGROUND Growing popularity of complementary and alternative medicine (CAM) in the public sector is reflected in the scientific community by an increased number of research articles assessing its therapeutic effects. Some suggest that publication biases occur in mainstream medicine, and may also occur in CAM. Homeopathy is one of the most widespread and most controversial forms of CAM. The purpose of this study was to compare the representation of homeopathic clinical trials published in traditional science and CAM journals. METHODS Literature searches were performed using Medline (PubMed), AMED and Embase computer databases. Search terms included "homeo-pathy, -path, and -pathic" and "clinical" and "trial". All articles published in English over the past 10 years were included. Our search yielded 251 articles overall, of which 46 systematically examined the efficacy of homeopathic treatment. We categorized the overall results of each paper as having either "positive" or "negative" outcomes depending upon the reported effects of homeopathy. We also examined and compared 15 meta-analyses and review articles on homeopathy to ensure our collection of clinical trials was reasonably comprehensive. These articles were found by inserting the term "review" instead of "clinical" and "trial". RESULTS Forty-six peer-reviewed articles published in a total of 23 different journals were compared (26 in CAM journals and 20 in conventional journals). Of those in conventional journals, 69% reported negative findings compared to only 30% in CAM journals. Very few articles were found to be presented in a "negative" tone, and most were presented using "neutral" or unbiased language. CONCLUSION A considerable difference exists between the number of clinical trials showing positive results published in CAM journals compared with traditional journals. We found only 30% of those articles published in CAM journals presented negative findings, whereas over twice that amount were published in traditional journals. These results suggest a publication bias against homeopathy exists in mainstream journals. Conversely, the same type of publication bias does not appear to exist between review and meta-analysis articles published in the two types of journals.