Location bias in controlled clinical trials of complementary/alternative therapies

To systematically investigate location bias of controlled clinical trials i n complementary/alternative medicine (CAM). Methods: Literature searches we re performed to identify systematic reviews and meta-analyses, which were u sed to retrieve controlled clinical trials. Trials were categorised by whet her they appeared in CAM-journals or mainstream medical (MM)-journals, and by their direction of outcome, methodological quality, and sample size. Res ults: 351 trials were analysed. A predominence of positive trials was seen in non-impact factor CAM- and MM-journals, 58/78 (74%) and 76/102 (75%) res pectively, and also in low impact factor CAM- and MM-journals. In high impa ct factor MM-journals there were equal numbers of positive and negative tri als, a distribution significantly (P < 0.05) different from all other journ al categories. Quality scores were significantly lower for positive than ne gative trials in non-impact factor CAM-journals (P < 0.02). A similar trend was seen in low-impact factor CAM journals, but not to a level of signific ance (P = 0.06). There were no significant differences between quality scor es of positive and negative trials published in MM-journals, except for hig h impact factor journals, in which positive trials had significantly lower scores than negative trials (P = 0.048). There was no difference between po sitive and negative trials in any category in terms of sample size. Conclus ion: More positive than negative trials of complementary therapies are publ ished, except in high-impact factor MM-journals. In non-impact factor CAM-j ournals positive studies were of poorer methodological quality than the cor responding negative studies. This was not the case in MM-journals which pub lished on a wider range of therapies, except in those with high impact fact ors. Thus location of trials in terms of journal type and impact factor sho uld be taken into account when the literature on complementary therapies is being examined. (C) 2000 Elsevier Science Inc. All rights reserved.