Meta-Analysis and Literature Review

Publication Bias

Searches of databases can yield many studies. However, these databases include only published studies. Such searches are unlikely to yield a representative sample because studies that show a positive result (usually in favor of a new or standard treatment) are more likely to be published. This selective publication is called publication bias. Consider the case of the publication status of studies on antidepressants. Based on studies registered with the FDA, 97% of the positive studies are published versus only 12% of the negative ones. Furthermore, when the nonpublished studies are not included, the positive effects of individual drugs increase between 11% and 69%.

One reason for publication bias is that drug manufacturers are not generally interested in publishing negative studies. Also, journal editors favor positive studies because these are the ones that make the headlines. To ameliorate the effect of publication bias, a serious effort should be made to identify unpublished studies. This is much easier now due to improved communication between researchers and by registries in which all the studies of a certain disease or treatment are reported. In some medical areas, the exclusion of studies conducted in non-English-speaking countries can increase publication bias.

The National Institutes of Health maintains a registry of all the studies it supports, and the U.S. Food and Drug Administration keeps a registry and database in which drug companies must register all trials they sponsor. Registries of published and unpublished trials supported by pharmaceutical companies are also available (e.g., GlaxoSmithKline's Clinical Study Register). The Cochrane collaboration keeps records of systematic reviews and meta-analyses of many diseases and interventions.

...