Reviewer Bias

Abstract

To the Editors: Several forms of publication bias distort the medical literature (1). To test the hypothesis that "reviewer bias" exists, a MEDLINE search of publications on transcutaneous electrical nerve stimulation (TENS) between 1983 and 1990 was carried out to identify investigators who had reported original data on this subject. Thirty-three investigators whose data were either clearly for or against the clinical effectiveness of TENS were selected. Investigators whose studies produced inconclusive results were excluded. The selected investigators were sent a fictitious "research paper" on TENS. It deliberately included strong points and flaws but reported a "positive" result. The manuscript