Bias in psychiatric case–control studies

Abstract

BackgroundCase–control studies are vulnerable to selection and information biases which may generate misleading findings.AimsTo assess the quality of methodological reporting of case–control studies published in general psychiatric journals.MethodAll the case–control studies published over a 2-year period in the six general psychiatric journals with impact factors of more than 3 were assessed by a group of psychiatrists with training in epidemiology using a structured assessment devised for the purpose. The measured study quality was compared across type of exposure and journal.ResultsThe reporting of methods in the 408 identified papers was generally poor, with basic information about recruitment of participants often absent. Reduction of selection bias was described best in the ‘pencil and paper’ studies and worst in the genetic studies. Neuroimaging studies reported the most safeguards against information bias. Measurement of exposure was reported least well in studies determining the exposure with a biological test.ConclusionsPoor reporting of recruitment strategies threatens the validity of reported results and reduces the generalisability of studies.