1. ¹Department of Psychiatry, University of Illinois at Chicago, Chicago, IL, USA
2. ²Maryland Psychiatric Research Center, Baltimore, MD, USA
3. ³Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, CA, USA

Correspondence: Dr JM Davis, Department of Psychiatry, University of Illinois at Chicago, 1601 W Taylor Street, Chicago, IL 60612, USA. Tel: +1 312 413 4570; Fax: +1 312 996 7658; E-mail: jmdavis@uic.edu

Received 3 January 2007; Revised 15 May 2007; Accepted 24 May 2007; Published online 4 July 2007.

Abstract

This study presents a meta-analysis of the influence of several potentially biasing factors (eg industry support, extrapyramidal side effects) on efficacy of studies comparing second-generation antipsychotic (SGA) with first-generation antipsychotic (FGA) medications. We used the dataset from our previously published meta-analysis of 124 randomized controlled trials (RCTs) comparing SGAs with FGAs, to evaluate whether certain possible biases could influence the actual outcome on the total score of the Positive and Negative Syndrome Scale (PANSS), Brief Psychiatric Rating Scale (BPRS), and Clinical Global Impressions (CGI) scores. Industry sponsorship was determined by contact with authors or publication statement. We calculated whether (1) industry sponsorship, (2) study quality, (3) extrapyramidal symptoms (EPS) properties, or (4) prophylactic antiparkinsonian medications influenced SGA vs FGA efficacy for each drug and averaged overall by two Hedges and Olkin-based meta-analyses. The analysis found that none of the factors was significantly associated with a particular outcome. While industry-sponsored articles may conclude their medication to be more favorable than that of a competitor in an RCT, we found that the observed efficacy was not influenced by sponsorship. Many attribute the finding that SGAs appears to be more efficacious than FGAs to be a result of EPS-decreasing efficacy (or its measurement). We were unable to confirm that the drug's EPS properties or antiparkinsonian management altered actual efficacy.