Assessing the Quality of Database Search Strategies Utilized by Meta-analyses That Evaluate Infectious Disease Drug Therapy.

Abstract

Meta-analyses are increasingly influencing clinical practice, but significant methodological flaws have been reported. The purpose of this study was to evaluate the quality of search strategies utilized by anti-infective meta-analyses. The Embase database was searched for meta-analyses evaluating anti-infective drug therapy; 103 of 268 identified citations met inclusion criteria and were evaluated. A total of 80.6% of meta-analyses used search terms, and an average of 4.3 databases (Medline, 98.1%; Cochrane, 93.2%; Embase, 76.7%) were searched to identify relevant articles for inclusion. The majority of meta-analyses used a quality assessment tool (84.5%) and reported positive results (59.2%). The average impact factor of journals publishing meta-analyses was 5.7 ± 3.4. The number of resources searched was associated with the impact factor ( P = .0013). The majority of anti-infective meta-analyses used rigorous search strategies to identify all relevant studies for evaluation. This finding is inconsistent with reports in other therapeutic areas that have questioned the quality of meta-analyses, and it may increase confidence in anti-infective meta-analyses.