Mapping the characteristics of meta-analyses of pharmacy services: a systematic review

Suboptimal meta-analyses with misleading conclusions are frequently published in the health areas, and they can compromise decision making in clinical practice.

This systematic review aimed to map the characteristics of published meta-analyses of pharmacy services and their association with the study conclusions.

We searched electronic databases (PubMed, Scopus, and Web of Science) to identify published meta-analyses of pharmacy services up to January 2019. Components of meta-analyses were extracted (i.e. studies' metadata; methods used in the systematic review; description of the statistical model used for the meta-analysis; main results; conflict of interest and funding source). The methodological quality was evaluated using the R-AMSTAR tool.

A total of 85 meta-analyses were included, with 2016 as the median publication year. Overall, the methodological quality of meta-analyses of pharmacy services was considered suboptimal. Only one-third of authors registered a protocol; complete search strategy and raw data were provided by 55.3% and 9.4% of studies, respectively. Evidence strength (GRADE) was evaluated in only 19.2% of studies. PRISMA and Cochrane recommendations were stated to be followed in 60% and 27.4% of articles, respectively. Around half of studies performed sensitivity analysis, however, the prediction interval was presented by only one meta-analysis. Studies that favoured the pharmacists' interventions poorly discussed the methodological quality and heterogeneity of primary trials.

Poor conduction and reporting were observed in meta-analyses of pharmacy services, especially in those that favoured the pharmacist's interventions. Reproducibility and transparency should be rigorously ensured by journal editors and peer-reviewers.