Original Abstracts from the 2018 European Meeting of ISMPP

Abstract

Objective: We evaluated compliance of recent narrative systematic reviews with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidance. Research design and methods: We searched EMBASE and MEDLINE databases on 21-September-2017, for English-language records with a print publication date of 01-June-2017, for "systematic review" (in the title field) and terms relating to drug therapy (in the subject field). Case studies, conference reports, letters, surveys, errata, editorials, and meta-analyses were excluded. A manual screen excluded protocols and reports providing statistical analysis. Articles were scored for fulfilment of PRISMA checklist items for the objectives, methods (information sources, search, study selection), and results (study selection, study characteristics), and whether they reference PRISMA. Results: Of the 99 abstracts identified, 46 (46.5%) were selected for analysis. Reasons for exclusion were: not drug related (n ¼ 35), statistical analysis conducted (n ¼ 10), protocol (n ¼ 7), and not in English (n ¼ 1). Twenty-seven (58.7%) publications did not fully adhere to our set of PRISMA items. More than 82% of publications appropriately reported on objectives, information sources, and study selection (methods) items, whereas 76.1% and 50.0% reported study selection (results) and search items, respectively. Publications citing PRISMA (n ¼ 28; 60.8%) tended to report on more items with the exception of search criteria. Conclusions: Just over half of these recent publications described as systematic reviews did not follow PRISMA criteria, despite referencing them. These findings suggest a need for improvement in performing systematic reviews and/or reporting how they were conducted. Such improvements may lead to greater confidence in the findings of systematic reviews.