Critical quality evaluation of network meta-analyses in dental care

Abstract

Objective: Network meta-analysis (NMA) is a new method to compare the effects of multiple treatments in dental care. We examined whether published NMA follow the key methodological reporting recommendations, the PRISMA Extension Statement.

Design: Collection of published systematic reviews with NMA in dental care DATA SOURCES: MEDLINE, EMBASE, and Web of Science searched from inception to May 1, 2017.

Study selection: All NMA published in dental journals comparing the clinical efficacy of three or more interventions with and without closed loops.

Data extraction and synthesis: NMA reporting was assessed according to the PRISMA Extension checklist. We assessed the reporting of general components (27 items) and key NMA methodological components (5 new items, S 1-5) of the systematic review process of NMA in dental care.

Results: Among 85 identified articles, 21 NMA (16 used the Bayesian statistical approach, and 5 the frequentist approach) were included. Overall, among 21 articles, 15 (71%) reports of NMA in dental care did not describe in detail the network regarding S1-5 based on the PRISMA Extension. Problem areas include exploring the geometry of the network (S1), assessment of insistency (S2) in the method section, presentation of network structure (S3), summary of network geometry (S4), exploration for inconsistency (S5) in the result section, risk of bias across studies, protocol registrations, and additional analysis.

Conclusions: The quality of NMA reporting was low in the dental field. An NMA with inadequate reporting reduces the end-user's confidence on the quality of the reported results. The PRISMA Extension guideline is relatively new and should be used more extensively to improve reporting practices in the field of dental care.

Keywords: Dental care; Meta-analysis; Methodological quality; Network meta-analysis; Quality evaluation.