Assessing the methodological and reporting quality of clinical systematic reviews and meta-analyses in paediatric urology: can practices on contemporary highest levels of evidence be built?

Introduction: Systematic reviews and meta-analyses provide a comprehensive summary of research studies and are used to assess clinical evidence, form policy and construct guidelines. This is pertinent to childhood surgery with issues of consent and condition prevalence. The aims of this study were to evaluate the methodological and reporting quality of these reviews and to identify how these reviews might guide clinical practice amongst those conditions most commonly encountered and managed by practicing paediatric urologists.

Methods: A systematic search of the English literature was performed to identify systematic reviews and meta-analyses focusing on clinical paediatric urology (1/1/1992-1/12/2018) to include common paediatric urological conditions managed by paediatric urology residents/fellows. To these reviews, Assessing the Methodological Quality of Systematic Reviews (AMSTAR)-2 and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) scores were applied. Univariate linear regression and descriptive statistical methods were performed.

Results: From an initial literature review of 1723 articles, 227 were included in the analysis. Inter-reviewer agreement was high amongst 3 independent reviewers (κ = 0.92). Eighty-four percent of systematic reviews and meta-analyses were published since 2009 following publication of the PRISMA guidelines. The overall impact factor was 3.38 (0.83-17.58), with adherence to AMSTAR-2 criteria 48.46% and PRISMA criteria 70.1%. From a methodological perspective, 15% of reviews were of moderate quality, 65% were of low quality and 20% reviews were of critically low quality, with none found to have good quality reporting.

Conclusions: Despite the continued increase of systematic reviews and meta-analyses in paediatric urology from which many guidelines are based, a significant number of reviews contain poor methodology and, to a lesser extent, poor reporting quality. Journals should consider having specific 'a priori' criteria based on checklists before publication of manuscripts to ensure the highest possible reporting quality.