A new risk of bias checklist applicable to randomized trials, observational studies, and systematic reviews was developed and validated to be used for systematic reviews focusing on drug adverse events

Jean-Luc Faillie (1, 2) , Pili Ferrer (3, 4) , Amandine Gouverneur (5, 6) , Damien Driot (7, 8) , Shoma Berkemeyer (9) , Xavier Vidal (3, 4) , Maria José Martinez-Zapata (10) , Consuelo Huerta (11) , Xavier Castells (12) , Marietta Rottenkolber (13) , Sven Schmiedl (14) , Mònica Sabaté (3, 4) , Elena Ballarin (3, 4) , Luisa Ibáñez (3, 4)

1 CHRU Montpellier - Centre Hospitalier Régional Universitaire [Montpellier]
2 AIDMP - Aide à la Décision pour une Médecine Personnalisé - Laboratoire de Biostatistique, Epidémiologie et Recherche Clinique - EA 2415
3 UAB - Universitat Autònoma de Barcelona = Autonomous University of Barcelona = Universidad Autónoma de Barcelona
4 Vall d'Hebron University Hospital [Barcelona]
5 BPH - Bordeaux population health
6 CHU Bordeaux
7 CHU Toulouse - Centre Hospitalier Universitaire de Toulouse
8 UT3 - Université Toulouse III - Paul Sabatier
9 RUB - Ruhr University Bochum = Ruhr-Universität Bochum
10 CIBERESP - Centro de Investigación Biomédica en Red de Epidemiología y Salud Pública = Consortium for Biomedical Research of Epidemiology and Public Health
11 AEMPS - Spanish Medicines Agency and Medical Devices [Madrid]
12 UdG - Universitat de Girona
13 Medizinische Klinik und Poliklinik IV
14 Universität Witten/Herdecke

The objective of the study was to develop and validate an adequate tool to evaluate the risk of bias of randomized controlled trials, observational studies, and systematic reviews assessing drug adverse events. STUDY DESIGN AND SETTING: We developed a structured risk of bias checklist applicable to randomized trials, cohort, case-control and nested case-control studies, and systematic reviews focusing on drug safety. Face and content validity was judged by three experienced reviewers. Interrater and intrarater reliability were determined using 20 randomly selected studies, assessed by three other independent reviewers including one performing a 3-week retest. RESULTS: The developed checklist examines eight domains: study design and objectives, selection bias, attrition, adverse events information bias, other information bias, statistical methods to control confounding, other statistical methods, and conflicts of interest. The total number of questions varied from 10 to 32 depending on the study design. Interrater and intrarater agreements were fair with Kendall's W of 0.70 and 0.74, respectively. Median time to complete the checklist was 8.5~minutes. CONCLUSION: The developed checklist showed face and content validity and acceptable reliability to assess the risk of bias for studies analyzing drug adverse events. Hence, it might be considered as a novel useful tool for systematic reviews and meta-analyses focusing on drug safety.

Mots clés: Adverse drug events; Checklist; Meta-analysis; Risk of bias; Study quality; Systematic review

Domaines: Santé publique et épidémiologie; Pharmacologie

Jean-Luc Faillie, Pili Ferrer, Amandine Gouverneur, Damien Driot, Shoma Berkemeyer, et al.. A new risk of bias checklist applicable to randomized trials, observational studies, and systematic reviews was developed and validated to be used for systematic reviews focusing on drug adverse events. Journal of Clinical Epidemiology, 2017, 86, pp.168--175. ⟨10.1016/j.jclinepi.2017.04.023⟩. ⟨hal-01761705⟩