Protocol for a Systematic Review and Individual Participant Data Meta-Analysis of Randomized Trials of Screening for Atrial Fibrillation to Prevent Stroke

Introduction Atrial fibrillation (AF) is a common cause of stroke. Timely diagnosis of AF and treatment with oral anticoagulation (OAC) can prevent up to two-thirds of AF-related strokes. Ambulatory electrocardiographic (ECG) monitoring can identify undiagnosed AF in at-risk individuals, but the impact of population-based ECG screening on stroke is uncertain, as ongoing and published randomized controlled trials (RCTs) have generally been underpowered for stroke.Methods and analysis The AF-SCREEN Collaboration, with support from AFFECT-EU, have begun a systematic review and individual participant data meta-analysis of RCTs evaluating ECG screening for AF. The primary outcome is stroke. Secondary outcomes include AF detection, OAC prescription, hospitalization, mortality, and bleeding.After developing a common data dictionary, anonymized data will be collated from individual trials into a central database. We will assess risk of bias using the Cochrane Collaboration tool, and overall quality of evidence with the Grading of Recommendations Assessment, Development and Evaluation approach.We will pool data using random effects models. Prespecified subgroup and multilevel meta-regression analyses will explore heterogeneity. We will perform prespecified trial sequential meta-analyses of published trials to determine when the optimal information size has been reached, and account for unpublished trials using the SAMURAI approach.Impact and Dissemination Individual participant data meta-analysis will generate adequate power to assess the risks and benefits of AF screening. Meta-regression will permit exploration of the specific patient, screening methodology, and health system factors that influence outcomes.

Automated summary

Atrial fibrillation is a common cause of stroke. Timely diagnosis and treatment with oral anticoagulation can prevent many AF-related strokes. However, the impact of population-based ECG screening on stroke is uncertain as current RCTs lack sufficient power.