Effectiveness of interventions to improve rates of intravenous thrombolysis using behaviour change wheel functions: a systematic review and meta-analysis

BACKGROUND: Despite being one of the few evidence-based treatments for acute ischemic stroke, intravenous thrombolysis has low implementation rates—mainly due to a narrow therapeutic window and the health system changes required to deliver it within the recommended time. This systematic review and meta-analyses explores the differential effectiveness of intervention strategies aimed at improving the rates of intravenous thrombolysis based on the number and type of behaviour change wheel functions employed. METHOD: The following databases were searched: MEDLINE, EMBASE, PsycINFO, CINAHL and SCOPUS. Multiple authors independently completed study selection and extraction of data. The review included studies that investigated the effects of intervention strategies aimed at improving the rates of intravenous thrombolysis and/or onset-to-needle, onset-to-door and door-to-needle time for thrombolysis in patients with acute ischemic stroke. Interventions were coded according to the behaviour change wheel nomenclature. Study quality was assessed using the QualSyst scoring system for quantitative research methodologies. Random effects meta-analyses were used to examine effectiveness of interventions based on the behaviour change wheel model in improving rates of thrombolysis, while meta-regression was used to examine the association between the number of behaviour change wheel intervention strategies and intervention effectiveness. RESULTS: Results from 77 studies were included. Five behaviour change wheel interventions, ‘Education’, ‘Persuasion’, ‘Training’, ‘Environmental restructuring’ and ‘Enablement’, were found to be employed among the included studies. Effects were similar across all intervention approaches regardless of type or number of behaviour change wheel-based strategies employed. High heterogeneity (I(2) > 75%) was observed for all the pooled analyses. Publication bias was also identified. CONCLUSION: There was no evidence for preferring one type of behaviour change intervention strategy, nor for including multiple strategies in improving thrombolysis rates. However, the study results should be interpreted with caution, as they display high heterogeneity and publication bias.