Economic evaluations of outpatient stroke rehabilitation models: a systematic review of evidence, methods, and gaps.

Background: Stroke imposes an ongoing burden on patients and health systems, making outpatient rehabilitation essential. The economic efficiency of alternative outpatient models remains uncertain, complicating resource allocation. We systematically reviewed and critically appraised full economic evaluations of outpatient stroke rehabilitation to synthesise evidence on reported cost-effectiveness, assess how costs and outcomes have been measured, and identify key methodological limitations and evidence gaps. Methods: We conducted a PRISMA-guided systematic review (PROSPERO: CRD420251151991) of full economic evaluations of outpatient stroke rehabilitation. Embase, Emcare, Medline, CINAHL, Cochrane Library, Web of Science, and PubMed were searched, with publication coverage including all available studies published prior to September 2025. Eligible studies included adults with ischemic or haemorrhagic stroke undergoing outpatient rehabilitation and reporting full economic evaluations. We excluded inpatient-only rehabilitation, transient ischaemic attack, partial evaluations, reviews, and non-English publications. Two reviewers independently screened, extracted, and appraised studies. Risk of bias (RoB 2, ROBINS-I) and methodological quality (Drummond checklist) were assessed, with data narratively synthesised by rehabilitation model and evaluation type. Findings: The search yielded 1800 records, of which 27 studies met inclusion criteria. Across studies, community-based, home-based, and early supported discharge rehabilitation models frequently demonstrated favourable cost-effectiveness when compared with inpatient or outpatient clinic-based rehabilitation. Study settings were global, with over one-third from the UK. Most were randomised controlled trials (56%), with cost-effectiveness analyses predominating (63%). Community- and home-based rehabilitation were most common (each 30%), followed by outpatient clinic-based (22%), early supported discharge (11%), and telerehabilitation (7%). Outcomes clustered into health-related quality of life (41%), functional recovery (37%), and process measures (22%). Methodological quality was mostly moderate (16/27), with seven high and four low. RCTs showed some reporting concerns, while most non-randomised studies carried serious risk of bias. Interpretation: Evidence from existing economic evaluations suggests that shifting stroke rehabilitation toward outpatient and community-based rehabilitation models may demonstrate more favourable cost-effectiveness profiles in certain contexts, while evidence on telerehabilitation remains limited. However, these findings should be interpreted with caution due to substantial heterogeneity across study designs, comparators, outcomes, and economic evaluation methods. To better inform decision-making, future studies should adopt standardised outcome measures, extend follow-up, and generate context-specific economic evidence, particularly in low- and middle-income countries. Funding: There is no funding source for this study.