Summary of findings for the main comparison.

Occupational therapy compared with standard care for care home residents with stroke
Patient or population: care home residents who have had a stroke Settings: care homes (nursing and residential homes) Intervention: occupational therapy Comparison: standard care
Outcomes	Illustrative comparative risks (95% CI)		Relative effect (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Standard care	Occupational therapy
Function in ADL at the end of scheduled follow‐up (Barthel ADL Index score) Barthel ADL Index score 6‐month follow‐up Analysis 1.1	The mean Barthel across the control group was 8	The mean Barthel in the intervention group was 2 points higher (mean Barthel of 10) (95% CI ‐0.11 to 0.90)		118 (1 study)	⊕⊕⊝⊝ low	Clustering design effect was accounted for
Global poor outcome Death or a drop in Barthel ADL score 6‐month follow‐up Analysis 1.2	Medium‐risk population		OR 0.34 (0.11 to 1.01)	118 (1 study)	⊕⊕⊝⊝ low	Clustering design effect was accounted for
	759 per 1000	516 per 1000 (258 to 759)
Function in ADL at the end of intervention (Barthel ADL Index score) Barthel ADL Index score 3‐month follow‐up Analysis 1.3	The mean Barthel across the control group was 8	The mean Barthel score in the intervention group was 3 points higher (mean Barthel of 11) (95% CI ‐0.03 to 0.99)		118 (1 study)	⊕⊕⊝⊝ low	Clustering design effect was accounted for
Death Number of deaths from any cause at 6‐month follow‐up Analysis 1.4	Medium‐risk population		OR (0.09 to 0.98)	118 (1 study)	⊕⊕⊝⊝ low	Clustering design effect was accounted for
	242 per 1000	485 per 1000 (28 to 239)
Quality of life	See comment	See comment	Not estimable	0	See comment	Included study did not measure 'quality of life' as an outcome
Mobility Rivermead Mobility Index (RMI) score 6‐month follow‐up Analysis 1.5	The mean RMI score across the control group was 4.5	The mean RMI score in the intervention group was 0.5 higher (mean RMI of 5) (95% CI ‐0.36 to 0.64)		118 (1 study)	⊕⊕⊝⊝ low	Clustering design effect was accounted for
Mood	See comment	See comment	Not estimable	0	See comment	Included study did not measure 'mood' as an outcome
Global cognition	See comment	See comment	Not estimable	0	See comment	Included study did not measure 'global cognition' as an outcome
Adverse events	See comment	See comment	Not estimable	0	See comment	Adverse events were not reported in the included study
Satisfaction with care	See comment	See comment	Not estimable	0	See comment	Included study did not measure 'satisfaction with care' as an outcome
Health economic outcomes	See comment	See comment	Not estimable	0	See comment	Included study did not measure 'health economic outcomes' as an outcome
CI: confidence interval; OR: odds ratio
GRADE Working Group grades of evidence: High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.