Dyad interventions for health-related quality of life, activity, and participation after stroke: a systematic review.

Nine of the 17 included studies had high or unclear risk of bias on three or more Cochrane items, most commonly blinding of participants and personnel (16 studies) and allocation concealment (11 studies). No meta-analysis was possible because of heterogeneity in intervention content, dose, delivery modality, study design, and outcome measures. Activity and participation outcomes were rarely measured for the caregiver, so the review cannot say whether dyad interventions affect the caregiver's daily life or social engagement. Three of the eight joint psychosocial studies were pilots without inferential statistics, so the positive signal in that category rests on only two trials. Most studies were conducted in high-income countries in Asia and Europe; only two were in upper middle-income countries, and race or ethnicity was reported in just two studies (both majority White). The search string was built around 'stroke AND social support AND intervention' because dyad interventions are not indexed in MeSH, so relevant studies may have been missed.

The authors declare no conflicts of interest. No funding source is named.

Do not read the large effect sizes from joint psychosocial interventions as proof these approaches work. Only two of seven studies in that category reported significant effects, two were pilots without inferential statistics, and one did not report between-group differences at all. The effect sizes come from just two trials, and the review as a whole found mixed results across all three intervention types.