Factors affecting representativeness in home-based dementia activity interventions: a literature review

Most persons living with dementia (PLwD) reside in community settings, which has driven the development of a wide range of home-based activity interventions. However, the external validity of these studies remains uncertain because the representativeness of study samples is rarely reported. Assessing representativeness is essential for estimating intervention reach and informing health equity and resource allocation decisions. This review evaluates the extent to which participants in home-based activity trials reflect the population of community-dwelling PLwD and identifies sources of selection bias. A targeted review was conducted using three complementary strategies: analyses of two prior systematic reviews, a PubMed search (2020 to 2025), and a review of relevant follow-up publications. Inclusion was limited to peer-reviewed studies that reported detailed recruitment flowcharts and screening data and had initial recruitment pools exceeding 70 participants. The review revealed that most studies provided limited detail on participant selection and frequently initiated recruitment reporting only after caregiver-care recipient dyads had already been assembled. Among the nine studies included, the mean refusal rate among eligible individuals was 57%. In addition, only 24% of initial referrals, on average, completed the first phase of an intervention. Primary barriers to participation included caregiver refusal, lack of an available caregiver/facilitator, and health-related issues. Based on these observed recruitment flows and pre-selection processes, home-based intervention models are estimated to reach fewer than 20% of the target population. These findings suggest that home-based activity interventions for PLwD are applicable to only a limited segment of the intended population. This "representativeness gap" appears largely attributable to reliance on facilitator-care recipient dyads, which inherently excludes the growing number of PLwD living alone and fails to account for the substantial burdens experienced by primary caregivers. Future research should prioritize the development and evaluation of alternative models, such as technology-assisted and community-supported approaches, to enhance scalability, inclusiveness, and equity.