Strategies to Improve Research Participation By Older People With Cognitive Impairment: A Systematic Review.

Abstract

Background and objectives: Older people with cognitive impairment are unrepresented in clinical research. Our objective was to review evidence for strategies to support their research inclusion and participation.

Research design and methods: Systematic review of published reports of inclusion and participation strategies for older people with cognitive impairment in clinical research (PROSPERO CRD42020212092). 5 databases were searched in September 2020, March 2023, and April 2024. Screening, full-text review, and data extraction were independently performed. The risk of bias was assessed using Scottish Intercollegiate Guidelines Network Methodology Checklists. Outcomes were participant characteristics, recruitment and consent processes, retention, experience, involvement of others, adverse events, and other reported outcomes.

Results: Of 4,564 identified sources, 12 studies were included. 9 compared recruitment strategies; 3 examined consent processes. Of 4,208 participants (mean age 78.3 years), 61% were female. Median (interquartile range) monthly recruitment rate was 10.3 (5.6-14.8). Participants had mild cognitive impairment, dementia, and/or delirium, and two-thirds of the studies involved proxies or study partners. Community outreach or population screening had higher recruitment compared with primary care referral and/or screening. Formal capacity and consent methods achieved lower rates of consent compared with informal. A memory and organizational aid increased participants' ability to provide informed consent compared with standard assessment. Few studies reported participants' or recruiters' subjective experience, and no studies reported participant retention or adverse events.

Discussion and implications: Targeted, tailored, and multi-pronged recruitment and consent strategies to support the inclusion of older people with cognitive impairment appear promising. Higher-quality studies are needed to confirm this finding.