Strategies to increase participation of adults aged ≥65 years in observational studies and RCTs: a systematic review
Primary care is increasingly managing older people (≥65 years) with multiple conditions and polypharmacy. However, older people tend to be excluded from observational studies and randomised controlled trials (RCTs) that could inform their care. Identifying effective recruitment and retention strategies will improve the inclusion of older people in such studies. The objective of this systematic review was to evaluate the effectiveness of strategies to improve recruitment and retention of older people in observational studies and RCTs.
Randomised and quasi-randomised trials were included that compared recruitment and retention strategies aimed at adults in observational studies and RCTs, in any setting. Numbers/proportions of participants aged ≥65 years recruited, responded or retained were defined as the outcomes. 10 databases were searched with no language restrictions from the earliest available date to January 2016 including MEDLINE, EMBASE, the Cochrane Library and AgeInfo. Two of the reviewing team independently screened each abstract and full text, with disagreements discussed and a third reviewer consulted if necessary. Trials without data for adults aged ≥65 years were discarded. Risk of bias was assessed using the Cochrane Risk of Bias tool. Results were synthesised narratively.
We included 33 trials: 7 trials of recruitment and/or retention to a trial, 4 trials of recruitment to an observational study, and 22 trials of response within an observational study. 11 trials were quasi-randomised. Minimum age was ≥65 years in 18 trials; the remaining 15 trials presented results stratified by age ≥65. The majority of trials were set in primary care (12 trials) or the community (12 trials). Trials showed wide variability in types of strategies. The most frequently evaluated strategies were different modes of approach (8 trials), incentives (5 trials), questionnaire length or content (5 trials), mode of reply (3 trials) and advance notification (3 trials). Most randomised trials had a low risk of bias although reporting of blinding was often unclear. The synthesis including updated searches is not yet complete. However, opt-out approaches seemed to result in increased recruitment in adults aged 65-74 and ≥70, recruiting to a primary care trial and primary care observational study, respectively. Trials using advance notification (telephone call or letter) found an increase in survey response compared to no notification in adults aged ≥70, 70-84 and ≥65 years. Willingness to participate in a hypothetical trial was increased in women aged ≥65 screened using tablet computers in a quasi-randomised trial in primary care.
Some recruitment strategies evaluated so far show promise, particularly opt-out approaches and advance notification. More research is needed to (i) confirm the value of these strategies, (ii) explore the use of new technology and electronic strategies, and (iii) investigate retention strategies further, to improve the participation of older people in research.