Comparing sampling methods to engage residents of underrepresented communities to identify research priorities / Barbara J. Turner [and four others].

Format:

Book

Author/Creator:

Turner, Barbara J., author.

Language:

English

Subjects (All):

Outcome assessment (Medical care).

Physical Description:

1 online resource (1 PDF file (121 pages)) : illustrations

Place of Publication:

Washington, DC : PCORI, Patient-Centered Outcomes Research Institute, 2019.

Summary:

BACKGROUND: For patient-centered outcomes research to improve health care decision-making and ultimately health status, patients and caregivers need to serve as research partners. Yet evidence-based community engagement methods are lacking, especially with hard-to-reach populations. This study evaluated sampling methods to recruit and engage residents of rural, Hispanic-majority communities to identify research priorities. OBJECTIVES: (Aim 1) To compare purposive sampling (PS) and respondent-driven sampling (RDS) to recruit patients and caregivers to serve as research participants (RPs) who then brainstormed and prioritized ideas about services, programs, and research to improve outcomes for persons with chronic noncancer pain (CNCP); and (Aim 2) to publish a practical guide to start and sustain hard-to-reach community partnerships. METHODS: Two county-based community advisory boards (cCABs) in 2 similar rural, Hispanic-majority counties with poor health outcomes directed implementation of PS and RDS. We predicted that PS would yield greater recruitment and retention and more research ideas. However, low recruitment using PS led the cCAB in 1 county to add purposive sampling plus convenience sampling (P + C) using a local recruiter. In the other county, the cCAB preferred a different chain-referral method--snowball sampling (SS)--to RDS because it was less complex and demanding for participants. In each county, 3 groups of consenting RPs, both caregivers and people with CNCP, attended 3 structured meetings to generate and rate ideas for research (total 12 meetings). Thematic analysis was used to categorize ideas. In a mixed methods analysis, RPs recruited by P + C vs SS were compared on demographics; participation (number consenting, number attending); number of ideas; and Likert scale ratings of ideas on importance to improve living with chronic pain and feasibility to implement. For external validity, a representative sample of Hispanic adults from 5 southwestern states was surveyed online to evaluate the RPs' ideas. RESULTS: Using P + C, 62 RPs were recruited; 36 (58%) attended the first meeting and 26 (42%) attended all 3 meetings. XXXX RPs were recruited using SS; 52 (95%) attended the first meeting and 36 (65%) attended all 3. SS yielded more Hispanic RPs than P + C (87% vs 73%; P = .049) and more recruits who were disabled (47% vs 10%; P < .001). SS RPs generated 115 ideas vs 101 ideas for P + C RPs. For 7 of 8 categories of ideas, ratings on importance and feasibility did not differ between SS and P + C RPs (P > .05). The 5-state online sample represented 10.5 million primarily urban Hispanics. Ratings on importance to improve living with chronic pain by this sample were all lower than the RPs' ratings (P < .002), but the order of relative ranking on the Likert scale and in a MaxDiff preference analysis was similar to that of the RPs. To fulfill Aim 2, the UP AHEAD Research Handbook was produced as a guide for investigators seeking to partner with hard-to-reach communities. CONCLUSIONS: Snowball sampling yielded greater and more sustained study participation--especially by Hispanics--than P + C sampling, but the SS method recruited more RPs who were disabled. RPs from both groups had similar ideas and ratings. A largely urban 5-state online Hispanic sample rated the rural RPs' ideas lower but, again, the highest priority ideas were similar. The advisory boards from hard-to-reach communities served a vital role in directing engagement. LIMITATIONS AND SUBPOPULATIONS: The project's methods were adjusted to align with real-world circumstances. Lessons learned may be most relevant to rural, Hispanic-majority communities.

Notes:

Description based on publisher supplied metadata and other sources.