Comparing Three Ways of Involving Patients with Low Back Pain in Setting Research Priorities—SMARTER [Internet]

Excerpt

Background: By participating in priority-setting activities in research, patients help ensure that questions of greatest importance are recognized and incorporated into future research agendas. Surveys, focus groups, and online crowdsourcing are used increasingly to obtain patient input into research, yet we know little about how they compare in the generation of research priorities or participant experience.

Objectives: SMARTER (Study of Methods for Assessing Research Topic Elicitation and pRioritization) had 2 objectives: (1) Evaluate how patient engagement methods compare in producing research priorities and participant satisfaction; and (2) determine if patient demographics and disease severity influence participation in research activities.

Methods: In phase 1, we sought to compare priorities for low back pain (LBP) research via a survey among 2 different populations: the Back pain Outcomes using Longitudinal Data (BOLD) registry of individuals aged 65 and older with LBP and the Amazon Mechanical Turk (MTurk) crowdsourcing platform. We adapted LBP research topics presented in the survey from a list generated by primary care clinicians and researchers in 2009. In phase 2, we randomly assigned phase 1 respondents from the BOLD population to participate in 1 of 3 interactive prioritization activities: online crowd voting, in-person focus groups using nominal group technique, and 2 rounds of a mailed survey (Delphi). To assess quality of experience, phase 2 respondents completed a second survey; a subset was subsequently interviewed. We used descriptive statistics to characterize the rankings produced by each method and rank-ordered logistic regression models to identify associations of the ranked priority topics with baseline characteristics of activity participants. We analyzed responses to the evaluation using a mixed methods approach: We tabulated responses to Likert scale questions and used thematic analysis of interviews to explore patient knowledge about the activity and perceptions of experience.

Results: In phase 1, a total of 924 BOLD registry participants from Henry Ford Health System in Michigan and Kaiser Permanente Northern California completed the initial survey and identified 5 LBP research topics as most important. On the MTurk platform, 350 people with LBP completed the prioritization activity. Prioritization results for the 2 groups were similar and included causes of LBP, self-care strategies, tests for diagnosing LBP, strategies for putting research results into practice, and physical health programs. In phase 2, 38 people participated in crowd voting, 39 in focus groups, and 74 in the Delphi survey. Phase 2 participants were predominantly White (83% crowd voting, 65% focus groups, 83% Delphi), non-Hispanic (100% crowd voting, 97% focus groups, 94% Delphi), and college educated (78% crowd voting, 60% focus group, 54% Delphi). Women outnumbered men in the focus groups and Delphi survey (71% and 61%, respectively); otherwise, demographics among groups were similar. Phase 2 activities generated similar lists of research priorities, including causes of LBP, improving physician–patient communication, and self-care strategies. In phase 2, 124 participants completed an evaluation survey (26 crowd voting, 35 focus group, 63 Delphi). Of these, we interviewed 31 participants about their experience (11 crowd voting, 10 focus group, 10 Delphi). Focus group participants rated their experience highest, in both post-activity evaluation and follow-up interviews as compared with other activities.

Conclusions: This study assessed how different methods perform in generating replicable research prioritization lists and how participants perceive the quality of the engagement experience. Results demonstrate that common methods for research prioritization yield similar priorities but differ on participant perceptions of experience. Such comparative studies are rare but important in understanding methods to involve patients in research.

Study Limitations: Our study focused on assessing prioritization methods among older adults with LBP and thus may not be generalizable for other health conditions or populations. Further, BOLD participants were patients in integrated health care systems with LBP lasting more than 6 months, and priorities for these participants may differ from those with acute or new onset LBP. Finally, we used preference-based randomization to address potential barriers for participation (eg, travel, internet access) and recognize that preference for a given approach may be associated with other patient characteristics that could affect participant prioritization.