Integrating community perspectives to improve survey completion rates in public health research by refining controversial survey elements.

Abstract

Background: Many factors can impact survey completion rates, including survey length, sensitivity of the topics addressed, and clarity of wording. This study used cognitive interviews (CIs), a methodological tool that can aid in developing and refining elements for multi-faceted assessments, and previous survey response patterns to refine, streamline, and increase response rates of RADx-UP Common Data Elements (CDEs) for survey/questionnaire use.

Methods: Ten previously enrolled CO-CREATE study participants were interviewed between May-June 2023. Interviewees identified CDEs that were "confusing, uncomfortable, and/or not applicable," along with their reasoning. Interview data were analyzed using a rapid qualitative analytic approach, resulting in a summary matrix categorized by language. For further contextualization, CDE response rates were calculated for the 9147 surveys administered during the CO-CREATE study (May 2021-March 2023) and compared against their survey position.

Results: Of the 94 CDEs evaluated in the CIs, 20 (21.3%) were flagged by one or more interviewees. Nine (9.6%) English while fourteen (14.9%) Spanish CDEs were flagged by interviewees, with some overlap. Also, CDE response rates differed according to position in the survey, with lower response rates for questions positioned later in the survey. Following review by the research team and the RADx-UP program, 10 English and 15 Spanish were revised, and seven were removed in both languages in the final survey.

Conclusion: Our findings underscore the importance of integrating community member perspectives to enhance the relevance and clarity of assessment instruments, optimizing the impact of public health research among underrepresented populations.