Early feedback elicitation and downstream participation behaviours: a retrospective observational analysis of study completion and disengagement

BACKGROUND: Many health research studies and clinical trials face threats to external validity and low statistical power because they are unable to recruit and retain a sufficient number of participants. Identifying study design improvements using participant feedback is an important avenue through which studies seek to address these challenges. However, the returns on this approach have been limited because experience measures are rarely linked to actual participation behaviour. We examine whether feedback collected at the beginning of a research study can provide a meaningful signal about future participation as measured by study completion, withdrawal, and communication unsubscription. METHODS: We use observational data from a large health research program (N = 35,838; 77% from groups historically underrepresented in biomedical research, UBR) in which feedback was collected across the participant experience. We correlate feedback responses (e.g., indicating satisfaction with the experience) with study completion and active disengagement 6 months later and use Artificial Intelligence (AI) processing of free-text comments to surface specific topics and sentiment. FINDINGS: Ratings given immediately after onboarding are associated with an 8.89 percentage-point (pp) higher study completion rate ('Good' vs 'Fair') and a 23.22 pp higher active disengagement rate ('Poor' vs 'Fair') within 6 months. Associations for feedback collected later are smaller (up to 4.17 pp for completion). These differences hold after adjusting for participant characteristics (age, income, educational attainment, sex at birth, UBR on race/ethnicity, and UBR on rurality). We find no statistically significant effect of the prompt itself. AI-extracted topics and sentiment agree with human reviewers 86% and 81% of the time, respectively. INTERPRETATION: Early, light-touch feedback can meaningfully differentiate participants by their likelihood of completing or disengaging from a study, creating opportunities to target support and address flagged concerns while the study is still in progress. Feedback collected at the end of a study is a weaker signal of future behaviour. Our approach, including AI processing of free-text comments, is readily portable to other studies and feedback elicitations. Whether feedback-informed changes improve retention is an important question for future work. FUNDING: National Institutes of Health.

Authors

Institutions

Publication Details

Journal
EBioMedicine
Published
2026-11-18
DOI
https://doi.org/10.1016/j.ebiom.2026.106522
Primary Topic
Ethics in Clinical Research
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Early feedback elicitation and downstream participation behaviours: a retrospective observational analysis of study completion and disengagement

Janna Ter Meer, Jacqueline Kathleen Kueper, Julia Moore Vogel, Giorgio Quer et al.
4 links
EBioMedicine
Ethics in Clinical Research
article

Early feedback elicitation and downstream participation behaviours: a retrospective observational analysis of study completion and disengagement

Janna Ter Meer, Jacqueline Kathleen Kueper, Julia Moore Vogel, Giorgio Quer, Ting-Yang Hung
article en

Abstract

BACKGROUND: Many health research studies and clinical trials face threats to external validity and low statistical power because they are unable to recruit and retain a sufficient number of participants. Identifying study design improvements using participant feedback is an important avenue through which studies seek to address these challenges. However, the returns on this approach have been limited because experience measures are rarely linked to actual participation behaviour. We examine whether feedback collected at the beginning of a research study can provide a meaningful signal about future participation as measured by study completion, withdrawal, and communication unsubscription. METHODS: We use observational data from a large health research program (N = 35,838; 77% from groups historically underrepresented in biomedical research, UBR) in which feedback was collected across the participant experience. We correlate feedback responses (e.g., indicating satisfaction with the experience) with study completion and active disengagement 6 months later and use Artificial Intelligence (AI) processing of free-text comments to surface specific topics and sentiment. FINDINGS: Ratings given immediately after onboarding are associated with an 8.89 percentage-point (pp) higher study completion rate ('Good' vs 'Fair') and a 23.22 pp higher active disengagement rate ('Poor' vs 'Fair') within 6 months. Associations for feedback collected later are smaller (up to 4.17 pp for completion). These differences hold after adjusting for participant characteristics (age, income, educational attainment, sex at birth, UBR on race/ethnicity, and UBR on rurality). We find no statistically significant effect of the prompt itself. AI-extracted topics and sentiment agree with human reviewers 86% and 81% of the time, respectively. INTERPRETATION: Early, light-touch feedback can meaningfully differentiate participants by their likelihood of completing or disengaging from a study, creating opportunities to target support and address flagged concerns while the study is still in progress. Feedback collected at the end of a study is a weaker signal of future behaviour. Our approach, including AI processing of free-text comments, is readily portable to other studies and feedback elicitations. Whether feedback-informed changes improve retention is an important question for future work. FUNDING: National Institutes of Health.

EBioMedicineVol. 134
Scripps Research Institute (US)
Foundation for the National Institutes of Health
Good health and well-being
Ethics in Clinical Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon