Presentation (CTI)

OBJECTIVES: Health preference research (HPR) increasingly informs patient-centered decision-making, and health technology assessment. However, concerns about data quality are increasing and are expected to grow further due to inattentive responding, ineligible participation, and potential fraud, particularly in online data collection. There is no structured guidance that systematically maps available measures across the HPR study cycle. This study develops a structured framework to guide the prevention, monitoring, and detection of low-quality and potentially fraudulent responses in HPR.
METHODS: A targeted narrative review was conducted to identify strategies for preventing, monitoring and detecting low-quality and potentially fraudulent responses for both general survey research and HPR study cycle. Searches combined PubMed queries, snowballing, and grey literature. Identified approaches were systematically mapped to stages of the HPR study life cycle and synthesized into a structured framework.
RESULTS: Evidence from general survey research and HPR literature converges on three complementary domains: (i) indicators capturing response behavior, identity, and plausibility; (ii) analytical approaches ranging from rule-based screening to likelihood-based diagnostics and models that distinguish preference heterogeneity from noise; and (iii) technical and procedural measures. These approaches are integrated into a structured, HPR phase-based framework spanning (1) design-phase prevention (e.g., trap questions, bot detection, and access controls), (2) monitoring during data collection (e.g., real-time behavioral indicators, soft-blocking, and source comparison), and (3) post hoc detection (e.g., internal validity tests, statistical exclusion rules, and model-based diagnostics).
CONCLUSIONS: The proposed Data Quality Management Framework provides a structured, phase-based approach to organizing measures across the HPR study life cycle. It maps available strategies and clarifies where different measures can be considered during study planning, fieldwork, and analysis. The framework supports transparent documentation and context-sensitive selection of quality assurance measures. It provides a foundation for future research evaluating the relative effectiveness, feasibility, and optimal combination of data quality measures in HPR.