OBJECTIVES: Research studies used to inform public policy should be rigorous, well-conducted and transparent. Primary data and analyses can be assessed once the study is complete, but at that point raw data quality cannot be improved. The UK valuation of EQ-5D-5L, which is urgently required to inform policy including health technology assessments submitted to NICE (National Institute for Health and Care Excellence,) is using in-built independent data quality assessments during primary data collection to make ongoing assessments of quality.

METHODS: Twelve hundred interviews are being undertaken using the time trade-off elicitation technique, using the EQ-VT (EuroQol Valuation Technology) interviewing system (version 2.1) with a representative sample of the UK population. The study benefits from public involvement and input from EuroQol’s EQ-VT support team, an independent Quality Control team and a Steering Group. The EQ-VT support team undertake weekly data monitoring at the aggregate and interviewer level throughout data collection to monitor data quality and identify issues. The Quality Control team independently assess the data at pilot stage, 20%, 40% and 60% data collection, and assess the final dataset and analyses. At each of these stages the decision to continue and/or refine data collection is made by the Steering Group, comprised of representatives from NICE, National Health Service (NHS) England, Department of Health and Social Care (DHSC), international experts and EuroQol scientific leaders.

RESULTS: Following a successful pilot, over 60% data collection has been undertaken, with no substantial modifications to data collection required. All interim analyses have indicated that data quality is acceptable. Data collection is ongoing.

CONCLUSIONS: Independent assessments of primary data during its collection can be undertaken to identify any issues arising promptly. This enables amendments to be made during data collection, where changes are possible, rather than finding the data has unresolvable issues once data collection is complete.