Development of a minimum checklist to assess the quality of evidence produced using registry data for the evaluation of medical device safety and performance.

Abstract

Objectives: Medical device registries in Europe report limited information about their structure and methodological characteristics. This hinders their utility for evaluation of medical device safety and performance under the Medical Device Regulation. This study aimed to define a minimum checklist of items necessary for regulators to assess the quality of evidence produced using registry data for the evaluation of medical device safety and performance.

Design: A three-round Delphi panel.

Setting: A task within the Coordinating Research and Evidence for Medical Devices project.

Participants: 101 experts in the medical device community (healthcare professionals, methodologists, registry experts, regulators, and assessors from notified bodies) were invited.

Interventions: Based on a literature review and expert advice, 27 items relating to the quality of registry data and the analysis of medical device safety and performance were selected. In round 1, participants selected which items were required for a minimum checklist. They could also propose new items. Items selected by ≥70% of participants indicated consensus. Remaining items were discussed in round 2, resulting in a final checklist that was ranked by participants for importance (round 3).

Main outcome measures: Consensus of items to be included in the minimum checklist.

Results: 51 experts participated in round 1, achieving consensus on 18 (67%) items and suggesting 12 items. After discussion in round 2, 5 additional items were selected, resulting in a final set of 15 data quality items and 8 data analysis items. The most important items were 'completeness of procedures' (data quality) and 'definition of outcome analyzed'" (quality of analysis).

Conclusions: Reporting all items from the minimum checklist will facilitate judgment of the utility of registry data to evaluate medical devices during post-market surveillance.