Director, Global HEOR Astellas Pharma, Inc. Crescent Springs, United States
Background: Real-world evidence (RWE) is gaining increased attention by health authorities for evaluation of effectiveness of drugs and vaccines. While documents such as Good Pharmacoepidemiology Practice (GPP) have provided important guidance for the conduct of RWE studies, guidance on the evaluation of the quality of real-world data (RWD) sources is limited.
Objectives: To identify key elements of RWD quality that could be used in a proposed quality management document supporting audits of RWD used for regulatory decision-making. Secondarily, to develop a practical list of considerations that could be used by scientists to help determine whether particular data would likely be deemed sufficiently trustworthy by regulators such that it could properly be used for regulatory decision-making.
Methods: A targeted literature of PubMed, MEDLINE, Embase, and the Cochrane Database of Systematic Reviews (1995-2021), as well as a review of publicly available industry guidance documents was conducted to identify publications related to RWD quality, identify common themes and elements required for high quality RWD. Focused interviews with RWD/RWE experts were conducted to identify potential gaps in the literature assessment, as well to understand the needs of regulatory decision makers. A survey of data providers was conducted to understand common practices used by RWD providers and the extent to which they share data quality metrics with researchers. The literature review and feedback received formed the basis for an exercise to develop an initial list of considerations under a set of common conceptual pillars. This document was then made available for public review and feedback for 60 days
Results: The targeted literature review returned 40 articles/guidance documents for consideration. Interviews were conducted with 8 external industry leaders. Over 18 data providers participated in the survey. Public review submissions were received from 15 organizations. The outcome of this process is a Considerations document that focuses on 5 key pillars of data quality: relevance, accrual, provenance, completeness, and accuracy. Under each pillar, the document defines specific considerations, such as available documentation, data handling procedures, or adherence to an applicable industry standard, as well as the purpose of each consideration.
Conclusions: Given the increased use of RWE and the explosive growth in RWD sources to generate such evidence, there is a strong need to develop systematic approaches for demonstrating the quality of data to regulators. This comprehensive list of data considerations should be a useful guide for data providers, researchers, and regulators in helping to determine whether RWD are sufficiently reliable for use in regulatory decision-making.
