Research on the Quality Evaluation Indicators of Real-World Data in the Review of Anti-Tumor Drugs for Chinese Regulatory Frameworks
摘要
This study addresses the quality of real-world data (RWD) in oncology and proposes a comprehensive evaluation system for assessing the quality of RWD in the review of anti-tumor drugs. The proposed system provides a solid theoretical foundation and methodological tools to support drug review decisions and offers insights for further enhancing data quality evaluation in regulatory processes.
MethodsPreliminary indicators for evaluating the quality of real-world data in anti-tumor drug review were identified based on established domestic and international theories and practices. Expert consultations were conducted to collect feedback from professionals across various fields, further refining the evaluation indicators. The Analytic Hierarchy Process (AHP) was then applied to develop an expert judgment matrix, assigning weights to the indicators according to their relative importance. The final weight values for each indicator were subsequently determined.
ResultsThis study developed a real-world data quality evaluation system for anti-tumor drug review, comprising 2 primary indicators, 8 secondary indicators, and 32 tertiary indicators. Weight analysis revealed that data accuracy, source credibility, and population representativeness were the most critical secondary indicators. Among the tertiary indicators, oncology-specific data standards, third-party ethical evaluations, and clarity in inclusion/exclusion criteria and screening processes were identified as the most significant factors. These findings suggest that regulatory agencies should emphasize these aspects more during the quality evaluation process.
ConclusionsBased on the key indicators identified in the evaluation system, it is recommended that drug regulatory agencies focus on the following aspects when assessing the quality of real-world data in anti-tumor drug reviews: (1) clear definition of core data elements and critical variables; (2) evaluation of data quality in alignment with research objectives; and (3) ensuring data security and consistency to enhance overall data quality.