A Systematic Review and Recommendation for Reporting of Surrogate Endpoint Evaluation Using Meta-analyses.
Background: Meta-analysis of randomized controlled trials (RCTs) has been widely conducted for the evaluation of surrogate endpoints in oncology, but little attention has been given to the adequacy of reporting and interpretation. This review evaluated the reporting quality of published meta-analyses on surrogacy evaluation and developed recommendations for future reporting. Methods: We searched PubMed through August 2017 to identify studies that evaluated surrogate endpoints using the meta-analyses of RCTs in oncology. Both individual patient data (IPD) and aggregate data (AD) meta-analyses were included for the review. Results: Eighty meta-analyses were identified: 22 used IPD and 58 used AD from multiple RCTs. We observed variability and reporting deficiencies in both IPD and AD meta-analyses, especially on reporting of trial selection, endpoint definition, study and patient characteristics for included RCTs, and important statistical methods and results. Based on these findings, we proposed a checklist and recommendations to improve completeness, consistency, and transparency of reports of meta-analytic surrogacy evaluation. We highlighted key aspects of the design and analysis of surrogate endpoints and presented explanations and rationale why these items should be clearly reported in surrogacy evaluation. Conclusions: Our reporting of surrogate endpoint evaluation using meta-analyses (ReSEEM) guidelines and recommendations will improve the quality in reporting and facilitate the interpretation and reproducibility of meta-analytic surrogacy evaluation. Also, they should help promote greater methodological consistency and could also serve as an evaluation tool in the peer review process for assessing surrogacy research.
Xie, W; Halabi, S; Tierney, JF; Sydes, MR; Collette, L; Dignam, JJ; Buyse, M; Sweeney, CJ; Regan, MM
Volume / Issue
Start / End Page
Pubmed Central ID
Electronic International Standard Serial Number (EISSN)
Digital Object Identifier (DOI)