Trials number, funding support, and intervention type associated with IPDMA data retrieval: a cross-sectional study

Objectives: This study aimed to investigate the predictors for high data retrieval and the reporting of individual participant data meta analyses (IPDMAs). Study Design and Setting: We searched EMBASE, MEDLINE, and the Cochrane Library for articles pertaining to IPDMA from 2011 to 2019. Only IPDMA assessing treatment effects, including randomized controlled trials (RCTs), were included. Adherence to the PRISMA-IPD guideline was checked. Results: A total of 210 IPDMA covering 18 diseases were sampled; 80 (38.1%) and 123 (58.6%) of the IPDMA retrieved IPD from all and >80% RCTs, respectively. Non-Cochrane reviews, IPDMA on nonpharmacological interventions, analyses of smaller numbers of RCTs, and having funding supports were predictors of complete IPD retrieval. Owners of RCTs had an increased probability of obtaining IPD. Only 4.3% described the eligibility criteria covering all the PICO components, 11.0% reported the methods for assessing risk of bias across studies, 11.4% mentioned the IPD integrity, and 9.0% presented detailed results of syntheses. Conclusion: IPD retrieval and reporting was not satisfactory among the published IPDMA. Encouraging RCT owners to conduct or join in the IPDMA is a potential strategy to maximize the IPD retrieval. IPDMA are suggested to adhere to the PRISMA-IPD guideline during reporting. (c) 2020 Elsevier Inc. All rights reserved. (c) 2020 Elsevier Inc. All rights reserved.

Automated summary

The study identified predictors for high data retrieval and reporting in IPDMA, including non-Cochrane reviews, IPDMA on nonpharmacological interventions, analyses of smaller numbers of RCTs, and having funding support. Only a small percentage of studies met all the PICO criteria, reported methods for assessing bias risk, mentioned IPD integrity, and presented detailed synthesis results, indicating room for improvement in reporting quality.