TY - JOUR T1 - Partial derivatives meta-analysis: pooled analyses when individual participant data cannot be shared JF - bioRxiv DO - 10.1101/038893 SP - 038893 AU - Hieab HH Adams AU - Hadie Adams AU - Lenore J Launer AU - Sudha Seshadri AU - Reinhold Schmidt AU - Joshua C Bis AU - Stephanie Debette AU - Paul A Nyquist AU - Jeroen Van der Grond AU - Thomas H Mosley, Jr AU - Jingyun Yang AU - Alexander Teumer AU - Saima Hilal AU - Gennady V Roshchupkin AU - Joanna M Wardlaw AU - Claudia L Satizabal AU - Edith Hofer AU - Ganesh Chauhan AU - Albert Smith AU - Lisa R Yanek AU - Sven J Van der Lee AU - Stella Trompet AU - Vincent Chouraki AU - Konstantinos A Arfanakis AU - James T Becker AU - Wiro J Niessen AU - Anton JM de Craen AU - Fabrice F Crivello AU - Li An Lin AU - Debra A Fleischman AU - Tien Yin Wong AU - Oscar H Franco AU - Katharina Wittfeld AU - J Wouter Jukema AU - Philip L De Jager AU - Albert Hofman AU - Charles DeCarli AU - Dimitris Rizopoulos AU - WT Longstreth, Jr AU - Bernard M Mazoyer AU - Vilmundar Gudnason AU - David A Bennett AU - Ian J Deary AU - M Kamran Ikram AU - Hans J Grabe AU - Myriam Fornage AU - Cornelia M Van Duijn AU - Meike W Vernooij AU - M Arfan Ikram AU - on behalf of the HD-READY Consortium Y1 - 2016/01/01 UR - http://biorxiv.org/content/early/2016/02/07/038893.abstract N2 - Joint analysis of data from multiple studies in collaborative efforts strengthens scientific evidence, with the gold standard approach being the pooling of individual participant data (IPD). However, sharing IPD often has legal, ethical, and logistic constraints for sensitive or high-dimensional data, such as in clinical trials, observational studies, and large-scale omics studies. Therefore, meta-analysis of study-level effect estimates is routinely done, but this compromises on statistical power, accuracy, and flexibility. Here we propose a novel meta-analytical approach, named partial derivatives meta-analysis, that is mathematically equivalent to using IPD, yet only requires the sharing of aggregate data. It not only yields identical results as pooled IPD analyses, but also allows post-hoc adjustments for covariates and stratification without the need for site-specific re-analysis. Thus, in case that IPD cannot be shared, partial derivatives meta-analysis still produces gold standard results, which can be used to better inform guidelines and policies on clinical practice. ER -