RT Journal Article
SR Electronic
T1 Partial derivatives meta-analysis: pooled analyses when individual participant data cannot be shared
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 038893
DO 10.1101/038893
A1 Adams, Hieab HH
A1 Adams, Hadie
A1 Launer, Lenore J
A1 Seshadri, Sudha
A1 Schmidt, Reinhold
A1 Bis, Joshua C
A1 Debette, Stephanie
A1 Nyquist, Paul A
A1 Van der Grond, Jeroen
A1 Mosley, Thomas H
A1 Yang, Jingyun
A1 Teumer, Alexander
A1 Hilal, Saima
A1 Roshchupkin, Gennady V
A1 Wardlaw, Joanna M
A1 Satizabal, Claudia L
A1 Hofer, Edith
A1 Chauhan, Ganesh
A1 Smith, Albert
A1 Yanek, Lisa R
A1 Van der Lee, Sven J
A1 Trompet, Stella
A1 Chouraki, Vincent
A1 Arfanakis, Konstantinos A
A1 Becker, James T
A1 Niessen, Wiro J
A1 de Craen, Anton JM
A1 Crivello, Fabrice F
A1 Lin, Li An
A1 Fleischman, Debra A
A1 Wong, Tien Yin
A1 Franco, Oscar H
A1 Wittfeld, Katharina
A1 Jukema, J Wouter
A1 De Jager, Philip L
A1 Hofman, Albert
A1 DeCarli, Charles
A1 Rizopoulos, Dimitris
A1 Longstreth, WT
A1 Mazoyer, Bernard M
A1 Gudnason, Vilmundar
A1 Bennett, David A
A1 Deary, Ian J
A1 Ikram, M Kamran
A1 Grabe, Hans J
A1 Fornage, Myriam
A1 Van Duijn, Cornelia M
A1 Vernooij, Meike W
A1 Ikram, M Arfan
A1 ,
YR 2016
UL http://biorxiv.org/content/early/2016/02/07/038893.abstract
AB 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.