Portal de Periódicos da CAPES

Assessing Risk Prediction Models Using Individual Participant Data From Multiple Studies

2013; Oxford University Press; Volume: 179; Issue: 5 Linguagem: Inglês

10.1093/aje/kwt298

1476-6256

Lisa Pennells, Stephen Kaptoge, Ian R. White, Simon G. Thompson, Angela Wood, Robert Tipping, Aaron R. Folsom, David Couper, Christie M. Ballantyne, Josef Coresh, S. Goya Wannamethee, Richard Morris, Stefan Kiechl, Johann Willeit, Peter Willeit, Georg Schett, Shah Ebrahim, Debbie A. Lawlor, J. W. G. Yarnell, John Gallacher, Mary Cushman, Bruce M. Psaty, Russ Tracy, Anne Tybjærg‐Hansen, Jackie F. Price, Amanda Lee, Stela McLachlan, Kay-Tee Khaw, Nicholas J. Wareham, Hermann Brenner, Ben Schöttker, Heiko Müller, Jan‐Håkan Jansson, Patrik Wennberg, Veikko Salomaa, Kennet Harald, Jari Lahti, Erkki Vartiainen, Mark Woodward, Ralph B. D’Agostino, Else‐Marie Bladbjerg, Torben Jørgensen, Yutaka Kiyohara, Hisatomi Arima, Yasufumi Doi, Toshiharu Ninomiya, Joost Dekker, Giel Nijpels, Coen D.A. Stehouwer, Jussi Kauhanen, Jukka T. Salonen, Tom Meade, Jackie A. Cooper, Mary Cushman, Aaron R. Folsom, Bruce M. Psaty, Steven Shea, Angela Döring, Lewis H. Kuller, Greg Grandits, Richard F Gillum, Michael E. Mussolino, Eric B. Rimm, S. E. Hankinson, JoAnn E. Manson, Jennifer K. Pai, Susan Kirkland, Jonathan A. Shaffer, Daichi Shimbo, Stephan J. L. Bakker, Ron T. Gansevoort, Hans L. Hillege, Philippe Amouyel, Dominique Arveiler, Alun Evans, Jean Ferrières, Naveed Sattar, Rudi G. J. Westendorp, Brendan M. Buckley, Bernard Cantin, Benoı̂t Lamarche, Elizabeth Barrett‐Connor, Deborah L. Wingard, Richele Bettencourt, Vilmundur Guðnason, Thor Aspelund, Gunnar Sigurðsson, Bolli Þórsson, Maryam Kavousi, Jacqueline C.M. Witteman, Albert Hofman, Oscar H. Franco, Barbara V. Howard, Ying Zhang, Lyle G. Best, Jason G. Umans, Altan Onat, Johan Sundström, J. Michael Gaziano, Meir J. Stampfer, Paul M. Ridker, J. Michael Gaziano, Paul M. Ridker, Michael Marmot, Robert Clarke, Rory Collins, Astrid Fletcher, Eric J. Brunner, Martin J. Shipley, Mika Kivimäki, Paul M. Ridker, Julie E. Buring, Nancy R. Cook, Ian Ford, James Shepherd, Stuart M. Cobbe, Michele Robertson, Matthew R. Walker, Sarah Watson, M Alexander, Adam S. Butterworth, Emanuele Di Angelantonio, Pei Gao, Philip Haycock, Stephen Kaptoge, Lisa Pennells, Simon G. Thompson, Matthew R. Walker, Sarah Watson, Ian R. White, Angela Wood, David Wormser, John Danesh,

Meta-analysis and systematic reviews

Individual participant time-to-event data from multiple prospective epidemiologic studies enable detailed investigation into the predictive ability of risk models. Here we address the challenges in appropriately combining such information across studies. Methods are exemplified by analyses of log C-reactive protein and conventional risk factors for coronary heart disease in the Emerging Risk Factors Collaboration, a collation of individual data from multiple prospective studies with an average follow-up duration of 9.8 years (dates varied). We derive risk prediction models using Cox proportional hazards regression analysis stratified by study and obtain estimates of risk discrimination, Harrell's concordance index, and Royston's discrimination measure within each study; we then combine the estimates across studies using a weighted meta-analysis. Various weighting approaches are compared and lead us to recommend using the number of events in each study. We also discuss the calculation of measures of reclassification for multiple studies. We further show that comparison of differences in predictive ability across subgroups should be based only on within-study information and that combining measures of risk discrimination from case-control studies and prospective studies is problematic. The concordance index and discrimination measure gave qualitatively similar results throughout. While the concordance index was very heterogeneous between studies, principally because of differing age ranges, the increments in the concordance index from adding log C-reactive protein to conventional risk factors were more homogeneous.