Send to

Choose Destination
Am J Epidemiol. 2014 Mar 1;179(5):621-32. doi: 10.1093/aje/kwt298. Epub 2013 Dec 22.

Assessing risk prediction models using individual participant data from multiple studies.

Collaborators (128)

Tipping RW, Folsom AR, Couper DJ, Ballantyne CM, Coresh J, Goya Wannamethee S, Morris RW, Kiechl S, Willeit J, Willeit P, Schett G, Ebrahim S, Lawlor DA, Yarnell JW, Gallacher J, Cushman M, Psaty BM, Tracy R, Tybjærg-Hansen A, Price JF, Lee AJ, McLachlan S, Khaw KT, Wareham NJ, Brenner H, Schöttker B, Müller H, Jansson JH, Wennberg P, Salomaa V, Harald K, Jousilahti P, Vartiainen E, Woodward M, D'Agostino RB, Bladbjerg EM, Jørgensen T, Kiyohara Y, Arima H, Doi Y, Ninomiya T, Dekker JM, Nijpels G, Stehouwer CD, Kauhanen J, Salonen JT, Meade TW, Cooper JA, Cushman M, Folsom AR, Psaty BM, Shea S, Döring A, Kuller LH, Grandits G, Gillum RF, Mussolino M, Rimm EB, Hankinson SE, Manson JE, Pai JK, Kirkland S, Shaffer JA, Shimbo D, Bakker SJ, Gansevoort RT, Hillege HL, Amouyel P, Arveiler D, Evans A, Ferrières J, Sattar N, Westendorp RG, Buckley BM, Cantin B, Lamarche B, Barrett-Connor E, Wingard DL, Bettencourt R, Gudnason V, Aspelund T, Sigurdsson G, Thorsson B, Kavousi M, Witteman JC, Hofman A, Franco OH, Howard BV, Zhang Y, Best L, Umans JG, Onat A, Sundström J, Michael Gaziano J, Stampfer M, Ridker PM, Michael Gaziano J, Ridker PM, Marmot M, Clarke R, Collins R, Fletcher A, Brunner E, Shipley M, Kivimäki M, Ridker PM, Buring J, Cook N, Ford I, Shepherd J, Cobbe SM, Robertson M, Walker M, Watson S, Alexander M, Butterworth AS, Di Angelantonio E, Gao P, Haycock P, Kaptoge S, Pennells L, Thompson SG, Walker M, Watson S, White IR, Wood AM, Wormser D, Danesh J.


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.


C index; D measure; coronary heart disease; individual participant data; inverse variance; meta-analysis; risk prediction; weighting

[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for Silverchair Information Systems Icon for PubMed Central
Loading ...
Support Center