Covariate-adjusted measures of discrimination for survival data.

Pubmed ID: 25530064

Pubmed Central ID: PMC4666552

Journal: Biometrical journal. Biometrische Zeitschrift

Publication Date: July 1, 2015

MeSH Terms: Humans, Male, Female, Cardiovascular Diseases, Risk Factors, Middle Aged, Survival Analysis, Clinical Trials as Topic, Analysis of Variance, Biometry, Discriminant Analysis

Grants: RG/08/014/24067, MR/K013351/1, UL1 TR000062, MR/L003120/1, G0700463, MC_UU_12013/5, UL1 TR001450, G19/35, G0100222, G8802774, G0902037, G1000616, RG/07/008/23674, RG/13/13/30194

Authors: White IR, Rapsomaniki E, Tipping RW, Davis BR, Simpson LM, Ballantyne CM, Folsom AR, Coresh J, Shaw JE, Atkins R, Zimmet PZ, Barr EL, Knuiman MW, Wannamethee SG, Morris RW, Willeit J, Willeit P, Santer P, Kiechl S, Wald N, Ebrahim S, Lawlor DA, Gallacher J, Yarnell JW, Ben-Shlomo Y, Casiglia E, Tikhonoff V, Sutherland SE, Nietert PJ, Keil JE, Bachman DL, Psaty BM, Cushman M, Nordestgaard BG, Tybjærg-Hansen A, Frikke-Schmidt R, Giampaoli S, Palmieri L, Panico S, Pilotto L, Vanuzzo D, Simons LA, Friedlander Y, McCallum J, Price JF, McLachlan S, Taylor JO, Guralnik JM, Wallace RB, Kohout FJ, Cornoni-Huntley JC, Guralnik JM, Blazer DG, Guralnik JM, Phillips CL, Phillips CL, Guralnik JM, Wareham NJ, Khaw KT, Brenner H, Schöttker B, Müller HT, Rothenbacher D, Nissinen A, Donfrancesco C, Giampaoli S, Harald K, Jousilahti PR, Vartiainen E, Salomaa V, D' Agostino RB, Wolf PA, Vasan RS, Daimon M, Oizumi T, Kayama T, Kato T, Chetrit A, Dankner R, Lubin F, Welin L, Svärdsudd K, Eriksson H, Lappas G, Lissner L, Mehlig K, Björkelund C, Nagel D, Kiyohara Y, Arima H, Ninomiya T, Hata J, Rodriguez B, Dekker JM, Nijpels G, Stehouwer CD, Iso H, Kitamura A, Yamagishi K, Noda H, Goldbourt U, Kauhanen J, Salonen JT, Tuomainen TP, DeStavola BL, Blokstra A, Verschuren WM, Cushman M, de Boer IH, Folsom AR, Psaty BM, Koenig W, Meisinger C, Peters A, Verschuren WM, Bueno-de-Mesquita HB, Blokstra A, Rosengren A, Wilhelmsen L, Lappas G, Kuller LH, Grandits G, Cooper JA, Bauer KA, Davidson KW, Kirkland S, Shaffer JA, Shimbo D, Kitamura A, Iso H, Sato S, Dullaart RP, Bakker SJ, Gansevoort RT, Ducimetiere P, Amouyel P, Arveiler D, Evans A, Ferrières J, Schulte H, Assmann G, Jukema JW, Westendorp RG, Sattar N, Cantin B, Lamarche B, Després JP, Barrett-Connor E, Wingard DL, Daniels LB, Gudnason V, Aspelund T, Trevisan M, Hofman A, Franco OH, Tunstall-Pedoe H, Tavendale R, Lowe GD, Woodward M, Howard WJ, Howard BV, Zhang Y, Best LG, Umans J, Ben-Shlomo Y, Davey-Smith G, Onat A, Nakagawa H, Sakurai M, Nakamura K, Morikawa Y, Njølstad I, Mathiesen EB, Wilsgaard T, Sundström J, Gaziano JM, Ridker PM, Marmot M, Clarke R, Collins R, Fletcher A, Brunner E, Shipley M, Kivimaki M, Ridker PM, Buring J, Rifai N, Cook N, Ford I, Robertson M, Ibañez Marín A, Feskens EJ, Geleijnse JM, Bolton T, Burgess S, Butterworth AS, di Angelantonio E, Gao P, Harshfield E, Kaptoge S, Pennells L, Peters S, Spackman S, Thompson S, Walker M, White I, Willeit P, Wood A, Danesh J

Cite As: White IR, Rapsomaniki E, Emerging Risk Factors Collaboration. Covariate-adjusted measures of discrimination for survival data. Biom J 2015 Jul;57(4):592-613. Epub 2014 Dec 20.

Studies:

Abstract

MOTIVATION: Discrimination statistics describe the ability of a survival model to assign higher risks to individuals who experience earlier events: examples are Harrell's C-index and Royston and Sauerbrei's D, which we call the D-index. Prognostic covariates whose distributions are controlled by the study design (e.g. age and sex) influence discrimination and can make it difficult to compare model discrimination between studies. Although covariate adjustment is a standard procedure for quantifying disease-risk factor associations, there are no covariate adjustment methods for discrimination statistics in censored survival data. OBJECTIVE: To develop extensions of the C-index and D-index that describe the prognostic ability of a model adjusted for one or more covariate(s). METHOD: We define a covariate-adjusted C-index and D-index for censored survival data, propose several estimators, and investigate their performance in simulation studies and in data from a large individual participant data meta-analysis, the Emerging Risk Factors Collaboration. RESULTS: The proposed methods perform well in simulations. In the Emerging Risk Factors Collaboration data, the age-adjusted C-index and D-index were substantially smaller than unadjusted values. The study-specific standard deviation of baseline age was strongly associated with the unadjusted C-index and D-index but not significantly associated with the age-adjusted indices. CONCLUSIONS: The proposed estimators improve meta-analysis comparisons, are easy to implement and give a more meaningful clinical interpretation.