Quantifying predictive accuracy in survival models
Abstract
For time-to-event outcomes in medical research, survival models are the most appropriate to use. Unlike logistic regression models, quantifying the predictive accuracy of these models is not a trivial task. We present the classes of concordance (C) statistics and R 2 statistics often used to assess the predictive ability of these models. The discussion focuses on Harrell's C, Kent and O'Quigley's R 2, and Royston and Sauerbrei's R 2. We present similarities and differences between the statistics, discuss the software options from the most widely used statistical analysis packages, and give a practical example using the Worcester Heart Attack Study dataset.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Harrell's version of the c-statistic appears most widespread and recommended for time-to-event outcomes.
“Although validation studies do not commonly report the definition and estimation method of presented c-statistics, Harrell’s version21 appears to be most widespread and recommended.22”
From A framework for meta-analysis of prediction model studies with binary and time-to-event outcomes (2019) · cites this paper as [22]
For time-to-event outcomes, concordance is not uniquely defined and several variations of the c-statistic have been proposed that adopt different strategies to account for ties or censoring and may lead to different values for c.
“For prediction models with a time-to-event outcome, concordance is not uniquely defined and several variations of the c-statistic have been proposed.22,23 These variations typically adopt different strategies to account for ties or censoring, and may thus lead to different values for c.”
From A framework for meta-analysis of prediction model studies with binary and time-to-event outcomes (2019) · cites this paper as [22]
Cites
Applied Survival Analysis: Regression Modeling of Time-to-Event Data
Hosmer, D. W., Lemeshow, S. A., & May, S. J. (2008) · Cited by 3,030 · see citing works
Uno, H., Cai, T., Pencina, M. J., D'Agostino, R. B., & Wei, L. J. (2011) · Statistics in Medicine · doi:10.1002/sim.4154 · Cited by 1,535 · see citing works
Concordance probability and discriminatory power in proportional hazards regression
Gönen, M., & Heller, G. (2005) · Biometrika · doi:10.1093/biomet/92.4.965 · Cited by 685 · see citing works
Lee, K. L., & Mark, D. B. (1996) · Cited by 485 · see citing works
A new measure of prognostic separation in survival data
Royston, P., & Sauerbrei, W. (2004) · Statistics in Medicine · doi:10.1002/sim.1621 · Cited by 450 · see citing works
Explained Variation for Survival Models
Royston, P. (2006) · The Stata Journal Promoting communications on statistics and Stata · doi:10.1177/1536867x0600600105 · Cited by 319 · see citing works
Estimation of time‐dependent area under the ROC curve for long‐term risk prediction
Chambless, L. E., & Diao, G. (2005) · Statistics in Medicine · doi:10.1002/sim.2299 · Cited by 309 · see citing works
Goldberg, R. J. (1986) · JAMA · doi:10.1001/jama.255.20.2774 · Cited by 218 · see citing works
Recent Changes in Attack and Survival Rates of Acute Myocardial Infarction (1975 Through 1981)
Goldberg, R. J. (1986) · JAMA · doi:10.1001/jama.1986.03370200076031 · Cited by 158 · see citing works
Measures of dependence for censored survival data
Kent, J. T., & O’Quigley, J. (1988) · Biometrika · doi:10.1093/biomet/75.3.525 · Cited by 155 · see citing works
Quantifying discrimination of Framingham risk functions with different survival C statistics
Pencina, M. J., D’Agostino, R. B., & Song, L. (2012) · Statistics in Medicine · doi:10.1002/sim.4508 · Cited by 95 · see citing works
Choodari‐Oskooei, B., Royston, P. J., & Parmar, M. (2011) · Statistics in Medicine · doi:10.1002/sim.4242 · Cited by 91 · see citing works
Estimating Harrell's Optimism on Predictive Indices Using Bootstrap Samples
Miao, Y., Kirby, A. C., & Boscardin, W. J. (2013) · Cited by 27 · see citing works
Choodari‐Oskooei, B., Royston, P. J., & Parmar, M. (2012) · Statistics in Medicine · doi:10.1002/sim.5460 · Cited by 25 · see citing works
Heinzl, H. (2000) · Computer Methods and Programs in Biomedicine · doi:10.1016/s0169-2607(00)00073-0 · Cited by 11 · see citing works
Harrell, F., Jr. (1996) · doi:10.1002/(sici)1097-0258(19960229)15:4
Flagged first, then by citations
Cited by
Sorted by citations
A framework for meta-analysis of prediction model studies with binary and time-to-event outcomes
Debray, T. P., Damen, J. A., Riley, R. D., Snell, K., Reitsma, J. B., Hooft, L., Collins, G. S., & Moons, K. G. (2019) · Statistical Methods in Medical Research · doi:10.1177/0962280218785504 · Cited by 278 · see citing works
“Although validation studies do not commonly report the definition and estimation method of presented c-statistics, Harrell’s version(21) appears to be most widespread and recommended.(22) Sometimes, discrimination is measured using Royston and Sauerbrei’s D index, usually referred to as the ‘‘D statistic’’.”
BackgroundMethod
- Authors
- Lirette, S. T., & Aban, I.
- Affiliations
- University of Mississippi Medical Center (US)
- University of Alabama at Birmingham (US)
- Corresponding author
- Lirette, Seth T.
- Venue
- Journal of Nuclear Cardiology
- Publisher
- Elsevier BV
- Published
- 2017-12
- Volume
- 24
- Issue
- 6
- Pages
- 1998-2003
- ISSN
- 1071-3581 (ISSN-L)
- OpenAlex
- W2131824620
- Cited by
- 8
- References
- 16
- Citations by year
- 2026: 1 · 2025: 2 · 2023: 1 · 2022: 2 · 2020: 1 · 2018: 1
- FWCI
- 0.30
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Journal of Nuclear Cardiology (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Epidemiology
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Inference
- Physical Sciences › Mathematics › Statistics and Probability › Probability and Statistical Research
- Keywords
- survival models, prediction accuracy, concordance index, time-to-event outcomes, model validation
- MeSH terms
- Humans, Software, Models, Statistical, Proportional Hazards Models, Survival Analysis
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
