EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

On the estimation of inverse-probability-of-censoring weights for the evaluation of survival prediction error

Thomas Prince iD, Andrea Bommert iD, Jörg Rahnenführer, Matthias Schmid

PLOS ONE · 31.01.2025 · 20 (1) · e0318349

DOI: 10.1371/journal.pone.0318349 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS ONE, online erschienen am 31.01.2025. Verfasst von Thomas Prince und 3 weiteren Autor:innen. Beteiligte Einrichtung: Institute for Medical Biometry, Informatics and Epidemiology, Medical Faculty, University of Bonn, Bonn, Germany. Hauptthemen laut MEDLINE: Models, Statistical. Die Arbeit wurde bisher 10-mal zitiert (Crossref) und nennt 56 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Inverse probability weighting (IPW) is a popular method for making inferences regarding unobserved or unobservable data of a target population based on observed data. This paper considers IPW applied to right-censored time-to-event data. We investigate the behavior of the inverse-probability-of-censoring weighted (IPCW) Brier score, which is frequently used to assess the predictive performance of time-to-event models. A key requirement of the IPCW Brier score is the estimation of the censoring distribution, which is needed to compute the weights. The established paradigm of splitting a dataset into a training and a test set for model fitting and evaluation raises the question which of these datasets to use in order to fit the censoring model. There seems to be considerable disagreement between authors with regards to this issue, and no standard has been established so far. To shed light on this important question, we conducted a comprehensive experimental study exploring various data scenarios and estimation schemes. We found that it is generally of little importance which dataset is used to model the censoring distribution. However, in some circumstances, such as in the case of a covariate-dependent censoring process, a small sample size, or when dealing with noisy data, it may be advisable to use the test set instead of the training set to model the censoring distribution. A detailed set of practical recommendations concludes our paper.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Models, StatisticalHumansProbabilitySurvival Analysis

Autor:innen und Institutionen

  • Institute for Medical Biometry, Informatics and Epidemiology, Medical Faculty, University of Bonn, Bonn, Germany.
  • Department of Statistics, TU Dortmund University, Dortmund, Germany.

ORCID-Profile: Thomas Prince, Andrea Bommert

Zitieren

APA

Prince, T., Bommert, A., Rahnenführer, J., & Schmid, M. (2025). On the estimation of inverse-probability-of-censoring weights for the evaluation of survival prediction error. PLOS ONE, 20(1), e0318349. https://doi.org/10.1371/journal.pone.0318349

Vancouver

Prince T, Bommert A, Rahnenführer J, Schmid M. On the estimation of inverse-probability-of-censoring weights for the evaluation of survival prediction error. PLOS ONE. 2025;20(1):e0318349. doi:10.1371/journal.pone.0318349

BibTeX
@article{Prince2025,
  author = {Prince, Thomas and Bommert, Andrea and Rahnenführer, Jörg and Schmid, Matthias},
  title = {On the estimation of inverse-probability-of-censoring weights for the evaluation of survival prediction error},
  journal = {PLOS ONE},
  year = {2025},
  volume = {20},
  number = {1},
  pages = {e0318349},
  doi = {10.1371/journal.pone.0318349}
}
Kostenlos starten →