Author ORCID Identifier
0000-0002-2397-5340
Document Type
Conference Paper
Disciplines
Statistics
Abstract
Clinical prediction models are developed to predict long-term patient outcomes following medical interventions. One example motivating this research is the prediction of graft and patient survival after kidney transplantation, using data from the German organ transplantation registry. A practical issue in this context is to deal with incomplete information due to right-censoring, which arises when patients are lost to follow-up or enter the study at different times, resulting in varying durations of observation. This is particularly relevant in the registry data, where follow-up is frequently incomplete or irregular. While traditional survival analysis methods handle censoring by modeling the hazard function, an alternative approach is to use classification algorithms to directly predict survival probabilities at fixed time points. This approach is often observed in machine learning, where classification algorithms are more commonly applied than survival analysis. However, if right-censoring is not properly addressed, the resulting incomplete survival information can negatively impact the accuracy of predictions [3, 4]. We show how classification algorithms can be adopted for rightcensored data by incorporating inverse probability of censoring (IPC) weights into the classification algorithm, either by modifying the loss function [1, 5] or by adjusting the resampling scheme in ensemble learning based on bootstrap samples [2].We demonstrate that IPC-weighted classification algorithms can yield accurate survival predictions. However, we also show that the modified weighting schemes can increase the variance of predictions, which may raise the need for alternative estimators of prediction uncertainty. In addition to simulation studies, we illustrate these concepts using national registry data to predict graft and patient survival following kidney transplantation.
DOI
https://doi.org/10.21427/e4fh-vc97
Recommended Citation
Jahn, Antje, "Survival Predictions from Classification Algorithms – Concepts and Application to Graft and Patient Survival after Kidney Transplantation" (2025). SAML-25 Workshop on Statistical and Machine Learning. 22.
https://arrow.tudublin.ie/saml/22
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-Share Alike 4.0 International License.
Publication Details
Statistical and Machine Learning: Methods and Applications (SAML-25) on June 5th and 6th, 2025 at TU Dublin, Ireland.
doi:10.21427/e4fh-vc97