Information criteria for detecting change-points in the Cox proportional hazards model

TL;DR

This paper develops AIC-type information criteria for detecting change-points in the Cox proportional hazards model, addressing irregularities and overfitting issues, with theoretical guarantees and practical validation.

Contribution

It introduces novel AIC-based criteria tailored for Cox models with change-points, overcoming limitations of traditional AIC in irregular settings.

Findings

Proposed criteria outperform traditional AIC in simulations.

Criteria effectively detect change-points in clinical trial data.

Explicit formulas derived for practical application.

Abstract

The Cox proportional hazards model, commonly used in clinical trials, assumes proportional hazards. However, it does not hold when, for example, there is a delayed onset of the treatment effect. In such a situation, an acute change in the hazard ratio function is expected to exist. This paper considers the Cox model with change-points and derives AIC-type information criteria for detecting those change-points. The change-point model does not allow for conventional statistical asymptotics due to its irregularity, thus a formal AIC that penalizes twice the number of parameters would not be analytically derived, and using it would clearly give overfitting analysis results. Therefore, we will construct specific asymptotics using the partial likelihood estimation method in the Cox model with change-points. Based on the original derivation method for AIC, we propose information criteria that are mathematically guaranteed. If the partial likelihood is used in the estimation, information criteria with penalties much larger than twice the number of parameters could be obtained in an explicit form. Numerical experiments confirm that the proposed criterion is clearly superior in terms of the original purpose of AIC, which is to provide an estimate that is close to the true structure. We also apply the proposed criterion to actual clinical trial data to indicate that it will easily lead to different results from the formal AIC.