On a Shape-Invariant Hazard Regression Model

TL;DR

This paper introduces a shape-invariant hazard regression model that accommodates covariates with effects on different scales, providing improved estimation and prediction in survival analysis.

Contribution

It proposes a novel shape-invariant hazard model with moment-based inference, addressing limitations of existing models in handling covariates on different effect scales.

Findings

Good finite sample performance demonstrated through simulations

Applied to VA lung cancer data showing model effectiveness

Identified significant treatment effects in HIVNET 012 data

Abstract

In survival analysis, Cox model is widely used for most clinical trial data. Alternatives include the additive hazard model, the accelerated failure time (AFT) model and a more general transformation model. All these models assume that the effects for all covariates are on the same scale. However, it is possible that for different covariates, the effects are on different scales. In this paper, we propose a shape-invariant hazard regression model that allows us to estimate the multiplicative treatment effect with adjustment of covariates that have non-multiplicative effects. We propose moment-based inference procedures for the regression parameters. We also discuss the risk prediction and goodness of fit test for our proposed model. Numerical studies show good finite sample performance of our proposed estimator. We applied our method to Veteran's Administration (VA) lung cancer data and the HIVNET 012 data. For the latter, we found that single-dose nevirapine treatment has a significant improvement for 18-month survival with appropriate adjustment for maternal CD4 counts and virus load.