CRISP-NAM: Competing Risks Interpretable Survival Prediction with Neural Additive Models

TL;DR

CRISP-NAM introduces an interpretable neural additive model for competing risks survival analysis, enabling cause-specific hazard modeling with feature-level interpretability and competitive predictive performance.

Contribution

It extends neural additive models to handle competing risks, providing cause-specific hazard estimation with transparent feature contributions.

Findings

Competitive performance on multiple datasets

Effective visualization of feature-risk relationships

Enhanced interpretability in survival analysis

Abstract

Competing risks are crucial considerations in survival modelling, particularly in healthcare domains where patients may experience multiple distinct event types. We propose CRISP-NAM (Competing Risks Interpretable Survival Prediction with Neural Additive Models), an interpretable neural additive model for competing risks survival analysis which extends the neural additive architecture to model cause-specific hazards while preserving feature-level interpretability. Each feature contributes independently to risk estimation through dedicated neural networks, allowing for visualization of complex non-linear relationships between covariates and each competing risk. We demonstrate competitive performance on multiple datasets compared to existing approaches.