An online updating approach for testing the proportional hazards assumption with streams of survival data

DOI10.1111/BIOM.13137zbMATH Open1451.62149arXiv1809.01291OpenAlexW3101489592WikidataQ92689766 ScholiaQ92689766MaRDI QIDQ5128786FDOQ5128786

Authors: Yishu Xue, Haiying Wang, Jun Yan, Elizabeth D. Schifano 

Publication date: 26 October 2020

Published in: Biometrics (Search for Journal in Brave)

Abstract: The Cox model, which remains as the first choice in analyzing time-to-event data even for large datasets, relies on the proportional hazards (PH) assumption. When survival data arrive sequentially in chunks, a fast and minimally storage intensive approach to test the PH assumption is desirable. We propose an online updating approach that updates the standard test statistic as each new block of data becomes available, and greatly lightens the computational burden. Under the null hypothesis of PH, the proposed statistic is shown to have the same asymptotic distribution as the standard version computed on the entire data stream with the data blocks pooled into one dataset. In simulation studies, the test and its variant based on most recent data blocks maintain their sizes when the PH assumption holds and have substantial power to detect different violations of the PH assumption. We also show in simulation that our approach can be used successfully with "big data" that exceed a single computer's computational resources. The approach is illustrated with the survival analysis of patients with lymphoma cancer from the Surveillance, Epidemiology, and End Results Program. The proposed test promptly identified deviation from the PH assumption that was not captured by the test based on the entire data.

Full work available at URL: https://arxiv.org/abs/1809.01291

Recommendations

• Data-driven smooth tests of the proportional hazards assumption
• Group sequential checking of the proportional hazards assumption based on a time-dependent covariate
• scientific article; zbMATH DE number 168112
• Testing proportional hazards for specified covariates
• A smooth test in proportional hazard survival models using local partial likelihood fitting

zbMATH Keywords

Cox modeldiagnosticsSchoenfeld residualsproportional hazards (PH) assumption

Mathematics Subject Classification ID

Diagnostics, and linear inference and regression (62J20) Reliability and life testing (62N05) Applications of statistics to biology and medical sciences; meta analysis (62P10) Point processes (e.g., Poisson, Cox, Hawkes processes) (60G55) Sequential estimation (62L12)

Cited In (21)

• Approximating Partial Likelihood Estimators via Optimal Subsampling
• Distributed subdata selection for big data via sampling-based approach
• Online updating of information based model selection in the big data setting
• Renewable learning for multiplicative regression with streaming datasets
• Divide-and-conquer information-based optimal subdata selection algorithm
• Scalable estimation for high velocity survival data able to accommodate addition of covariates
• Optimal subsampling for multiplicative regression with massive data
• A model robust subsampling approach for generalised linear models in big data settings
• Online estimation of hazard rate under random censoring
• Parameter estimation of linear mixed effects model based on online update
• Renewable risk assessment of heterogeneous streaming time-to-event cohorts
• Sequential Data Integration Under Dataset Shift
• Distributed quantile regression in decentralized optimization
• Subsampling approach for least squares fitting of semi-parametric accelerated failure time models to massive survival data
• Survival analysis on data streams: analyzing temporal events in dynamically changing environments
• Online updating method to correct for measurement error in big data streams
• Online Updating of Survival Analysis
• Optimal subsampling for modal regression in massive data
• Divide and conquer for accelerated failure time model with massive time‐to‐event data
• Sampling-based estimation for massive survival data with additive hazards model
• Optimal subsampling for parametric accelerated failure time models with massive survival data

This page was built for publication: An online updating approach for testing the proportional hazards assumption with streams of survival data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5128786)