A Random Forest Inverse Probability Weighted Pseudo-Observation Framework for Alternating Recurrent Events

2025-10-27

AI Summary

This paper addresses estimation bias in alternating recurrent event settings—where primary and secondary events alternate—arising from informative missingness of the at-risk period for the primary event due to right-censoring and occurrence of secondary events. We propose a pseudo-observation method based on random forest inverse probability weighting (RF-IPW) to obtain unbiased estimation of the τ-restricted mean time to the primary event. Our key innovation lies in employing random forests to flexibly model dynamic weights, thereby accurately correcting nonrandom missingness induced by the alternating state structure—applicable whether primary and secondary events are dependent or independent. Simulation studies demonstrate superior performance in terms of bias reduction, standard error accuracy, and confidence interval coverage. An empirical application to a mobile health study quantifies the causal effect of self-care notifications on caregivers’ psychological outcomes among patients with brain injury.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Intelligent Robots: State Estimation

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Security and Privacy: Large-scale security measurements

Abstract

Alternating recurrent events, where subjects experience two potentially correlated event types over time, are common in healthcare, social, and behavioral studies. Often there is a primary event of interest that, when triggered, initiates a period of treatment and recovery measured via a secondary time-to-event. For example, cancer patients can experience repeated blood clotting emergencies that require hospitalization followed by discharge, people with alcohol use disorder can have periods of addiction and sobriety, or care partners can experience periods of depression and recovery. Potential censoring of the data requires special handling. Overlaying this are the missing at-risk periods for the primary event type when individuals have initiated the primary event but not reached the subsequent secondary event. In this paper, we develop a framework for regression analysis of censored alternating recurrent events that uses a random forest inverse probability weighting strategy to avoid bias in the analysis of the time to the primary event due to informative missingness from the alternate secondary state. The proposed regression model estimates $τ$-restricted mean time to the primary event of interest while taking into account complexities of censored. Simulations show good performance of our method when the alternate times-to-event are either independent or correlated. We analyze a mobile health study data to evaluate the impact of self-care push notifications on the mental state of caregivers of traumatic brain injury patients.

Problem

• Analyzing censored alternating recurrent events with correlated event types
• Addressing bias in primary event time analysis due to informative missingness
• Estimating restricted mean time to primary event under complex censoring

Innovation

• Random forest inverse probability weighting for bias avoidance
• Regression model for τ-restricted mean time estimation
• Handling censored alternating recurrent events with missing periods