cumulative incidence estimation

This study addresses the incidence time bias arising in biobank cohorts where prevalent cases (diagnosed before recruitment) coexist with incident cases (diagnosed during follow-up). To tackle this challenge, the authors propose a novel nonparametric estimator for the cumulative incidence function (CIF) that jointly models all diseased individuals—regardless of their disease onset time or survival trajectory—within a unified framework grounded in truncation data theory. The estimator is rigorously supported by asymptotic analysis to ensure desirable statistical properties. Application to UK Biobank cancer data and extensive simulation studies demonstrate that the proposed method substantially outperforms existing approaches, achieving markedly reduced bias while maintaining excellent convergence behavior, particularly for diseases characterized by early onset and long-term survival.

This study addresses the substantial loss of power experienced by existing two-sample tests for competing risks when cumulative incidence functions (CIFs) cross. To overcome this limitation, the authors propose a novel test statistic based on the integrated difference between the two CIFs and develop an asymptotically valid inference procedure using the wild bootstrap, thereby circumventing the intractable limiting distribution of the statistic. Theoretical analysis and extensive simulations demonstrate that the proposed method achieves higher statistical power and greater stability in finite samples, particularly in scenarios involving crossing CIFs. This approach offers a robust and practical solution for comparing groups in the presence of competing risks.

This study addresses the challenge of evaluating updated vaccine regimens against pathogens with multiple serotypes by proposing a causal inference framework that integrates individual-level data from historical phase III vaccine trials and immunobridging studies to estimate counterfactual and etiology-specific cumulative incidence curves. Methodologically, the authors develop a multiply robust, efficient estimator capable of testing the no-controlled direct effect assumption and validate its finite-sample performance through simulations. Applied to data from the COVAIL trial, the approach successfully reconstructs hypothetical cumulative incidence curves for bivalent mRNA booster vaccines, corroborating the plausibility of key causal assumptions and offering a novel tool for assessing the effectiveness of variant-matched vaccines.

Accurate estimation and inference for the average treatment effect (ATE) in multi-covariate randomized controlled trials (RCTs) remain challenging, particularly under high-dimensional covariates and small sample sizes. Method: Building on the Neyman finite-population framework, we propose a bias-corrected regression-adjustment estimator with cross-fitting and introduce, for the first time, an HC3-type heteroskedasticity-robust standard error tailored to high-dimensional settings. We rigorously derive first- and second-order stochastic expansions of the random component of regression-adjustment estimators, identifying the source of higher-order bias in conventional inference; leveraging this insight, we design a cross-fitting procedure to eliminate bias and extend HC3 standard errors to stratified experimental designs. Results: Simulations and reanalysis of Angrist et al. (2009)’s education RCT demonstrate that our method substantially improves estimation accuracy, confidence interval coverage, and statistical power—especially in small-sample and high-dimensional scenarios.

This paper addresses unbiased estimation of the marginal counterfactual cumulative incidence function (CIF) in observational studies under competing risks. We propose a modified Nelson–Aalen estimator based on inverse probability weighting (IPW), incorporating estimated propensity scores. Our key contribution is the first rigorous asymptotic theory for such an adjusted estimator within the competing risks framework: we explicitly characterize the additional variability induced by propensity score estimation and prove its asymptotically negligible impact on final inference. Leveraging counting process theory and influence function analysis, we derive influence functions for both the counterfactual cumulative hazard and the CIF, establishing consistency and asymptotic normality of the estimator. Simulation studies and real-data analyses demonstrate low bias, robust performance across scenarios, and accurate standard error estimation. The proposed method thus provides a theoretically sound and practically viable tool for causal inference in competing risks settings.

In studies with competing risks, conventional incidence measures are susceptible to censoring mechanisms and often fail to accurately reflect the true burden of events. This work proposes the Average Cause-Specific Hazard (ACSH), a novel incidence metric based on survival weighting that does not rely on assumptions about the censoring distribution. The authors develop a nonparametric estimator for ACSH and a corresponding two-sample comparison procedure. ACSH retains the intuitive interpretability of incidence rates while fully eliminating bias induced by censoring, and enables group comparisons without requiring strong modeling assumptions. Simulation studies demonstrate its favorable finite-sample properties, and its application to the CANVAS trial yields robust and interpretable estimates of between-group differences.

本文探讨了在病例对照研究中使用倾向评分方法进行因果推断，提出了几种新的估计方法和诊断工具以解决混杂偏差问题。

This study addresses the challenge of estimating causal effects on survival outcomes in clinical trials when intercurrent events—such as treatment discontinuation—distort interpretation. While principal stratification offers a theoretically sound framework for handling such complications, its practical application has been limited by methodological complexity and poor accessibility. Building upon the ICH E9(R1) addendum, this work systematically extends principal stratification to time-to-event settings by integrating mixture models with inverse probability weighting. The approach explicitly formalizes modeling assumptions and incorporates a structured sensitivity analysis workflow. Its performance is evaluated through both an oncology case study and comprehensive simulation experiments. To enhance usability, the project provides reproducible R code and a practical implementation guide, substantially improving the feasibility and utility of principal stratification for regulatory decision-making and clinical research.

This study addresses the coarsening of self-reported numeric variables in surveys—often caused by rounding or heaping—by proposing a novel approach that integrates design-based inference with latent variable modeling. Treating observed values as coarsened manifestations of an underlying continuous latent variable, the method jointly models the coarsening mechanism and the latent distribution via a survey-weighted pseudo-likelihood. It generates posterior predictive replicates to propagate coarsening-induced uncertainty into standard design-based estimators. This framework is the first to explicitly correct for coarsening bias under complex sampling designs, enabling unbiased estimation of means, quantiles, and threshold-based prevalence measures. Simulation studies demonstrate robustness across various model misspecifications and sampling scenarios, and empirical application to Italy’s PASSI behavioral surveillance data shows effective correction of coarsening-related estimation bias.

This study addresses recurrent event data subject to right censoring and a terminal event by proposing a pseudo-value–based regression approach to model the effects of covariates on the mean cumulative function (MCF) and its area under the curve (AUMCF) at fixed time points. The method systematically extends the pseudo-value regression framework to estimate covariate effects on both MCF and AUMCF, constructing pseudo-observations via influence functions and enabling efficient inference through generalized estimating equations or ordinary least squares. Computationally straightforward and interpretable, the approach is compatible with standard regression software. Simulation studies demonstrate its favorable performance across diverse recurrent event settings, exhibiting accurate estimation, proper confidence interval coverage, controlled Type I error rates, and high statistical power. The method is successfully applied to the ORATORIO clinical trial data.