Fast and accurate recurrent event analysis for genome-wide association studies.

Many diseases recur after recovery, for example, recurrences in cancer and infections. However, research is often focused on analysing only time-to-first recurrence, thereby ignoring any subsequent recurrences that may occur after the first. Statistical models for the analysis of recurrent events are available, of which the extended Cox proportional hazards frailty model is the current state-of-the-art. However, this model is too statistically complex for computationally efficient application in high-dimensional data sets, including genome-wide association studies (GWAS). Here, we develop an application for fast and accurate recurrent event analysis in GWAS, called SPARE (SaddlePoint Approximation for Recurrent Event analysis). In SPARE, every DNA variant is tested for association with recurrence risk using a modified score statistic. A saddlepoint approximation is implemented to achieve statistical accuracy. SPARE controls the Type I error, and its statistical power is similar to existing recurrent event models, yet SPARE is significantly faster. An application of SPARE in a recurrent event GWAS on bladder cancer for 6.2 million DNA variants in 1,443 individuals required less than 15 min, whereas existing recurrent event methods would require several weeks.

Bayesian design of clinical trials using joint models for recurrent and terminating events.

Joint models for recurrent event and terminating event data are increasingly used for the analysis of clinical trials. However, few methods have been proposed for designing clinical trials using these models. In this article, we develop a Bayesian clinical trial design methodology focused on evaluating the effect of an investigational product (IP) on both recurrent event and terminating event processes considered as multiple primary endpoints, using a multifrailty joint model. Dependence between the recurrent and terminating event processes is accounted for using a shared frailty. Inferences for the multiple primary outcomes are based on posterior model probabilities corresponding to mutually exclusive hypotheses regarding the benefit of IP with respect to the recurrent and terminating event processes. We propose an approach for sample size determination to ensure the trial design has a high power and a well-controlled type I error rate, with both operating characteristics defined from a Bayesian perspective. We also consider a generalization of the proposed parametric model that uses a nonparametric mixture of Dirichlet processes to model the frailty distributions and compare its performance to the proposed approach. We demonstrate the methodology by designing a colorectal cancer clinical trial with a goal of demonstrating that the IP causes a favorable effect on at least one of the two outcomes but no harm on either.

Evaluating dynamic and predictive discrimination for recurrent event models: use of a time-dependent C-index.

Interest in analyzing recurrent event data has increased over the past few decades. One essential aspect of a risk prediction model for recurrent event data is to accurately distinguish individuals with different risks of developing a recurrent event. Although the concordance index (C-index) effectively evaluates the overall discriminative ability of a regression model for recurrent event data, a local measure is also desirable to capture dynamic performance of the regression model over time. Therefore, in this study, we propose a time-dependent C-index measure for inferring the model's discriminative ability locally. We formulated the C-index as a function of time using a flexible parametric model and constructed a concordance-based likelihood for estimation and inference. We adapted a perturbation-resampling procedure for variance estimation. Extensive simulations were conducted to investigate the proposed time-dependent C-index's finite-sample performance and estimation procedure. We applied the time-dependent C-index to three regression models of a study of re-hospitalization in patients with colorectal cancer to evaluate the models' discriminative capability.

Self-Rated Health in Middle Age and Risk of Hospitalizations and Death: Recurrent Event Analysis of the ARIC Study.

BACKGROUND: Self-rated health is a simple measure that may identify individuals who are at a higher risk for hospitalization or death. OBJECTIVE: To quantify the association between a single measure of self-rated health and future risk of recurrent hospitalizations or death. PARTICIPANTS: Atherosclerosis Risk in Communities (ARIC) study, a community-based prospective cohort study of middle-aged men and women with follow-up beginning from 1987 to 1989. MAIN MEASURES: We quantified the associations between initial self-rated health with risk of recurrent hospitalizations and of death using a recurrent events survival model that allowed for dependency between the rates of hospitalization and hazards of death, adjusted for demographic and clinical factors. KEY RESULTS: Of the 14,937 ARIC cohort individuals with available self-rated health and covariate information, 34% of individuals reported "excellent" health, 47% "good," 16% "fair," and 3% "poor" at study baseline. After a median follow-up of 27.7 years, 1955 (39%), 3569 (51%), 1626 (67%), and 402 (83%) individuals with "excellent," "good," "fair," and "poor" health, respectively, had died. After adjusting for demographic factors and medical history, a less favorable self-rated health status was associated with increased rates of hospitalization and death. As compared to those reporting "excellent" health, adults with "good," "fair," and "poor" health had 1.22 (1.07 to 1.40), 2.01 (1.63 to 2.47), and 3.13 (2.39 to 4.09) times the rate of hospitalizations, respectively. The hazards of death also increased with worsening categories of self-rated health, with "good," "fair," and "poor" health individuals experiencing 1.30 (1.12 to 1.51), 2.15 (1.71 to 2.69), and 3.40 (2.54 to 4.56) times the hazard of death compared to "excellent," respectively. CONCLUSIONS: Even after adjusting for demographic and clinical factors, having a less favorable response on a single measure of self-rated health taken in middle age is a potent marker of future hospitalizations and death.

Clustering blood donors via mixtures of product partition models with covariates.

Motivated by the problem of accurately predicting gap times between successive blood donations, we present here a general class of Bayesian nonparametric models for clustering. These models allow for the prediction of new recurrences, accommodating covariate information that describes the personal characteristics of the sample individuals. We introduce a prior for the random partition of the sample individuals, which encourages two individuals to be co-clustered if they have similar covariate values. Our prior generalizes product partition models with covariates (PPMx) models in the literature, which are defined in terms of cohesion and similarity functions. We assume cohesion functions that yield mixtures of PPMx models, while our similarity functions represent the denseness of a cluster. We show that including covariate information in the prior specification improves the posterior predictive performance and helps interpret the estimated clusters in terms of covariates in the blood donation application.

Proportional rates models for multivariate panel count data.

Multivariate panel count data arise when there are multiple types of recurrent events, and the observation for each study subject consists of the number of recurrent events of each type between two successive examinations. We formulate the effects of potentially time-dependent covariates on multiple types of recurrent events through proportional rates models, while leaving the dependence structures of the related recurrent events completely unspecified. We employ nonparametric maximum pseudo-likelihood estimation under the working assumptions that all types of events are independent and each type of event is a nonhomogeneous Poisson process, and we develop a simple and stable EM-type algorithm. We show that the resulting estimators of the regression parameters are consistent and asymptotically normal, with a covariance matrix that can be estimated consistently by a sandwich estimator. In addition, we develop a class of graphical and numerical methods for checking the adequacy of the fitted model. Finally, we evaluate the performance of the proposed methods through simulation studies and analysis of a skin cancer clinical trial.

A joint frailty model for recurrent and competing terminal events: Application to delirium in the ICU.

Joint models linking longitudinal biomarkers or recurrent event processes with a terminal event, for example, mortality, have been studied extensively. Motivated by studies of recurrent delirium events in patients receiving care in an intensive care unit (ICU), we devise a joint model for a recurrent event process and multiple terminal events. Being discharged alive from the ICU or experiencing mortality may be associated with a patient's hazard of delirium, violating the assumption of independent censoring. Moreover, the direction of the association between the hazards of delirium and mortality may be opposite of the direction of association between the hazards of delirium and ICU discharge. Hence treating either terminal event as independent censoring may bias inferences. We propose a competing joint model that uses a latent frailty to link a patient's recurrent and competing terminal event processes. We fit our model to data from a completed placebo-controlled clinical trial, which studied whether Haloperidol could prevent death and delirium among ICU patients. The clinical trial served as a foundation for a simulation study, in which we evaluate the properties, for example, bias and confidence interval coverage, of the competing joint model. As part of the simulation study, we demonstrate the shortcomings of using a joint model with a recurrent delirium process and a single terminal event to study delirium in the ICU. Lastly, we discuss limitations and possible extensions for the competing joint model. The competing joint model has been added to frailtypack, an R package for fitting an assortment of joint models.

Robust inference for causal mediation analysis of recurrent event data.

Recurrent events, including cardiovascular events, are commonly observed in biomedical studies. Understanding the effects of various treatments on recurrent events and investigating the underlying mediation mechanisms by which treatments may reduce the frequency of recurrent events are crucial tasks for researchers. Although causal inference methods for recurrent event data have been proposed, they cannot be used to assess mediation. This study proposed a novel methodology of causal mediation analysis that accommodates recurrent outcomes of interest in a given individual. A formal definition of causal estimands (direct and indirect effects) within a counterfactual framework is given, and empirical expressions for these effects are identified. To estimate these effects, a semiparametric estimator with triple robustness against model misspecification was developed. The proposed methodology was demonstrated in a real-world application. The method was applied to measure the effects of two diabetes drugs on the recurrence of cardiovascular disease and to examine the mediating role of kidney function in this process.

Distributional imputation for the analysis of censored recurrent events.

Longitudinal clinical trials for which recurrent events endpoints are of interest are commonly subject to missing event data. Primary analyses in such trials are often performed assuming events are missing at random, and sensitivity analyses are necessary to assess robustness of primary analysis conclusions to missing data assumptions. Control-based imputation is an attractive approach in superiority trials for imposing conservative assumptions on how data may be missing not at random. A popular approach to implementing control-based assumptions for recurrent events is multiple imputation (MI), but Rubin's variance estimator is often biased for the true sampling variability of the point estimator in the control-based setting. We propose distributional imputation (DI) with corresponding wild bootstrap variance estimation procedure for control-based sensitivity analyses of recurrent events. We apply control-based DI to a type I diabetes trial. In the application and simulation studies, DI produced more reasonable standard error estimates than MI with Rubin's combining rules in control-based sensitivity analyses of recurrent events.

Estimation of trajectory of protective efficacy in infectious disease prevention trials using recurrent event times.

In studies of infectious disease prevention, the level of protective efficacy of medicinal products such as vaccines and prophylactic drugs tends to vary over time. Many products require administration of multiple doses at scheduled times, as opposed to one-off or continual intervention. Accurate information on the trajectory of the level of protective efficacy over time facilitates informed clinical recommendations and implementation strategies, for example, with respect to the timing of administration of the doses. Based on concepts from pharmacokinetic and pharmacodynamic modeling, we propose a non-linear function for modeling the trajectory after each dose. The cumulative effect of multiple doses of the products is captured by an additive series of the function. The model has the advantages of parsimony and interpretability, while remaining flexible in capturing features of the trajectories. We incorporate this series into the Andersen-Gill model for analysis of recurrent event time data and compare it with alternative parametric and non-parametric functions. We use data on clinical malaria disease episodes from a trial of four doses of an anti-malarial drug combination for chemoprevention to illustrate, and evaluate the performance of the methods using simulation. The proposed method out-performed the alternatives in the analysis of real data in terms of Akaike and Bayesian Information Criterion. It also accurately captured the features of the protective efficacy trajectory such as the area under curve in simulations. The proposed method has strong potential to enhance the evaluation of disease prevention measures and improve their implementation strategies.

τ $$ \tau $$ -Inflated beta regression model for censored recurrent events.

This research introduces a multivariate τ $$ \tau $$ -inflated beta regression ( τ $$ \tau $$ -IBR) modeling approach for the analysis of censored recurrent event data that is particularly useful when there is a mixture of (a) individuals who are generally less susceptible to recurrent events and (b) heterogeneity in duration of event-free periods amongst those who experience events. The modeling approach is applied to a restructured version of the recurrent event data that consists of censored longitudinal times-to-first-event in τ $$ \tau $$ length follow-up windows that potentially overlap. Multiple imputation (MI) and expectation-solution (ES) approaches appropriate for censored data are developed as part of the model fitting process. A suite of useful analysis outputs are provided from the τ $$ \tau $$ -IBR model that include parameter estimates to help interpret the (a) and (b) mixture of event times in the data, estimates of mean τ $$ \tau $$ -restricted event-free duration in a τ $$ \tau $$ -length follow-up window based on a patient's covariate profile, and heat maps of raw τ $$ \tau $$ -restricted event-free durations observed in the data with censored observations augmented via averages across MI datasets. Simulations indicate good statistical performance of the proposed τ $$ \tau $$ -IBR approach to modeling censored recurrent event data. An example is given based on the Azithromycin for Prevention of COPD Exacerbations Trial.

Modelling alternately recurring events using subject specific hazard estimation approach.

The motivation for this paper is to account for subject specific variations in a Cox proportional hazard model for alternating recurrent events. This is done through two sets of frailty components, whose marginal distributions are bound together by a copula function. The likelihood function involves unobservable variables, which requires the use of the EM algorithm. This leads to intractable integrals, which after some approximations, are solved using computationally intensive techniques. The results are applied to a real-life data. A simulation study is also carried out to check for consistency.

Sample size estimation for recurrent event data using multifrailty and multilevel survival models.

In epidemiology and clinical research, recurrent events refer to individuals who are likely to experience transient clinical events repeatedly over an observation period. Examples include hospitalizations in patients with heart failure, fractures in osteoporosis studies and the occurrence of new lesions in oncology. We provided an in-depth analysis of the sample size required for the analysis of recurrent time-to-event data using multifrailty or multilevel survival models. We covered the topic from the simple shared frailty model to models with hierarchical or joint frailties. We relied on a Wald-type test statistic to estimate the sample size assuming either a single or multiple endpoints. Simulations revealed that the sample size increased as heterogeneity increased. We also observed that it was more attractive to include more patients and reduce the duration of follow-up than to include fewer patients and increase the duration of follow-up to obtain the number of events required. Each model investigated can address the question of the number of subjects for recurrent events. However, depending on the research question, one model will be more suitable than another. We illustrated our methodology with the AFFIRM-AHF trial investigating the effect of intravenous ferric carboxymaltose in patients hospitalised for acute heart failure.

Dissecting the restricted mean time in favor of treatment.

The restricted mean time in favor (RMT-IF) summarizes the treatment effect on a hierarchical composite endpoint with mortality at the top. Its crude decomposition into "stage-wise effects," i.e., the net average time gained by the treatment prior to each component event, does not reveal the patient state in which the extra time is spent. To obtain this information, we break each stage-wise effect into subcomponents according to the specific state to which the reference condition is improved. After re-expressing the subcomponents as functionals of the marginal survival functions of outcome events, we estimate them conveniently by plugging in the Kaplan -- Meier estimators. Their robust variance matrices allow us to construct joint tests on the decomposed units, which are particularly powerful against component-wise differential treatment effects. By reanalyzing a cancer trial and a cardiovascular trial, we acquire new insights into the quality and composition of the extra survival times, as well as the extra time with fewer hospitalizations, gained by the treatment in question. The proposed methods are implemented in the rmt package freely available on the Comprehensive R Archive Network (CRAN).

Simulation-based sample size calculations of marginal proportional means models for recurrent events with competing risks.

In randomised controlled trials, the outcome of interest could be recurrent events, such as hospitalisations for heart failure. If mortality rates are non-negligible, both recurrent events and competing terminal events need to be addressed when formulating the estimand and statistical analysis is no longer trivial. In order to design future trials with primary recurrent event endpoints with competing risks, it is necessary to be able to perform power calculations to determine sample sizes. This paper introduces a simulation-based approach for power estimation based on a proportional means model for recurrent events and a proportional hazards model for terminal events. The simulation procedure is presented along with a discussion of what the user needs to specify to use the approach. The method is flexible and based on marginal quantities which are easy to specify. However, the method introduces a lack of a certain type of dependence. This is explored in a sensitivity analysis which suggests that the power is robust in spite of that. Data from a randomised controlled trial, LEADER, is used as the basis for generating data for a future trial. Finally, potential power gains of recurrent event methods as opposed to first event methods are discussed.

Causal inference with recurrent and competing events.

Many research questions concern treatment effects on outcomes that can recur several times in the same individual. For example, medical researchers are interested in treatment effects on hospitalizations in heart failure patients and sports injuries in athletes. Competing events, such as death, complicate causal inference in studies of recurrent events because once a competing event occurs, an individual cannot have more recurrent events. Several statistical estimands have been studied in recurrent event settings, with and without competing events. However, the causal interpretations of these estimands, and the conditions that are required to identify these estimands from observed data, have yet to be formalized. Here we use a formal framework for causal inference to formulate several causal estimands in recurrent event settings, with and without competing events. When competing events exist, we clarify when commonly used classical statistical estimands can be interpreted as causal quantities from the causal mediation literature, such as (controlled) direct effects and total effects. Furthermore, we show that recent results on interventionist mediation estimands allow us to define new causal estimands with recurrent and competing events that may be of particular clinical relevance in many subject matter settings. We use causal directed acyclic graphs and single world intervention graphs to illustrate how to reason about identification conditions for the various causal estimands based on subject matter knowledge. Furthermore, using results on counting processes, we show that our causal estimands and their identification conditions, which are articulated in discrete time, converge to classical continuous time counterparts in the limit of fine discretizations of time. We propose estimators and establish their consistency for the various identifying functionals. Finally, we use the proposed estimators to compute the effect of blood pressure lowering treatment on the recurrence of acute kidney injury using data from the Systolic Blood Pressure Intervention Trial.

Assessing model prediction performance for the expected cumulative number of recurrent events.

In a recurrent event setting, we introduce a new score designed to evaluate the prediction ability, for a given model, of the expected cumulative number of recurrent events. This score can be seen as an extension of the Brier Score for single time to event data but works for recurrent events with or without a terminal event. Theoretical results are provided that show that under standard assumptions in a recurrent event context, our score can be asymptotically decomposed as the sum of the theoretical mean squared error between the model and the true expected cumulative number of recurrent events and an inseparability term that does not depend on the model. This decomposition is further illustrated on simulations studies. It is also shown that this score should be used in comparison with a reference model, such as a nonparametric estimator that does not include the covariates. Finally, the score is applied for the prediction of hospitalisations on a dataset of patients suffering from atrial fibrillation and a comparison of the prediction performances of different models, such as the Cox model, the Aalen Model or the Ghosh and Lin model, is investigated.

Inverse Probability Weighting to Estimate Exposure Effects on the Burden of Recurrent Outcomes in the Presence of Competing Events.

Recurrent events-outcomes that an individual can experience repeatedly over the course of follow-up-are common in epidemiologic and health services research. Studies involving recurrent events often focus on time to first occurrence or on event rates, which assume constant hazards over time. In this paper, we contextualize recurrent event parameters of interest using counterfactual theory in a causal inference framework and describe an approach for estimating a target parameter referred to as the mean cumulative count. This approach leverages inverse probability weights to control measured confounding with an existing (and underutilized) nonparametric estimator of recurrent event burden first proposed by Dong et al. in 2015. We use simulations to demonstrate the unbiased estimation of the mean cumulative count using the weighted Dong-Yasui estimator in a variety of scenarios. The weighted Dong-Yasui estimator for the mean cumulative count allows researchers to use observational data to flexibly estimate and contrast the expected number of cumulative events experienced per individual by a given time point under different exposure regimens. We provide code to ease application of this method.

A Dynamic Risk Model for Multitype Recurrent Events.

Recurrent events can occur more than once in the same individual; such events may be of different types, known as multitype recurrent events. They are very common in longitudinal studies. Often there is a terminating event, after which no further events can occur. The risk of any event, including terminating events such as death or cure, is typically affected by prior events. We propose a flexible joint multitype recurrent-events model that explicitly provides estimates of the change in risk for each event due to subject characteristics, including number and type of prior events and the absolute risk for every event type (terminating and nonterminating), and predicts event-free survival probability over a desired time period. The model is fully parametric, and therefore a standard likelihood function and robust standard errors can be constructed. We illustrate the model with applications to the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (1994-2002) and provide discussion of the results and model features.

Bayesian inference for a principal stratum estimand on recurrent events truncated by death.

Recurrent events are often important endpoints in randomized clinical trials. For example, the number of recurrent disease-related hospitalizations may be considered as a clinically meaningful endpoint in cardiovascular studies. In some settings, the recurrent event process may be terminated by an event such as death, which makes it more challenging to define and estimate a causal treatment effect on recurrent event endpoints. In this paper, we focus on the principal stratum estimand, where the treatment effect of interest on recurrent events is defined among subjects who would be alive regardless of the assigned treatment. For the estimation of the principal stratum effect in randomized clinical trials, we propose a Bayesian approach based on a joint model of the recurrent event and death processes with a frailty term accounting for within-subject correlation. We also present Bayesian posterior predictive check procedures for assessing the model fit. The proposed approaches are demonstrated in the randomized Phase III chronic heart failure trial PARAGON-HF (NCT01920711).