Statistics in Medicine最新文献

Motivated by the experience of COVID-19 trials, we consider clinical trials in the setting of an emerging disease in which the uncertainty of natural disease course and potential treatment effects makes advance specification of a sample size challenging. One approach to such a challenge is to use a group sequential design to allow the trial to stop on the basis of interim analysis results as soon as a conclusion regarding the effectiveness of the treatment under investigation can be reached. As such a trial may be halted before a formal stopping boundary is reached, we consider the final analysis under such a scenario, proposing alternative methods for when the decision to halt the trial is made with or without knowledge of interim analysis results. We address the problems of ensuring that the type I error rate neither exceeds nor falls unnecessarily far below the nominal level. We also propose methods in which there is no maximum sample size, the trial continuing either until the stopping boundary is reached or it is decided to halt the trial.

Understanding the underlying causes of maternal death across all regions of the world is essential to inform policies and resource allocation to reduce the mortality burden. However, in many countries there exists very little data on the causes of maternal death, and data that do exist do not capture the entire population at risk. In this article, we present a Bayesian hierarchical multinomial model to estimate maternal cause of death distributions globally, regionally, and for all countries worldwide. The framework combines data from various sources to inform estimates, including data from civil registration and vital systems, smaller-scale surveys and studies, and high-quality data from confidential enquiries and surveillance systems. The framework accounts for varying data quality and coverage, and allows for situations where one or more causes of death are missing. We illustrate the results of the model on three case-study countries that have different data availability situations.

In recent years, there has been a growing interest in the prediction of individualized treatment effects. While there is a rapidly growing literature on the development of such models, there is little literature on the evaluation of their performance. In this paper, we aim to facilitate the validation of prediction models for individualized treatment effects. The estimands of interest are defined based on the potential outcomes framework, which facilitates a comparison of existing and novel measures. In particular, we examine existing measures of discrimination for benefit (variations of the c-for-benefit), and propose model-based extensions to the treatment effect setting for discrimination and calibration metrics that have a strong basis in outcome risk prediction. The main focus is on randomized trial data with binary endpoints and on models that provide individualized treatment effect predictions and potential outcome predictions. We use simulated data to provide insight into the characteristics of the examined discrimination and calibration statistics under consideration, and further illustrate all methods in a trial of acute ischemic stroke treatment. The results show that the proposed model-based statistics had the best characteristics in terms of bias and accuracy. While resampling methods adjusted for the optimism of performance estimates in the development data, they had a high variance across replications that limited their accuracy. Therefore, individualized treatment effect models are best validated in independent data. To aid implementation, a software implementation of the proposed methods was made available in R.

A causal decomposition analysis allows researchers to determine whether the difference in a health outcome between two groups can be attributed to a difference in each group's distribution of one or more modifiable mediator variables. With this knowledge, researchers and policymakers can focus on designing interventions that target these mediator variables. Existing methods for causal decomposition analysis either focus on one mediator variable or assume that each mediator variable is conditionally independent given the group label and the mediator-outcome confounders. In this article, we propose a flexible causal decomposition analysis method that can accommodate multiple correlated and interacting mediator variables, which are frequently seen in studies of health behaviors and studies of environmental pollutants. We extend a Monte Carlo-based causal decomposition analysis method to this setting by using a multivariate mediator model that can accommodate any combination of binary and continuous mediator variables. Furthermore, we state the causal assumptions needed to identify both joint and path-specific decomposition effects through each mediator variable. To illustrate the reduction in bias and confidence interval width of the decomposition effects under our proposed method, we perform a simulation study. We also apply our approach to examine whether differences in smoking status and dietary inflammation score explain any of the Black-White differences in incident diabetes using data from a national cohort study.

Quality measurement plays an increasing role in U.S. health care. Measures inform quality improvement efforts, public reporting of variations in quality of care across providers and hospitals, and high-stakes financial decisions. To be meaningful in these contexts, measures should be reliable and not heavily impacted by chance variations in sampling or measurement. Several different methods are used in practice by measure developers and endorsers to evaluate reliability; however, there is uncertainty and debate over differences between these methods and their interpretations. We review methods currently used in practice, pointing out differences that can lead to disparate reliability estimates. We compare estimates from 14 different methods in the case of two sets of mental health quality measures within a large health system. We find that estimates can differ substantially and that these discrepancies widen when sample size is reduced.

Why does a beneficial treatment effect on a longitudinal biomarker not translate into overall treatment benefit on survival, when the biomarker is in fact a prognostic factor of survival? In a recent exploratory data analysis in oncology, we were faced with this seemingly paradoxical result. To address this problem, we applied a theoretically principled methodology called dynamic path analysis, which allows us to perform mediation analysis with a longitudinal mediator and survival outcome. The aim of the analysis is to decompose the total treatment effect into a direct treatment effect and an indirect treatment effect mediated through a carefully constructed mediation path. The dynamic nature of the underlying methodology enables us to describe how these effects evolve over time, which can add to the mechanistic understanding of the underlying processes. In this paper, we present a detailed description of the dynamic path analysis framework and illustrate its application to survival mediation analysis using simulated and real data. The use case analysis provides clarity on the specific exploratory question of interest while the methodology generalizes to a wide range of applications in drug development where time-to-event is the primary clinical outcome of interest.

Marginal structural models have been increasingly used by analysts in recent years to account for confounding bias in studies with time-varying treatments. The parameters of these models are often estimated using inverse probability of treatment weighting. To ensure that the estimated weights adequately control confounding, it is possible to check for residual imbalance between treatment groups in the weighted data. Several balance metrics have been developed and compared in the cross-sectional case but have not yet been evaluated and compared in longitudinal studies with time-varying treatment. We have first extended the definition of several balance metrics to the case of a time-varying treatment, with or without censoring. We then compared the performance of these balance metrics in a simulation study by assessing the strength of the association between their estimated level of imbalance and bias. We found that the Mahalanobis balance performed best. Finally, the method was illustrated for estimating the cumulative effect of statins exposure over one year on the risk of cardiovascular disease or death in people aged 65 and over in population-wide administrative data. This illustration confirms the feasibility of employing our proposed metrics in large databases with multiple time-points.

The Cox proportional hazards model is commonly used to analyze time-to-event data in clinical trials. Standard inference procedures for the Cox model are based on asymptotic approximations and may perform poorly when there are few events in one or both treatment groups, as may be the case when the event of interest is rare or when the experimental treatment is highly efficacious. In this article, we propose an exact test of equivalence and efficacy under a proportional hazard model with treatment effect as the only fixed effect, together with an exact confidence interval that is obtained by inverting the exact test. The proposed test is based on a conditional error method originally proposed for sample size reestimation problems. In the present context, the conditional error method is used to combine information from a sequence of hypergeometric distributions, one at each observed event time. The proposed procedures are evaluated in simulation studies and illustrated using real data from an HIV prevention trial. A companion R package "ExactCox" is available for download on CRAN.

Joint models for longitudinal and time-to-event data are receiving increasing attention owing to its capability of capturing the possible association between these two types of data. Typically, a joint model consists of a longitudinal submodel for longitudinal processes and a survival submodel for the time-to-event response, and links two submodels by common covariates that may carry both fixed and random effects. However, research gaps still remain on how to simultaneously select fixed and random effects from the two submodels under the joint modeling framework efficiently and effectively. In this article, we propose a novel block-coordinate gradient descent (BCGD) algorithm to simultaneously select multiple longitudinal covariates that may carry fixed and random effects in the joint model. Specifically, for the multiple longitudinal processes, a linear mixed effect model is adopted where random intercepts and slopes serve as essential covariates of the trajectories, and for the survival submodel, the popular proportional hazard model is employed. A penalized likelihood estimation is used to control the dimensionality of covariates in the joint model and estimate the unknown parameters, especially when estimating the covariance matrix of random effects. The proposed BCGD method can successfully capture the useful covariates of both fixed and random effects with excellent selection power, and efficiently provide a relatively accurate estimate of fixed and random effects empirically. The simulation results show excellent performance of the proposed method and support its effectiveness. The proposed BCGD method is further applied on two real data sets, and we examine the risk factors for the effects of different heart valves, differing on type of tissue, implanted in the aortic position and the risk factors for the diagnosis of primary biliary cholangitis.