研究园简易版 · 经典书籍 · 期刊文章 · 有原文 · 完整版(可多选筛选、搜索) · 首页
2023 年的期刊文章 · 第 13 页
共 1511 篇,按发表时间排,新的在前,每页 40 篇。这一页是 2023-07-04 到 2023-06-20。
按期刊、按主题看 →
Biometrics Survival Analysis Nonparametric & Semiparametric · 发表于 2023-07-04
Xi Ning, Yinghao Pan, Yanqing Sun, Peter B. Gilbert
摘要
We propose a broad class of so-called Cox–Aalen transformation models that incorporate both multiplicative and additive covariate effects on the baseline hazard function within a transformation. The proposed models provide a highly flexible and versatile class of semiparametric models that include the transformation models and the Cox–Aalen model as special cases. Specifically, it extends the transformation models by allowing potentially time-dependent covariates to work additively on the baseline hazard and extends the Cox–Aalen model through a predetermined transformation function. We propose an estimating equation approach and devise an expectation-solving (ES) algorithm that involves fast and robust calculations. The resulting estimator is shown to be consistent and asymptotically normal via modern empirical process techniques. The ES algorithm yields a computationally simple method for estimating the variance of both parametric and nonparametric estimators. Finally, we demonstrate the performance of our procedures through extensive simulation studies and applications in two randomized, placebo-controlled human immunodeficiency virus (HIV) prevention efficacy trials. The data example shows the utility of the proposed Cox–Aalen transformation models in enhancing statistical power for discovering covariate effects.
JASA Spatial & Time Series High-Dimensional Testing & Uncertainty · 发表于 2023-07-03
Daisuke Kurisu, Kengo Kato, Xiaofeng Shao
摘要
In this article, we establish a high-dimensional CLT for the sample mean of p-dimensional spatial data observed over irregularly spaced sampling sites in Rd, allowing the dimension p to be much larger than the sample size n. We adopt a stochastic sampling scheme that can generate irregularly spaced sampling sites in a flexible manner and include both pure increasing domain and mixed increasing domain frameworks. To facilitate statistical inference, we develop the spatially dependent wild bootstrap (SDWB) and justify its asymptotic validity in high dimensions by deriving error bounds that hold almost surely conditionally on the stochastic sampling sites. Our dependence conditions on the underlying random field cover a wide class of random fields such as Gaussian random fields and continuous autoregressive moving average random fields. Through numerical simulations and a real data analysis, we demonstrate the usefulness of our bootstrap-based inference in several applications, including joint confidence interval construction for high-dimensional spatial data and change-point detection for spatio-temporal data. Supplementary materials for this article are available online. *(abstract via OpenAlex)*
SiM Testing & Uncertainty · 发表于 2023-07-03
Aryana Arsham, Ionut Bebu, Thomas Mathew
摘要
Probability based criteria are proposed for the assessment of cost‐effectiveness of a new treatment compared to a standard treatment when there are multiple effectiveness measures. Depending on the preferences of a policy maker, there are several options to define such criteria. Two such metrics are investigated in detail. One metric is defined as the conditional probability that a new treatment is more effective with respect to the multiple effectiveness measures for patients having lower costs under the new treatment. A second metric is defined as the conditional probability that a new treatment is less costly for patients having greater health benefits under the new treatment. The metrics offer considerable flexibility to a policy maker as thresholds for cost and effectiveness can be incorporated into the metrics. Parametric confidence limits are developed using a percentile bootstrap approach assuming multivariate normality for the joint distribution of the log(cost) and effectiveness measures. A non‐parametric estimation procedure is also developed using the theory of U‐statistics. Numerical results indicate that the proposed confidence limits accurately maintain coverage probabilities. The methodologies are illustrated using a study on the treatment of type two diabetes. Code implementing the proposed methods are provided in the supporting information.
SiM Missing Data & Measurement Error Causal Inference Clinical Trials · 发表于 2023-07-02
Chia‐Rui Chang, Yue Song, Fan Li, Rui Wang
摘要
When analyzing data from randomized clinical trials, covariate adjustment can be used to account for chance imbalance in baseline covariates and to increase precision of the treatment effect estimate. A practical barrier to covariate adjustment is the presence of missing data. In this article, in the light of recent theoretical advancement, we first review several covariate adjustment methods with incomplete covariate data. We investigate the implications of the missing data mechanism on estimating the average treatment effect in randomized clinical trials with continuous or binary outcomes. In parallel, we consider settings where the outcome data are fully observed or are missing at random; in the latter setting, we propose a full weighting approach that combines inverse probability weighting for adjusting missing outcomes and overlap weighting for covariate adjustment. We highlight the importance of including the interaction terms between the missingness indicators and covariates as predictors in the models. We conduct comprehensive simulation studies to examine the finite‐sample performance of the proposed methods and compare with a range of common alternatives. We find that conducting the proposed adjustment methods generally improves the precision of treatment effect estimates regardless of the imputation methods when the adjusted covariate is associated with the outcome. We apply the methods to the Childhood Adenotonsillectomy Trial to assess the effect of adenotonsillectomy on neurocognitive functioning scores.
Biometrics Survival Analysis Nonparametric & Semiparametric · 发表于 2023-06-29
Zexi Cai, Tony Sit
摘要
Quantile regression has emerged as a useful and effective tool in modeling survival data, especially for cases where noises demonstrate heterogeneity. Despite recent advancements, non-smooth components involved in censored quantile regression estimators may often yield numerically unstable results, which, in turn, lead to potentially self-contradicting conclusions. We propose an estimating equation-based approach to obtain consistent estimators of the regression coefficients of interest via the induced smoothing technique to circumvent the difficulty. Our proposed estimator can be shown to be asymptotically equivalent to its original unsmoothed version, whose consistency and asymptotic normality can be readily established. Extensions to handle functional covariate data and recurrent event data are also discussed. To alleviate the heavy computational burden of bootstrap-based variance estimation, we also propose an efficient resampling procedure that reduces the computational time considerably. Our numerical studies demonstrate that our proposed estimator provides substantially smoother model parameter estimates across different quantile levels and can achieve better statistical efficiency compared to a plain estimator under various finite-sample settings. The proposed method is also illustrated via four survival datasets, including the HMO (health maintenance organizations) HIV (human immunodeficiency virus) data, the primary biliary cirrhosis (PBC) data, and so forth.
SMMR Clinical Trials Econometrics & Economics · 发表于 2023-06-29
Joseph A Chiarappa, Donald R Hoover, Alison Pedley
摘要
Comparative Poisson trials of an experimental treatment versus a control typically condition on the total number of events that occur across both arms (Design A). Inference is based on the binomial distribution. Recently, an approach termed Design C to compare K experimental treatments to the same control was introduced. Under Design C without curtailment, the trial continues until a prespecified number of events occur in the control arm, leading to inference based on the negative multinomial distribution. The question remains of how advantageous it is to conduct one Design C trial comparing K experimental treatment arms to the same control arm as opposed to conducting K separate Design A trials each comparing one experimental treatment arm to a different control arm. This paper, therefore, compares the expected number of subjects to enroll for the two designs under uncurtailed and curtailed settings. The designs are evaluated when the null hypothesis and various assumptions for the alternative hypothesis hold. We simulate a variety of combinations for the Type 1 error, power, and ratio of the incidence rate of events in the experimental treatment to control arms. Design C frequently offers significant savings in terms of sample size relative to Design A.
SMMR Survival Analysis Missing Data & Measurement Error · 发表于 2023-06-29
Lillian A Boe, Pamela A Shaw
摘要
In large epidemiologic studies, it is typical for an inexpensive, non-invasive procedure to be used to record disease status during regular follow-up visits, with less frequent assessment by a gold standard test. Inexpensive outcome measures like self-reported disease status are practical to obtain, but can be error-prone. Association analysis reliant on error-prone outcomes may lead to biased results; however, restricting analyses to only data from the less frequently observed error-free outcome could be inefficient. We have developed an augmented likelihood that incorporates data from both error-prone outcomes and a gold standard assessment. We conduct a numerical study to show how we can improve statistical efficiency by using the proposed method over standard approaches for interval-censored survival data that do not leverage auxiliary data. We extend this method for the complex survey design setting so that it can be applied in our motivating data example. Our method is applied to data from the Hispanic Community Health Study/Study of Latinos to assess the association between energy and protein intake and the risk of incident diabetes. In our application, we demonstrate how our method can be used in combination with regression calibration to additionally address the covariate measurement error in self-reported diet.
SMMR Bayesian Methods Genetics & Omics High-Dimensional · 发表于 2023-06-28
Xiaotian Dai, Xuewen Lu, Thierry Chekouo
摘要
Coronary artery disease is one of the most common types of cardiovascular disease. Death from coronary heart disease is influenced by genetic factors in both women and men. In this article, we propose a novel Bayesian variable selection framework for the identification of important genetic variants associated with coronary artery disease disease status. Instead of treating each feature independently as in conventional Bayesian variable selection methods, we propose an innovative prior for the inclusion probabilities of genetic variants that accounts for their ordering structure. We assume that neighboring variants are more likely to be selected together as they tend to be highly correlated and have similar biological functions. Additionally, we propose to group participating subjects based on underlying population structure and fit separate regressions, so that the regression coefficients can better reflect different disease risks in different population groups. Our approach borrows strength across regression models through an innovative prior inspired by the Markov random fields. The proposed framework can improve variable selection and prediction performances as demonstrated in the simulation studies. We also apply the proposed framework to the CATHeterization GENetics data with binary Coronary artery disease disease status.
SiM High-Dimensional Nonparametric & Semiparametric · 发表于 2023-06-27
Seyoung Park, Eun Ryung Lee, Hyokyoung G. Hong
摘要
Health outcomes, such as body mass index and cholesterol levels, are known to be dependent on age and exhibit varying effects with their associated risk factors. In this paper, we propose a novel framework for dynamic modeling of the associations between health outcomes and risk factors using varying‐coefficients (VC) regional quantile regression via K‐nearest neighbors (KNN) fused Lasso, which captures the time‐varying effects of age. The proposed method has strong theoretical properties, including a tight estimation error bound and the ability to detect exact clustered patterns under certain regularity conditions. To efficiently solve the resulting optimization problem, we develop an alternating direction method of multipliers (ADMM) algorithm. Our empirical results demonstrate the efficacy of the proposed method in capturing the complex age‐dependent associations between health outcomes and their risk factors.
JASA Survival Analysis Networks & Graphs · 发表于 2023-06-27
Fangda Song, Jing Chu, Shuangge Ma, Yingying Wei
摘要
Whenever we send a message via a channel such as e-mail, Facebook, WhatsApp, WeChat, or LinkedIn, we care about the response rate—the probability that our message will receive a response—and the response time—how long it will take to receive a reply. Recent studies have made considerable efforts to model the sending behaviors of messages in social networks with point processes. However, statistical research on modeling response rates and response times on social networks is still lacking. Compared with sending behaviors, which are often determined by the sender’s characteristics, response rates and response times further depend on the relationship between the sender and the receiver. Here, we develop a survival mixed membership blockmodel (SMMB) that integrates semiparametric cure rate models with a mixed membership stochastic blockmodel to analyze time-to-event data observed for node pairs in a social network, and we are able to prove its model identifiability without the pure node assumption. We develop a Markov chain Monte Carlo algorithm to conduct posterior inference and select the number of social clusters in the network according to the conditional deviance information criterion. The application of the SMMB to the Enron e-mail corpus offers novel insights into the company’s organization and power relations. *(abstract via OpenAlex)*
Biometrics Bayesian Methods Longitudinal Data Spatial & Time Series · 发表于 2023-06-27
Shuangshuang Xu, Marco A. R. Ferreira, Erica M. Porter, Christopher T. Franck
摘要
We propose a Bayesian model selection approach for generalized linear mixed models (GLMMs). We consider covariance structures for the random effects that are widely used in areas such as longitudinal studies, genome-wide association studies, and spatial statistics. Since the random effects cannot be integrated out of GLMMs analytically, we approximate the integrated likelihood function using a pseudo-likelihood approach. Our Bayesian approach assumes a flat prior for the fixed effects and includes both approximate reference prior and half-Cauchy prior choices for the variances of random effects. Since the flat prior on the fixed effects is improper, we develop a fractional Bayes factor approach to obtain posterior probabilities of the several competing models. Simulation studies with Poisson GLMMs with spatial random effects and overdispersion random effects show that our approach performs favorably when compared to widely used competing Bayesian methods including deviance information criterion and Watanabe–Akaike information criterion. We illustrate the usefulness and flexibility of our approach with three case studies including a Poisson longitudinal model, a Poisson spatial model, and a logistic mixed model. Our proposed approach is implemented in the R package GLMMselect that is available on CRAN.
Biometrika Spatial & Time Series Nonparametric & Semiparametric · 发表于 2023-06-27
Ottmar Cronie, Mehdi Moradi, Christophe A N Biscio
摘要
Motivated by the general ability of cross-validation to reduce overfitting and mean square error, we develop a cross-validation-based statistical theory for general point processes. It is based on the combination of two novel concepts for general point processes: cross-validation and prediction errors. Our cross-validation approach uses thinning to split a point process/pattern into pairs of training and validation sets, while our prediction errors measure discrepancy between two point processes. The new statistical approach, which may be used to model different distributional characteristics, exploits the prediction errors to measure how well a given model predicts validation sets using associated training sets. Having indicated that our new framework generalizes many existing statistical approaches, we then establish different theoretical properties for it, including large sample properties. We further recognize that nonparametric intensity estimation is an instance of Papangelou conditional intensity estimation, which we exploit to apply our new statistical theory to kernel intensity estimation. Using independent thinning-based cross-validation, we numerically show that the new approach substantially outperforms the state-of-the-art in bandwidth selection. Finally, we carry out intensity estimation for a dataset in forestry and a dataset in neurology.
SiM Missing Data & Measurement Error · 发表于 2023-06-26
Lisa N. Yelland, Jennie Louise, Brennan C. Kahan, Tim P. Morris, Katherine J. Lee, Thomas R. Sullivan
摘要
Many trials use stratified randomisation, where participants are randomised within strata defined by one or more baseline covariates. While it is important to adjust for stratification variables in the analysis, the appropriate method of adjustment is unclear when stratification variables are affected by misclassification and hence some participants are randomised in the incorrect stratum. We conducted a simulation study to compare methods of adjusting for stratification variables affected by misclassification in the analysis of continuous outcomes when all or only some stratification errors are discovered, and when the treatment effect or treatment‐by‐covariate interaction effect is of interest. The data were analysed using linear regression with no adjustment, adjustment for the strata used to perform the randomisation (randomisation strata), adjustment for the strata if all errors are corrected (true strata), and adjustment for the strata after some errors are discovered and corrected (updated strata). The unadjusted model performed poorly in all settings. Adjusting for the true strata was optimal, while the relative performance of adjusting for the randomisation strata or the updated strata varied depending on the setting. As the true strata are unlikely to be known with certainty in practice, we recommend using the updated strata for adjustment and performing subgroup analyses, provided the discovery of errors is unlikely to depend on treatment group, as expected in blinded trials. Greater transparency is needed in the reporting of stratification errors and how they were addressed in the analysis.
JASA Survival Analysis Meta-Analysis & Data Integration · 发表于 2023-06-26
Ziyi Li, Yu Shen, Jing Ning
摘要
Transfer learning has attracted increasing attention in recent years for adaptively borrowing information across different data cohorts in various settings. Cancer registries have been widely used in clinical research because of their easy accessibility and large sample size. Our method is motivated by the question of how to utilize cancer registry data as a complement to improve the estimation precision of individual risks of death for inflammatory breast cancer (IBC) patients at The University of Texas MD Anderson Cancer Center. When transferring information for risk estimation based on the cancer registries (i.e., source cohort) to a single cancer center (i.e., target cohort), time-varying population heterogeneity needs to be appropriately acknowledged. However, there is no literature on how to adaptively transfer knowledge on risk estimation with time-to-event data from the source cohort to the target cohort while adjusting for time-varying differences in event risks between the two sources. Our goal is to address this statistical challenge by developing a transfer learning approach under the Cox proportional hazards model. To allow data-adaptive levels of information borrowing, we impose Lasso penalties on the discrepancies in regression coefficients and baseline hazard functions between the two cohorts, which are jointly solved in the proposed transfer learning algorithm. As shown in the extensive simulation studies, the proposed method yields more precise individualized risk estimation than using the target cohort alone. Meanwhile, our method demonstrates satisfactory robustness against cohort differences compared with the method that directly combines the target and source data in the Cox model. We develop a more accurate risk estimation model for the MD Anderson IBC cohort given various treatment and baseline covariates, while adaptively borrowing information from the National Cancer Database to improve risk assessment. *(abstract via PubMed)*
Biometrics Clinical Trials Survival Analysis · 发表于 2023-06-25
Michael J. Martens, Soyoung Kim, Kwang Woo Ahn
摘要
Sample size and power determination are crucial design considerations for biomedical studies intending to formally test the effects of key variables on an outcome. Other known prognostic factors may exist, necessitating the use of techniques for covariate adjustment when conducting this evaluation. Moreover, the main interest often includes assessing the impact of more than one variable on an outcome, such as multiple treatments or risk factors. Regression models are frequently employed for these purposes, formalizing this assessment as a test of multiple regression parameters. But, the presence of multiple variables of primary interest and correlation between covariates can complicate sample size/power calculation. Given the paucity of available sample size formulas for this context, these calculations are often performed via simulation, which can be both time-consuming as well as demanding extensive probability modeling. We propose a simpler, general approach to sample size and power determination that may be applied when testing multiple parameters in commonly used regression models, including generalized linear models as well as ordinary and stratified versions of the Cox and Fine–Gray models. Through both rigorous simulations and theoretical derivations, we demonstrate the formulas' accuracy in producing sample sizes that will meet the type I error rate and power specifications of the study design.
JASA Spatial & Time Series · 发表于 2023-06-23
Piotr Fryzlewicz
摘要
We propose Narrowest Significance Pursuit (NSP), a general and flexible methodology for automatically detecting localised regions in data sequences which each must contain a change-point (understood as an abrupt change in the parameters of an underlying linear model), at a prescribed global significance level. NSP works with a wide range of distributional assumptions on the errors, and guarantees important stochastic bounds which directly yield exact desired coverage probabilities, regardless of the form or number of the regressors. In contrast to the widely studied “post-selection inference” approach, NSP paves the way for the concept of “post-inference selection”. An implementation is available in the R package nsp. *(abstract via OpenAlex)*
SiM Bayesian Methods Longitudinal Data · 发表于 2023-06-23
Felix M. Pabon‐Rodriguez, Grant D. Brown, Breanna M. Scorza, Christine A. Petersen
摘要
While many Bayesian state‐space models for infectious disease processes focus on population infection dynamics (eg, compartmental models), in this work we examine the evolution of infection processes and the complexities of the immune responses within the host using these techniques. We present a joint Bayesian state‐space model to better understand how the immune system contributes to the control of Leishmania infantum infections over the disease course. We use longitudinal molecular diagnostic and clinical data of a cohort of dogs to describe population progression rates and present evidence for important drivers of clinical disease. Among these results, we find evidence for the importance of co‐infection in disease progression. We also show that as dogs progress through the infection, parasite load is influenced by their age, ectoparasiticide treatment status, and serology. Furthermore, we present evidence that pathogen load information from an earlier point in time influences its future value and that the size of this effect varies depending on the clinical stage of the dog. In addition to characterizing the processes driving disease progression, we predict individual and aggregate patterns of Canine Leishmaniasis progression. Both our findings and the application to individual‐level predictions are of direct clinical relevance, presenting possible opportunities for application in veterinary practice and motivating lines of additional investigation to better understand and predict disease progression. Finally, as an important zoonotic human pathogen, these results may support future efforts to prevent and treat human Leishmaniosis.
SiM Latent Variables & Mixtures Longitudinal Data Bayesian Methods · 发表于 2023-06-23
Benedict Orindi, Adrian Quintero, Luk Bruyneel, Baoyue Li, Emmanuel Lesaffre
摘要
Li et al developed a multilevel covariance regression (MCR) model as an extension of the covariance regression model of Hoff and Niu. This model assumes a hierarchical structure for the mean and the covariance matrix. Here, we propose the combined multilevel factor analysis and covariance regression model in a Bayesian framework, simultaneously modeling the MCR model and a multilevel factor analysis (MFA) model. The proposed model replaces the responses in the MCR part with the factor scores coming from an MFA model. Via a simulation study and the analysis of real data, we show that the proposed model is quite efficient when the responses of the MCR model are not measured directly but are latent variables such as the patient experience measurements in our motivating dataset.
JASA Optimization & Learning Theory Bayesian Methods · 发表于 2023-06-23
Zhehui Chen, Simon Mak, C. F. Jeff Wu
摘要
The Expected Improvement (EI) method, proposed by Jones, Schonlau, andWelch, is a widely used Bayesian optimization method, which makes use of a fitted Gaussian process model for efficient black-box optimization. However, one key drawback of EI is that it is overly greedy in exploiting the fitted Gaussian process model for optimization, which results in suboptimal solutions even with large sample sizes. To address this, we propose a new hierarchical EI (HEI) framework, which makes use of a hierarchical Gaussian process model. HEI preserves a closed-form acquisition function, and corrects the over-greediness of EI by encouraging exploration of the optimization space. We then introduce hyperparameter estimation methods which allow HEI to mimic a fully Bayesian optimization procedure, while avoiding expensive Markov-chain Monte Carlo sampling steps. We prove the global convergence of HEI over a broad function space, and establish near-minimax convergence rates under certain prior specifications. Numerical experiments show the improvement of HEI over existing Bayesian optimization methods, for synthetic functions and a semiconductor manufacturing optimization problem. Supplementary materials for this article are available online. *(abstract via OpenAlex)*
SiM Clinical Trials · 发表于 2023-06-22
Steven Teerenstra, Jessica Kasza, Ruslan Leontjevas, Andrew B. Forbes
摘要
In a randomized controlled trial, outcomes of different subjects may be independent at baseline, but correlated at a follow‐up measurement due to treatment. This treatment‐related clustering at follow‐up can arise for instance because the treatment is given in a group or because subjects are treated individually but by the same therapist (therapist effect). There is substantial literature on the design and analysis of such trials when estimation of the intervention effect is based on a follow‐up measurement (eg, directly after treatment or at a later time point). However, often the baseline measurement of the outcome is highly correlated with the follow‐up measurement, and this information can be used in the analysis. For a randomized design with a baseline and a follow‐up measurement, we compare sample size requirements for analyses with and without adjustment for this baseline measure. We show that adjusting for baseline reduces required sample size. This reduction depends on the variance of the difference between arms at baseline, the variance of this difference at follow‐up, and the correlation between the two. From this, we derive sample size formulas for partially or fully nested designs, and cluster randomized trials with treatment as a partially or fully cross‐classified factor. Also, we discuss situations where clusters are already present at baseline or where treatment by cluster interaction is present. For the partially nested design, we work out practical design considerations (eg, use of content‐matter input, design factors and optimal allocation ratio) and investigate small sample properties of the sample size formula.
JASA Testing & Uncertainty High-Dimensional · 发表于 2023-06-22
Haoyu Geng, Xiaolong Cui, Haojie Ren, Changliang Zou
摘要
Two-sample multiple testing has a wide range of applications. Most of the literature considers simultaneous tests of equality of parameters. The article takes a different perspective and investigates the null hypotheses that the two support sets are equal. This formulation of the testing problem is motivated by the fact that in many applications where the two parameter vectors being compared are both sparse, one might be more concerned about the detection of differential sparsity structures rather than the difference in parameter magnitudes. Focusing on this type of problem, we develop a general approach, which adapts the newly proposed symmetry data aggregation tool combined with a novel double thresholding (DT) filter. The DT filter first constructs a sequence of pairs of ranking statistics that fulfill global symmetry properties and then chooses two data-driven thresholds along the ranking to simultaneously control the False Discovery Rate (FDR) and maximize the number of rejections. Several applications of the methodology are given including high-dimensional linear models and Gaussian graphical models. We show that the proposed method is able to asymptotically control the FDR and have power guarantee under certain conditions. Numerical results confirm the effectiveness and robustness of DT in FDR control and detection ability. Supplementary materials for this article are available online. *(abstract via OpenAlex)*
Biometrics Meta-Analysis & Data Integration Optimization & Learning Theory · 发表于 2023-06-22
Sangyoon Yi, Raymond Ka Wai Wong, Irina Gaynanova
摘要
The prevalence of data collected on the same set of samples from multiple sources (i.e., multi-view data) has prompted significant development of data integration methods based on low-rank matrix factorizations. These methods decompose signal matrices from each view into the sum of shared and individual structures, which are further used for dimension reduction, exploratory analyses, and quantifying associations across views. However, existing methods have limitations in modeling partially-shared structures due to either too restrictive models, or restrictive identifiability conditions. To address these challenges, we propose a new formulation for signal structures that include partially-shared signals based on grouping the views into so-called hierarchical levels with identifiable guarantees under suitable conditions. The proposed hierarchy leads us to introduce a new penalty, hierarchical nuclear norm (HNN), for signal estimation. In contrast to existing methods, HNN penalization avoids scores and loadings factorization of the signals and leads to a convex optimization problem, which we solve using a dual forward–backward algorithm. We propose a simple refitting procedure to adjust the penalization bias and develop an adapted version of bi-cross-validation for selecting tuning parameters. Extensive simulation studies and analysis of the genotype-tissue expression data demonstrate the advantages of our method over existing alternatives.
SiM Causal Inference · 发表于 2023-06-22
Samrat Roy, Ting Ye, Ashkan Ertefaie, Tat‐Thang Vo, James Flory, Sean Hennessy, Dylan Small
摘要
Unmeasured confounding is a major obstacle to reliable causal inference based on observational studies. Instrumented difference‐in‐differences (iDiD), a novel idea connecting instrumental variable and standard DiD, ameliorates the above issue by explicitly leveraging exogenous randomness in an exposure trend. In this article, we utilize the above idea of iDiD, and propose a novel group sequential testing method that provides valid inference even in the presence of unmeasured confounders. At each time point, we estimate the average or conditional average treatment effect under iDiD setting using the data accumulated up to that time point, and test the significance of the treatment effect. We derive the joint distribution of the test statistics under the null using the asymptotic properties of M‐estimation, and the group sequential boundaries are obtained using the ‐spending functions. The performance of our proposed approach is evaluated on both synthetic data and Clinformatics Data Mart Database (OptumInsight, Eden Prairie, MN) to examine the association between rofecoxib and acute myocardial infarction, and our method detects significant adverse effect of rofecoxib much earlier than the time when it was finally withdrawn from the market.
JASA High-Dimensional · 发表于 2023-06-22
Junlong Zhao, Yang Zhou, Yufeng Liu
摘要
High dimensional linear models are commonly used in practice. In many applications, one is interested in linear transformations β ⊤ x of regression coefficients β ∈ R p , where x is a specific point and is not required to be identically distributed as the training data. One common approach is the plug-in technique which first estimates β , then plugs the estimator in the linear transformation for prediction. Despite its popularity, estimation of β can be difficult for high dimensional problems. Commonly used assumptions in the literature include that the signal of coefficients β is sparse and predictors are weakly correlated. These assumptions, however, may not be easily verified, and can be violated in practice. When β is non-sparse or predictors are strongly correlated, estimation of β can be very difficult. In this paper, we propose a novel pointwise estimator for linear transformations of β . This new estimator greatly relaxes the common assumptions for high dimensional problems, and is adaptive to the degree of sparsity of β and strength of correlations among the predictors. In particular, β can be sparse or non-sparse and predictors can be strongly or weakly correlated. The proposed method is simple for implementation. Numerical and theoretical results demonstrate the competitive advantages of the proposed method for a wide range of problems. *(abstract via PubMed)*
Biometrics Survival Analysis High-Dimensional Causal Inference · 发表于 2023-06-22
Kelly Van Lancker, Oliver Dukes, Stijn Vansteelandt
摘要
The problem of how to best select variables for confounding adjustment forms one of the key challenges in the evaluation of exposure effects in observational studies, and has been the subject of vigorous recent activity in causal inference. A major drawback of routine procedures is that there is no finite sample size at which they are guaranteed to deliver exposure effect estimators and associated confidence intervals with adequate performance. In this work, we will consider this problem when inferring conditional causal hazard ratios from observational studies under the assumption of no unmeasured confounding. The major complication that we face with survival data is that the key confounding variables may not be those that explain the censoring mechanism. In this paper, we overcome this problem using a novel and simple procedure that can be implemented using off-the-shelf software for penalized Cox regression. In particular, we will propose tests of the null hypothesis that the exposure has no effect on the considered survival endpoint, which are uniformly valid under standard sparsity conditions. Simulation results show that the proposed methods yield valid inferences even when covariates are high-dimensional.
Biometrics Missing Data & Measurement Error Bayesian Methods Latent Variables & Mixtures · 发表于 2023-06-22
Michael J. Daniels, Minji Lee, Wei Feng
摘要
In longitudinal studies, it is not uncommon to make multiple attempts to collect a measurement after baseline. Recording whether these attempts are successful provides useful information for the purposes of assessing missing data assumptions. This is because measurements from subjects who provide the data after numerous failed attempts may differ from those who provide the measurement after fewer attempts. Previous models for these designs were parametric and/or did not allow sensitivity analysis. For the former, there are always concerns about model misspecification and for the latter, sensitivity analysis is essential when conducting inference in the presence of missing data. Here, we propose a new approach which minimizes issues with model misspecification by using Bayesian nonparametrics for the observed data distribution. We also introduce a novel approach for identification and sensitivity analysis. We re-analyze the repeated attempts data from a clinical trial involving patients with severe mental illness and conduct simulations to better understand the properties of our approach.
JASA Survival Analysis Nonparametric & Semiparametric · 发表于 2023-06-22
Chi Wing Chu, Tony Sit
摘要
Conventionally, censored quantile regression stipulates a specific, pointwise conditional quantile of the survival time given covariates.Despite its model flexibility and straightforward interpretation, the pointwise formulation oftentimes yields rather unstable estimates across neighboring quantile levels with large variances.In view of this phenomenon, we propose a new class of quantile-based regression models with timedependent covariates for censored data.The models proposed aim to capture the relationship between the failure time and the covariate processes of a target population that falls within a specific quantile bracket.The pooling of information within a homogeneous neighborhood facilitates more efficient estimates hence, more consistent conclusion on statistical significances of the variables concerned.This new formulation can also be regarded as a generalization of the accelerated failure time model for survival data in the sense that it relaxes the assumption of global homogeneity for the error at all quantile levels.By introducing a class of weighted rank-based estimation procedure, our framework allows a quantile-based inference on the covariate effect with a less restrictive set of assumptions.Numerical studies demonstrate that the proposed estimator outperforms existing alternatives under various settings in terms of smaller empirical biases and standard deviations.A perturbation-based resampling method is also developed to reconcile the asymptotic distribution of the parameter estimates.Finally, consistency and weak convergence of the proposed estimator are established via empirical process theory. *(abstract via OpenAlex)*
JASA Survival Analysis · 发表于 2023-06-22
Nir Keret, Malka Gorfine
摘要
Massive sized survival datasets become increasingly prevalent with the development of the healthcare industry, and pose computational challenges unprecedented in traditional survival analysis use cases. In this work we analyze the UK-biobank colorectal cancer data with genetic and environmental risk factors, including a time-dependent coefficient, which transforms the dataset into “pseudo-observation” form, thus, critically inflating its size. A popular way for coping with massive datasets is downsampling them, such that the computational resources can be afforded by the researcher. Cox regression has remained one of the most popular statistical models for the analysis of survival data to-date. This work addresses the settings of right censored and possibly left-truncated data with rare events, such that the observed failure times constitute only a small portion of the overall sample. We propose Cox regression subsampling-based estimators that approximate their full-data partial-likelihood-based counterparts, by assigning optimal sampling probabilities to censored observations, and including all observed failures in the analysis. The suggested methodology is applied on the UK-biobank for building a colorectal cancer risk-prediction model, while reducing the computation time and memory requirements. Asymptotic properties of the proposed estimators are established under suitable regularity conditions, and simulation studies are carried out to evaluate their finite sample performance. Supplementary materials for this article are available online. *(abstract via OpenAlex)*
SiM Clinical Trials Causal Inference Networks & Graphs · 发表于 2023-06-21
Alex Ocampo, Jemar R. Bather
摘要
Confusion often arises when attempting to articulate target estimand(s) of a clinical trial in plain language. We aim to rectify this confusion by using a type of causal graph called the Single‐World Intervention Graph (SWIG) to provide a visual representation of the estimand that can be effectively communicated to interdisciplinary stakeholders. These graphs not only display estimands, but also illustrate the assumptions under which a causal estimand is identifiable by presenting the graphical relationships between the treatment, intercurrent events, and clinical outcomes. To demonstrate its usefulness in pharmaceutical research, we present examples of SWIGs for various intercurrent event strategies specified in the ICH E9(R1) addendum, as well as an example from a real‐world clinical trial for chronic pain. code to generate all the SWIGs shown is this paper is made available. We advocate clinical trialists adopt the use of SWIGs in their estimand discussions during the planning stages of their studies.
SiM Testing & Uncertainty Clinical Trials · 发表于 2023-06-21
Samuel I. Watson, Joshua O. Akinyemi, Karla Hemming
摘要
In this article, we derive and compare methods to derive P ‐values and sets of confidence intervals with strong control of the family‐wise error rates and coverage for estimates of treatment effects in cluster randomized trials with multiple outcomes. There are few methods for P ‐value corrections and deriving confidence intervals, limiting their application in this setting. We discuss the methods of Bonferroni, Holm, and Romano and Wolf and adapt them to cluster randomized trial inference using permutation‐based methods with different test statistics. We develop a novel search procedure for confidence set limits using permutation tests to produce a set of confidence intervals under each method of correction. We conduct a simulation‐based study to compare family‐wise error rates, coverage of confidence sets, and the efficiency of each procedure in comparison to no correction using both model‐based standard errors and permutation tests. We show that the Romano‐Wolf type procedure has nominal error rates and coverage under non‐independent correlation structures and is more efficient than the other methods in a simulation‐based study. We also compare results from the analysis of a real‐world trial.
JASA Survival Analysis Machine Learning · 发表于 2023-06-21
Moritz Berger, Ana Kowark, Rolf Rossaint, Mark Coburn, Matthias Schmid, and the Group POSE-Study
摘要
Elderly patients are at a high risk of suffering from postoperative death. Personalized strategies to improve their recovery after intervention are therefore urgently needed. A popular way to analyze postoperative mortality is to develop a prognostic model that incorporates risk factors measured at hospital admission, for example, comorbidities. When building such models, numerous issues must be addressed, including censoring and the presence of competing events (such as discharge from hospital alive). Here we present a novel survival modeling approach to investigate 30-day inpatient mortality following intervention. The proposed method accounts for both grouped event times, for example, measured in 24-hour intervals, and competing events. Conceptually, the method is embedded in the framework of generalized additive models for location, scale, and shape (GAMLSS). Model fitting is performed using a component-wise gradient boosting algorithm, which allows for additional regularization steps via stability selection. We used this new modeling approach to analyze data from the Peri-interventional Outcome Study in the Elderly (POSE), which is a recent cohort study that enrolled 9862 elderly inpatients undergoing intervention under anesthesia. Application of the proposed boosting algorithm yielded six important risk factors (including both clinical variables and interventional characteristics) that either contributed to the hazard of death or to discharge from hospital alive. Supplementary materials for this article are available online. *(abstract via OpenAlex)*
Biometrics Other · 发表于 2023-06-20
James S. Hodges
摘要
This review assignment was timely because I just re-submitted a proposal to the US government's National Institutes of Health (NIH) and writing it was … difficult. Lisa Chasan-Taber is a professor of epidemiology who has had continuous grant funding for decades, been on many grant-review panels, mentored many junior colleagues in grant-writing, and taught a proposal-writing course for >20 years (the basis for this book). This book succeeds brilliantly: it is well organized, easy to read, and packed with practical advice. Its main target is early-career faculty but even senior researchers may find it educational—I did, and I have helped write about 250 proposals, though I have been the principal investigator (PI) of only four. The book's first chapter covers Prof. Chasan-Taber's “Ten Top Tips for Successful Grant Proposal Writing.”. These illustrate the author's pragmatic focus and style, while bringing out non-obvious aspects of apparently obvious things. Here are the 10 tips quoted in bold-face capitals, which I discuss using my recent proposal as an example. I will use “grantsman” because female colleagues use it to refer to themselves. START SMALL BUT HAVE A BIG VISION. FOCUS ON SMALL GRANTS TARGETED TO EARLY-CAREER INVESTIGATORS. “Each small grant … should be viewed as providing preliminary data for one or two of the specific aims of your ultimate larger grant. … [T]ry to envision your ultimate large project [, then s]tep by step, you start biting off small chunks of this larger grant through writing small grants designed to support one or two of these ultimate aims.” Like the book generally, this tip's discussion gives forward references to other tips and to later chapters, for example, the subsection “A Pitfall to Avoid: Interdependent Aims” states briefly why you should avoid them and refers to Chapter 6, “Specific Aims.” When I first wrote a proposal as a PI, I was 25 years out of grad school and never considered starting with small grants. That was a mistake and would be an even bigger mistake for a young investigator. Prof. Chasan-Taber says, “In my experience as an NIH review panel member, [applying for an R01 as your first proposal] is almost certainly destined to fail” and she's right. LOOK AT WHO AND WHAT THEY FUNDED BEFORE YOU. Prof. Chasan-Taber advises learning who is on the panel that will review your proposal, to get a sense of the reviewers. This is necessary but not foolproof. For the first submission of the grant I just re-submitted, our review panel differed from the roster the program officer sent us when we were writing the proposal; none of our three reviewers was on the roster we wrote for. SPEND HALF YOUR TIME ON THE SPECIFIC AIMS AND PROJECT SUMMARY/ABSTRACT; send it to people to review, rewrite it, send it to more people to review, etc. Prof. Chasan-Taber also suggests studying aims pages of successful proposals responding to similar calls for proposals, to get a sense of the scope of successful projects, number of aims, and so on. This is, again, excellent advice. On the proposal I just re-submitted, a co-investigator who is a successful grantsman insisted that we spend what seemed an inordinate amount of time developing a conceptual diagram for the project. But he was right, because doing so served the same purpose as this tip: it clarified our thinking and summarized the project for reviewers, and we wrote the rest of the proposal around it. SHOW THAT YOU CAN PULL IT OFF. This seems obvious but in our initial submission, we bungled it by assuming reviewers would read our bio-sketches and other secondary materials, and we did not say enough about our previous work or our relationship with our overseas collaborators. This is too important to leave to secondary materials. YOUR METHODS SHOULD MATCH YOUR AIMS AND VICE VERSA. “Include methods that address each of your study aims while taking care not to include additional methods that do not correspond to any aims.” This again seems obvious but in some ways it is not. For example, our original submission said we have other datasets we would use opportunistically in developing our statistical methods and the reviewers dinged us for not giving a rationale for these specific datasets. When writing the revision, I suggested just re-wording this because I could not imagine how it could be bad to try our methods on other datasets. But our experienced grantsman agreed with Prof. Chasan-Taber and the resubmitted proposal does not mention other datasets; it even extols our main application as providing everything needed for our methods–development aims. As advice about how to do research, I think this is simply wrong but “what's sensible in doing methods research” ≠ “what you should put in the proposal.” A PROPOSAL CAN NEVER HAVE TOO MANY FIGURES OR TABLES. “As compared to dense text, figures and tables make it easy for the reviewers to quickly grasp your proposal. … [T]he act of creating these figures and tables will help you to crystallize your specific aims and study methods … consider having a figure or table on almost every page of your proposal,” even the specific aims page. My gut reaction to this was: Yuk! I hate reading documents like that and some pictures are not worth the words they displace. But the point is not to please the grant writer, the point is to get the grant, and successful grantsmen do this. SEEK EXTERNAL REVIEW PRIOR TO SUBMISSION. “It is … generally acknowledged that a local mock review panel can double your chances of funding.” This is good advice, the hard part is starting early enough to have time for external reviews. BE KIND TO YOUR REVIEWERS. “The most effective way to make a reviewer happy is to help them complete their review forms. Do this by using the grant review criteria as subheadings in your application” and highlighting key sentences. I used to save space by hyphenating words at the ends of lines but a successful grantsman convinced me that the small saving in length was not worth the extra annoyance to reviewers from solid blocks of text with no white space. IF AT ALL POSSIBLE, CHOOSE A TOPIC THAT YOU FIND INTERESTING! This sad comment on academia—is this how you envisioned a life in science, when you were a kid watching NOVA? —points to something I wish Prof. Chasan-Taber had addressed. This tip acknowledges indirectly that academics face an incentive to chase what is fundable rather than what they consider good and exciting science. Prof. Chasan-Taber could address this by explaining how successful grantsmen find proposal writing stimulating, or at least not soul-deadening: do they—does she—see grant-writing as an opportunity to dream up future projects? If it is possible to cultivate attitudes, this attitude, whatever it is, is worth cultivating. I, on the contrary, find proposal writing to be a dead loss apart from the small chance of getting the grant, and I am not alone. One able colleague recently left our university for a pharma company's consulting operation mainly because he wanted to spend his time doing statistics instead of spending his time hustling grants so he could pay grad students to do statistics. The book's other chapters flesh out this summary with a systematic treatment of every step of writing a proposal. I will survey these briefly. Part I, “Preparing to Write the Grant Proposal” has three chapters on things to do before you start writing and a chapter about scientific writing. For example, the chapter about time management recommends starting 4–6 months before the deadline and lists major tasks that should be done 4 months before the deadline, 3 months, etc. Section 3 asks “HOW BIG A RESEARCH GAP DO I NEED TO FILL?” and gives a helpful table of gaps that reviewers find convincing. Much of this part is about being strategic in your academic career, which could be helpful to people whose advisors (like mine) threw them in the deep end of the pool and let them figure out how to swim. As for the chapter on scientific writing, if you have time to read only a short piece about writing, this is as good as it gets. (I fell in love with this chapter when the first page gave Orwell's six rules for writing.) Part II, “The Grant Proposal: Section by Section” has 11 relentlessly practical chapters about writing the sections of an NIH proposal. As an example, Chapter 14, “Study Limitations to Consider,” discusses eight (!) general classes of limitations and the section “Issues for Critical Reading” has a table for each of cohort studies, randomized trials, and case–control/cross-sectional studies, listing possible limitations with specific questions you can address. No matter how long you have been writing proposals, this kind of checklist is valuable. Part III, “Submission and Resubmission,” has chapters on boring but necessary parts of an NIH proposal, and on reviews and re-submitting a proposal. It has two other chapters for young researchers, on fellowships and early-career awards. The latter exemplify this book's attention to detail with, for example, advice for writing the “Contribution to Science” section of an NIH bio-sketch, which is hard for grad students who might reasonably think they have not yet made any contributions to science. One possible weakness of this book is that it is totally focused on NIH proposals. For people outside the US or writing proposals to other agencies, the NIH-specific material will not be useful but the rest of the book is still immensely valuable. The book also emphasizes considerations pertinent to proposals in epidemiology, so people in other fields may find some chapters less useful, for example, Chapters 14 and 15 on study limitations. As much as I admire this book, it lacked something that I will illustrate with a story. When I was 22, I decided I could make a living as a mandolin player. I spent the following summer with a fiddle player who was vastly more talented and enthusiastic than I was and I realized that for all that, he would still be hard-pressed to support himself and probably would not be able to afford health insurance until he was old enough for Medicare. I knew I did not have a prayer and decided instead to do work for which I had some talent. For grant-writing, as for music or athletics or math, there is such a thing as talent and for some people, trying to make a living by writing grants is simply futile. My recent proposal convinced me, finally, that I am one of those people. Prof. Chasan-Taber would do her readers a favor if she told them that grant-writing requires talent, that everyone must decide at some point whether they have enough to justify cultivating it, and then giving frank but kind advice about how to make that decision. *(abstract via OpenAlex)*
SMMR Causal Inference · 发表于 2023-06-20
Yangqing Deng, Dongsheng Tu, Chris J O'Callaghan, Geoffrey Liu, Wei Xu
摘要
In clinical research, it is important to study whether certain clinical factors or exposures have causal effects on clinical and patient-reported outcomes such as toxicities, quality of life, and self-reported symptoms, which can help improve patient care. Usually, such outcomes are recorded as multiple variables with different distributions. Mendelian randomization (MR) is a commonly used technique for causal inference with the help of genetic instrumental variables to deal with observed and unobserved confounders. Nevertheless, the current methodology of MR for multiple outcomes only focuses on one outcome at a time, meaning that it does not consider the correlation structure of multiple outcomes, which may lead to a loss of statistical power. In situations with multiple outcomes of interest, especially when there are mixed correlated outcomes with different distributions, it is much more desirable to jointly analyze them with a multivariate approach. Some multivariate methods have been proposed to model mixed outcomes; however, they do not incorporate instrumental variables and cannot handle unmeasured confounders. To overcome the above challenges, we propose a two-stage multivariate Mendelian randomization method (MRMO) that can perform multivariate analysis of mixed outcomes using genetic instrumental variables. We demonstrate that our proposed MRMO algorithm can gain power over the existing univariate MR method through simulation studies and a clinical application on a randomized Phase III clinical trial study on colorectal cancer patients.
Biometrics Survey & Sampling · 发表于 2023-06-20
David Banks
摘要
Alicia Carriquiry, Judith Tanur, William Eddy, and Margaret Smykla (2022). Statistics in the Public Interest: In Memory of Stephen E. Fienberg. New York City: Springer. It is difficult to write this review. Steve was a friend, who died too young. Despite a life of contribution and community, he still had much to do. And the subsequent death of Joyce, his wife, from a brutal, futile, and stupid act of bigoted terrorism, inescapably darkens my memories of them both. But one of Steve's cardinal virtues was his commitment to getting a job done. And I am certainly delighted that so many of his friends and admirers have come together to honor his memory with this volume. Appropriately, it is a large book, and it contains multitudes. We all owe thanks to the editors, Alicia Carriquiry, Judy Tanur, Bill Eddy, and their special assistant, Margie Smykla. The chapter authors are outstanding researchers in their disparate fields. It is an easy inference that Steve's stature in the profession made it easy for the editors to recruit top-shelf contributors. The book has six technical sections, followed by a set of remembrances. The technical sections are Theory and Methods for Categorical Data, Bayesian Theory and Applications, Statistics and the Law, Causal Analyses, Surveys and Censuses, and Official Statistics. But for most sections, the chapters within cover a wide range of topics. There are a total of 23 such chapters, and each highlights a distinct aspect of Steve's work. The first section, Theory and Methods for Categorical Data, is typical. It has three chapters, each very different. The first, by Larry Wasserman, extends Steve's famously influential work on categorical data (Google Scholar reports that the 1974 book Discrete Multivariate Analysis: Theory and Practice, by Bishop, Fienberg, and Holland, has been cited 11,035 times as of March 13, 2023). Unsurprisingly, Larry goes in a fresh direction, looking at prediction rather than inference, and focusing on set-valued prediction rather than point estimates. It builds out both Steve's work on log-linear models and his research on exponential random graph models in a fresh way. The second chapter is on capture–recapture models and their use for estimating the number of unreported civilian casualties in conflict regions (the idea is that various groups, such as newspapers and NGOs, keep lists of casualties; if the lists have a lot of overlap, then there are few unreported deaths, but if not, then there are many—the overlap enables one to estimate the total number of casualties). The chapter is written by Daniel Manrique-Vallier, Patrick Ball, and Mauricio Sadinle. Daniel and Mauricio were Steve's Ph.D. students, and Patrick was the first person to apply capture–recapture techniques to casualty estimation in Guatemala. All three are leaders in the field, and their chapter traces the last 15 years of research, as strong assumptions, such as everyone having an equal probability of “capture” (being on a list), have been steadily weakened in ways that make the methodology far more realistic and practical. The third chapter is on algebraic statistics. Written by Elizabeth Gross, Vishesh Karwa, and Sonja Petrović, it reviews work that largely began in the 2000s, but which formalized and generalized Steve's geometric approach to contingency tables from the 1970s. The field is an important marriage of pure mathematics with statistics, and the authors provide a gentle introduction to some deep mathematics—including toric ideals for goodness-of-fit testing. In particular, the chapter is a Steve-centric description of algebraic statistics and emphasizes his work on log-linear network models and hypergraphs. The next four sections are similar. They have leaders in the field elaborating upon and reflecting upon the remarkable scope of Steve's research. It is an impressive tribute to an impressive statistician. The sixth section is somewhat different and merits a different kind of attention. It reviews Steve's almost unique role in official statistics. Much, but certainly not all of his activity on this front, grew from his work with the Committee on National Statistics. There had been highly placed statisticians on that committee before Steve, but he brought fresh energy to it and recognized that it could stimulate novel methodology. In particular, he pressed for cognitive studies to guide survey design and analysis, evidence-based public policy, and better ways of balancing the need for privacy with the value of the information in official records. This side of his contribution was clearly statistical, but often not has mathematical as his other work, and the section lays out the special challenges in this kind of work. This seventh section is a celebration of both Steve and his wife Joyce. For those who knew them, it is especially moving and inspirational. In those 12 pieces, I was delighted to learn new things about them and to be reminded of old stories. And, with wonderfully surprising appropriateness, Margie Smykla has the last word, with her section on “Reminiscences of Joyce.” For many years, Margie was the lead administrative assistant at Carnegie Mellon's Department of Statistics, and everyone who spent any time in its halls loved her. Over a remarkably long span of time, almost everyone in our field knew Steve and Joyce. In their different ways, they helped a huge number of people forward with their careers and with their lives. The first six sections discuss Steve's legacy of statistical research. But I think the seventh section shows that the larger contribution is the one they made jointly, building the next two, maybe three, generations of statisticians. *(abstract via OpenAlex)*
SMMR Causal Inference · 发表于 2023-06-20
Bas BL Penning de Vries, Rolf HH Groenwold
摘要
Unmeasured confounding is a well-known obstacle in causal inference. In recent years, negative controls have received increasing attention as a important tool to address concerns about the problem. The literature on the topic has expanded rapidly and several authors have advocated the more routine use of negative controls in epidemiological practice. In this article, we review concepts and methodologies based on negative controls for detection and correction of unmeasured confounding bias. We argue that negative controls may lack both specificity and sensitivity to detect unmeasured confounding and that proving the null hypothesis of a null negative control association is impossible. We focus our discussion on the control outcome calibration approach, the difference-in-difference approach, and the double-negative control approach as methods for confounding correction. For each of these methods, we highlight their assumptions and illustrate the potential impact of violations thereof. Given the potentially large impact of assumption violations, it may sometimes be desirable to replace strong conditions for exact identification with weaker, easily verifiable conditions, even when these imply at most partial identification of unmeasured confounding. Future research in this area may broaden the applicability of negative controls and in turn make them better suited for routine use in epidemiological practice. At present, however, the applicability of negative controls should be carefully judged on a case-by-case basis.