2023 年的期刊文章 · 第 21 页

共 1511 篇,按发表时间排,新的在前,每页 40 篇。这一页是 2023-03-27 到 2023-03-14。

按期刊、按主题看 →

第 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 页

SiM Survival Analysis · 发表于 2023-03-27

Adjusting for informative cluster size in pseudo‐value‐based regression approaches with clustered time to event data

Samuel Anyaso‐Samuel, Somnath Datta

摘要

Informative cluster size (ICS) arises in situations with clustered data where a latent relationship exists between the number of participants in a cluster and the outcome measures. Although this phenomenon has been sporadically reported in the statistical literature for nearly two decades now, further exploration is needed in certain statistical methodologies to avoid potentially misleading inferences. For inference about population quantities without covariates, inverse cluster size reweightings are often employed to adjust for ICS. Further, to study the effect of covariates on disease progression described by a multistate model, the pseudo‐value regression technique has gained popularity in time‐to‐event data analysis. We seek to answer the question: “How to apply pseudo‐value regression to clustered time‐to‐event data when cluster size is informative?” ICS adjustment by the reweighting method can be performed in two steps; estimation of marginal functions of the multistate model and fitting the estimating equations based on pseudo‐value responses, leading to four possible strategies. We present theoretical arguments and thorough simulation experiments to ascertain the correct strategy for adjusting for ICS. A further extension of our methodology is implemented to include informativeness induced by the intracluster group size. We demonstrate the methods in two real‐world applications: (i) to determine predictors of tooth survival in a periodontal study and (ii) to identify indicators of ambulatory recovery in spinal cord injury patients who participated in locomotor‐training rehabilitation.

SiM High-Dimensional Survival Analysis · 发表于 2023-03-25

Penalized maximum likelihood inference under the mixture cure model in sparse data

Changchang Xu, Shelley B. Bull

摘要

Introduction: When a study sample includes a large proportion of long‐term survivors, mixture cure (MC) models that separately assess biomarker associations with long‐term recurrence‐free survival and time to disease recurrence are preferred to proportional‐hazards models. However, in samples with few recurrences, standard maximum likelihood can be biased. Objective and Methods: We extend Firth‐type penalized likelihood (FT‐PL) developed for bias reduction in the exponential family to the Weibull‐logistic MC, using the Jeffreys invariant prior. Via simulation studies based on a motivating cohort study, we compare parameter estimates of the FT‐PL method to those by ML, as well as type 1 error (T1E) and power obtained using likelihood ratio statistics. Results: In samples with relatively few events, the Firth‐type penalized likelihood estimates (FT‐PLEs) have mean bias closer to zero and smaller mean squared error than maximum likelihood estimates (MLEs), and can be obtained in samples where the MLEs are infinite. Under similar T1E rates, FT‐PL consistently exhibits higher statistical power than ML in samples with few events. In addition, we compare FT‐PL estimation with two other penalization methods (a log‐F prior method and a modified Firth‐type method) based on the same simulations. Discussion: Consistent with findings for logistic and Cox regressions, FT‐PL under MC regression yields finite estimates under stringent conditions, and better bias‐and‐variance balance than the other two penalizations. The practicality and strength of FT‐PL for MC analysis is illustrated in a cohort study of breast cancer prognosis with long‐term follow‐up for recurrence‐free survival.

Biometrics Spatial & Time Series Latent Variables & Mixtures · 发表于 2023-03-25

Detecting the Spatial Clustering of Exposure–Response Relationships with Estimation Error: A Novel Spatial Scan Statistic

Wei Wang, Sheng Li, Tao Zhang, Fei Yin, Yue Ma

摘要

Detecting the spatial clustering of the exposure–response relationship (ERR) between environmental risk factors and health-related outcomes plays important roles in disease control and prevention, such as identifying highly sensitive regions, exploring the causes of heterogeneous ERRs, and designing region-specific health intervention measures. However, few studies have focused on this issue. A possible reason is that the commonly used cluster-detecting tool, spatial scan statistics, cannot be used for multivariate spatial datasets with estimation error, such as the ERR, which is often defined by a vector with its covariance estimated by a regression model. Such spatial datasets have been produced in abundance in the last decade, which suggests the importance of developing a novel cluster-detecting tool applicable for multivariate datasets with estimation error. In this work, by extending the classic scan statistic, we developed a novel spatial scan statistic called the estimation-error-based scan statistic (EESS), which is applicable for both univariate and multivariate datasets with estimation error. Then, a two-stage analytic process was proposed to detect the spatial clustering of ERRs in practical studies. A published motivating example and a simulation study were used to validate the performance of EESS. The results show that the clusters detected by EESS can efficiently reflect the clustering heterogeneity and yield more accurate ERR estimates by adjusting for such heterogeneity.

Biometrika Testing & Uncertainty · 发表于 2023-03-24

Hybrid confidence intervals for informative uniform asymptotic inference after model selection

A McCloskey

摘要

I propose a new type of confidence interval for correct asymptotic inference after using data to select a model of interest without assuming any model is correctly specified. This hybrid confidence interval is constructed by combining techniques from the selective inference and post-selection inference literatures to yield a short confidence interval across a wide range of data realizations. I show that hybrid confidence intervals have correct asymptotic coverage, uniformly over a large class of probability distributions that do not bound scaled model parameters. I illustrate the use of these confidence intervals in the problem of inference after using the lasso objective function to select a regression model of interest and provide evidence of their desirable length and coverage properties in small samples via a set of Monte Carlo experiments that entail a variety of different data distributions as well as an empirical application to the predictors of diabetes disease progression.

Biometrics Meta-Analysis & Data Integration · 发表于 2023-03-24

An Efficient Data Integration Scheme for Synthesizing Information from Multiple Secondary Datasets for the Parameter Inference of the Main Analysis

Chixiang Chen, Ming Wang, Shuo Chen

摘要

Many observational studies and clinical trials collect various secondary outcomes that may be highly correlated with the primary endpoint. These secondary outcomes are often analyzed in secondary analyses separately from the main data analysis. However, these secondary outcomes can be used to improve the estimation precision in the main analysis. We propose a method called multiple information borrowing (MinBo) that borrows information from secondary data (containing secondary outcomes and covariates) to improve the efficiency of the main analysis. The proposed method is robust against model misspecification of the secondary data. Both theoretical and case studies demonstrate that MinBo outperforms existing methods in terms of efficiency gain. We apply MinBo to data from the Atherosclerosis Risk in Communities study to assess risk factors for hypertension.

Biometrics Latent Variables & Mixtures · 发表于 2023-03-23

Supervised Convex Clustering

Minjie Wang, Tianyi Yao, Genevera I. Allen

摘要

Clustering has long been a popular unsupervised learning approach to identify groups of similar objects and discover patterns from unlabeled data in many applications. Yet, coming up with meaningful interpretations of the estimated clusters has often been challenging precisely due to their unsupervised nature. Meanwhile, in many real-world scenarios, there are some noisy supervising auxiliary variables, for instance, subjective diagnostic opinions, that are related to the observed heterogeneity of the unlabeled data. By leveraging information from both supervising auxiliary variables and unlabeled data, we seek to uncover more scientifically interpretable group structures that may be hidden by completely unsupervised analyses. In this work, we propose and develop a new statistical pattern discovery method named supervised convex clustering (SCC) that borrows strength from both information sources and guides towards finding more interpretable patterns via a joint convex fusion penalty. We develop several extensions of SCC to integrate different types of supervising auxiliary variables, to adjust for additional covariates, and to find biclusters. We demonstrate the practical advantages of SCC through simulations and a case study on Alzheimer's disease genomics. Specifically, we discover new candidate genes as well as new subtypes of Alzheimer's disease that can potentially lead to better understanding of the underlying genetic mechanisms responsible for the observed heterogeneity of cognitive decline in older adults.

SMMR Survival Analysis High-Dimensional Nonparametric & Semiparametric · 发表于 2023-03-23

Identifying regions of interest in mammogram images

Shu Jiang, Jiguo Cao, Graham A. Colditz

摘要

Screening mammography is the primary preventive strategy for early detection of breast cancer and an essential input to breast cancer risk prediction and application of prevention/risk management guidelines. Identifying regions of interest within mammogram images that are associated with 5- or 10-year breast cancer risk is therefore clinically meaningful. The problem is complicated by the irregular boundary issue posed by the semi-circular domain of the breast area within mammograms. Accommodating the irregular domain is especially crucial when identifying regions of interest, as the true signal comes only from the semi-circular domain of the breast region, and noise elsewhere. We address these challenges by introducing a proportional hazards model with imaging predictors characterized by bivariate splines over triangulation. The model sparsity is enforced with the group lasso penalty function. We apply the proposed method to the motivating Joanne Knight Breast Health Cohort to illustrate important risk patterns and show that the proposed method is able to achieve higher discriminatory performance.

SiM Survival Analysis Diagnostics & Prediction High-Dimensional · 发表于 2023-03-23

Disease progression based feature screening for ultrahigh‐dimensional survival‐associated biomarkers

Mengjiao Peng, Liming Xiang

摘要

Summary: The increased availability of ultrahigh‐dimensional biomarker data and the high demand of identifying biomarkers importantly related to survival outcomes made feature screening methods commonplace in the analysis of cancer genome data. When survival outcomes include endpoints of overall survival (OS) and time‐to‐progression (TTP), a high concordance is typically found in both endpoints in cancer studies, namely, patients' OS would most likely be extended when tumour progression is delayed. Existing screening procedures are often performed on a single survival endpoint only and may result in biased selection of features for OS in ignorance of disease progression. We propose a novel feature screening method by incorporating information of TTP into the selection of important biomarker predictors for more accurate inference of OS subsequent to disease progression. The proposal is based on the rank of correlation between individual features and the conditional distribution of OS given observations of TTP. It is advantageous for its flexible model nature, which requires no marginal model assumption for each endpoint, and its minimal computational cost for implementation. Theoretical results show its ranking consistency, sure screening and false rate control properties. Simulation results demonstrate that the proposed screener leads to more accurate feature selection than the method without considering the prior observations of disease progression. An application to breast cancer genome data illustrates its practical utility and facilitates disease classification using selected biomarker predictors.

JASA Networks & Graphs Latent Variables & Mixtures · 发表于 2023-03-22

Network Inference Using the Hub Model and Variants

Zhibing He, Yunpeng Zhao, Peter Bickel, Charles Weko, Dan Cheng, Jirui Wang

摘要

Statistical network analysis primarily focuses on inferring the parameters of an observed network. In many applications, especially in the social sciences, the observed data is the groups formed by individual subjects. In these applications, the network is itself a parameter of a statistical model. Zhao and Weko propose a model-based approach, called the hub model, to infer implicit networks from grouping behavior. The hub model assumes that each member of the group is brought together by a member of the group called the hub. The set of members which can serve as a hub is called the hub set. The hub model belongs to the family of Bernoulli mixture models. Identifiability of Bernoulli mixture model parameters is a notoriously difficult problem. This article proves identifiability of the hub model parameters and estimation consistency under mild conditions. Furthermore, this article generalizes the hub model by introducing a model component that allows hubless groups in which individual nodes spontaneously appear independent of any other individual. We refer to this additional component as the null component. The new model bridges the gap between the hub model and the degenerate case of the mixture model—the Bernoulli product. Identifiability and consistency are also proved for the new model. In addition, a penalized likelihood approach is proposed to estimate the hub set when it is unknown. Supplementary materials for this article are available online. *(abstract via OpenAlex)*

JRSSB Bayesian Methods · 发表于 2023-03-22

Monotone response surface of multi-factor condition: estimation and Bayes classifiers

Ying Kuen Cheung, Keith M Diaz

摘要

We formulate the estimation of monotone response surface of multiple factors as the inverse of an iteration of partially ordered classifier ensembles. Each ensemble (called product-of-independent-probability-escalation (PIPE)-classifiers) is a projection of Bayes classifiers on the constrained space. We prove that the inverse of PIPE-classifiers (iPIPE) exists, and propose algorithms to efficiently compute iPIPE by reducing the space over which optimisation is conducted. The methods are applied in analysis and simulation settings where the surface dimension is higher than what the isotonic regression literature typically considers. Simulation shows that iPIPE-based credible intervals achieve nominal coverage probability and are more precise compared to unconstrained estimation.

Biometrics Machine Learning Genetics & Omics · 发表于 2023-03-22

Gene expression data analysis: a statistical and machine learning perspective. PankajBarah, DhrubaKumar Bhattacharyya, JugalKumar Kalita (2022). Boca Raton, Florida and London. CRC Press; Taylor and Francis. https://doi.org/10.1201/9780429322655

Amrita Chattopadhyay

摘要

Gene expression profiling and analysis is an important research area that helps fathoming the underlying molecular mechanisms of complex diseases. Current availability of high-throughput technologies, laboratory investigation protocols along with bioinformatics methodologies allow identification of transcripts and quantification of gene function, toward decoding dynamic biological processes played by RNA as the intermediary step between the genome and the proteome. Explicit understanding of the above enables reliable scientific conclusions. RNA sequencing (RNA-seq) technology, which was introduced in 2008 (Jazayeri et al., 2015), has revolutionized exponentially in the past decade and have evolved from being only a gene expression quantification tool to a powerful resource for discovery of new transcripts, characterization of alternative splicing variants or new cell types (via single-cell RNA sequencing (scRNA-seq)), and biomarker detection for drug screening (Hong et al., 2020). Applications of RNA-seq is not only restricted to the genomics community, but expands to being utilized for developing toolkits potentially serving varied scientific purposes for the research community (Stark et al., 2019). Over the years, multitude of protocols and analysis pipelines for RNA-seq have been published, making it an intriguing and challenging discipline for new users to delve into. Therefore, it is important to gain perspectives into the fundamental concepts of biology and functional genomics, data types, analysis strategies, existing methodologies: their advantages and caveats, and an explanation of the results that are revealed at the end of each analytic route. This would potentially allow readers a holistic knowledge about the dynamic biological processes at both cellular and organismal level. This book attempts to provide an easy understanding and guidance for users who have no prior experience of gene expression analysis and lacks the concepts needed to comprehend it. The book consists of a total of eight chapters which can be broadly classified into four distinct sections. The first section includes an introductory chapter (Chapter 1) with brief explanations of each of the topics that have been touched, along the way. The second section contains Chapters 2–4 comprising mostly the fundamental concepts and backgrounds required for readers to learn gene-expression analysis. Chapter 2 introduces systems theory, particularly its application in the hierarchical information flow of genetic information within complex biological systems via top-down or bottom-up models, pertaining to different Omics data. It then proceeds to explain in a succinct way, the central dogma, DNA-replication, transcription and translation, and the ambiguities of central dogma, that led to the understanding of reverse transcriptase and complimentary-DNAs (cDNAs). Chapter 3 focuses on gene expression, its history, descriptions on low-throughput (northern blotting, ribonuclease protection assay, qRT-PCR (real-time quantitative reverse transcription-PCR), and SAGE (serial analysis of gene expression)), and high-throughput technologies (microarray, RNA-seq, scRNA-seq); types of RNA-Seq data and gene-expression databases. The authors use small paragraphs to explain each of the topics, summarizing the relevant information. Procedures to access datasets are important information for freshers of this discipline, which could have added value to this chapter, however, this information was missing. Chapter 4 discusses statistical and machine learning foundations as preceding knowledge for gene expression analysis. The statistical concepts are explained briefly with equations and comprises topics such as probability theory, hypothesis testing methods, exact tests, common data distributions, false discovery rates and, maximal likelihood estimation. Although examples for some of the abovementioned topics have been provided, better illustrations of applications that could have provided more clarity for the readers are lacking. The next part includes machine learning topic that spans almost 73 pages and is quite comprehensive and well explained, providing readers with a glimpse into variety of models, training–testing–validation and evaluation procedures, learning methods, similarity and dissimilarity measures, and partitioning and clustering approaches. Machine learning applications in healthcare services assists in crucial aspects such as clinical decision making, detection of diseases, and personalized medicine toward saving lives. As machine learning is utilized profusely toward identification of biomarkers, it is an important technique that needs to be understood well. The goal is to use machine-learning based equipment and software that can replace or support the existing diagnostic and treatment options for an overall improvement of the healthcare systems while lowering the cost of care. Hence, this is an important and relevant topic and the authors have provided well researched material towards educating the readers and providing them a foundation course toward application into gene-expression data. The third section is composed of Chapters 5 and 6 that discusses methods and techniques for conducting co-expression analysis and differential gene expression analysis, respectively. Co-expression analysis enables identification of genes with similar expression patterns that regulate various biological conditions or phenotypes, based on certain statistical proximity metrics. On the other hand, differential expression analysis enables identification of genes that demonstrate quantitative changes in expression levels between different experimental groups. Both the above methods are important and profusely used in disease-associated transcriptome profiling toward prioritizing regulatory candidate disease genes. Methodologies and approaches to identify co-expression patterns were primarily the focus of Chapter 5. Explanations on types and patterns of gene co-expression, and measures to identify co-expressed genes are provided briefly with clear examples. Descriptions are provided for various aspects and stages of co-expression analysis pipelines such as preprocessing, clustering, query-specific cluster ranking and selection, cluster-based analysis, and validation for different data types (microarray, and sequencing data). In-depth discussions are then provided for each of the above-mentioned analysis approaches with respect to the data type to be processed; such as microarray data, RNAseq data, or scRNAseq data. This was impressively done and can be used as a reference while conducting data analysis on real genomic data. Chapter 6, similarly, gives descriptions of analysis pipelines comprising of detailed preprocessing procedures followed by methodologies to conduct differential gene expression (DE) analysis. A list of statistical tests, both parametric and non-parametric along with numerical descriptions and simplified examples are provided. Validation procedures and downstream network analysis are also described. R-packages and other tools and repositories that allow both co-expression pattern and DE analysis are also described in a succinct way. The authors have succeeded to cover all necessary information, competently through simple and easy to understand descriptions which make the information appropriate for readers without any prior experience in this field. The final section includes Chapters 7 and 8. Chapter 7 focuses on tools and repositories related to the topic, that are popularly used by bioinformaticians and computational biologists. Microarray, RNA-seq, scRNAseq techniques lead to clinically relevant data and large repositories, that houses multitude of samples. The unprecedented rate at which such scientific knowledge is being generated and shared, computational, mathematical and statistical expertise are important toward archiving, retrieving, and analyzing such huge biological data. This chapter provides a comprehensively curated content of tools, databases, and webservers for various steps of the pipeline leading to gene-expression analysis. The contents of this chapter are of extreme relevance and significantly contribute to the value of the book. The chapter is divided into subtopics such as preprocessing tools, discretization techniques, dimensionality reduction and quality control steps, and gene expression analysis tools. Under each topic, the authors provide a detailed description of an extensive list of software and repositories, with tabulated information, and comparative comments that would guide readers to choose the relevant tool based on the suitability and the aim of their study. Furthermore, hyperlinks are provided for convenience of readers. Finally, Chapter 8 summarizes the contents by highlighting the major topics that is handled through this book, notably, co-expression analysis, differential co-expression analysis, and differential expression analysis using statistical and machine-learning approaches. Finally, the authors conclude by discussing about the challenges and issues that are associated with transcriptome research. The target audience of this book is future computational biologists who are venturing into the world of genomics and clinico-genomics studies. The material is relevant and important. The topics are organized well. The writing is smooth and easy to understand even for a non-native English speaker. The topics does not go into too much depth but provides enough information to have an overall understanding. The chapters on fundamentals of machine learning and software and tools, deserve special mention for the carefully collected information and the extensive research that went behind them. The book however suffers from poorly represented and amateurish figures and illustrations. Good illustrations are a way to attract readers especially the inexperienced ones, into exploring the material. In my opinion, the book may fall short in this area and as a result may lose some potential readers. Also, more example problems with example data illustrating each of the contents would have really enhanced the material. Lastly, some typographical error exists and should be rectified in the future editions. Despite few shortcomings, the book is a good reference material in this rapidly evolving and important field of study. I would recommend it as a reference textbook for undergraduate and masters’ students who are pursuing genetics, biostatistics, and bioinformatics. I appreciate that I was given an opportunity to review the book and I praise the authors for their excellent achievement. *(abstract via OpenAlex)*

Biometrics Testing & Uncertainty Clinical Trials Bayesian Methods · 发表于 2023-03-22

Confidence intervals for discrete data in clinical research By VivekPradhan, Ashis K.Gangopadhyay, Sandeep M.Menon, CynthiaBasu, TathagataBanerjee, Chapman and Hall. 2021. pp. 226. $119.95. (hbk). ISBN: 978‐1138048980

Naitee Ting

摘要

In the clinical development of new drugs, research pharmaceutical industry and biotech companies (known as sponsors) design and execute clinical trials in order to evaluate the efficacy and safety of the study drug. Clinical data are observed and collected from trial participants. These data are analyzed and organized to support clinical study reports. Sponsors that develop the new drugs will submit these study reports to U.S. Food and Drug Administration (FDA) for review. If FDA considers this new drug is efficacious, and, relatively safe, then this new drug would be approved for general patient population. Discrete data are very common in clinical trials. One example is known as “responder analysis.” After the completion of a clinical trial, the physician evaluates each patient and classifies the patient as a “responder” or a “nonresponder.” This is a binary classification of treatment results for each patient. The responder analysis is to study the proportion of responders among all trial participants. If this proportion is high, then it is likely that the study drug can be approved by FDA. In most of clinical trials, two treatment groups (one group of patients receives the test drug treatment, and the other receives control treatment) are compared for efficacy. If it turns out that more proportion of responders is observed from the test drug group than the control group, then the test drug can be approved for general patient population. In this book, Chapters 4 and 5 provide extensive materials to cover treatment comparisons of two independent binomial proportions. These methods can be very useful to practitioners. In some situations, when developing drugs for the treatment of rare diseases, because of difficulty in patient recruitment, “single arm trial” can be found in practice. A single arm trial means the clinical trial treats every participant only with the study medication, without any control group. When this is the case, methods for analysis of one binomial population can be very useful. Another application can be found in crossover clinical trials, or matched case-control epidemiology studies—these research work can be based on the paired binomial data. Chapter 3 of this book covers the one binomial proportion situation, and Chapter 6 focuses on paired binomial proportions. In clinical trial applications, sometimes the interest is in recurring events. For example, in treatment of diabetic patients, hypoglycemia can be a very severe adverse event. In the duration of the study treatment, most patients are not expected to develop hypoglycemia. Some patients could develop one episode. Other patients may develop repeated events. In this application, confidence intervals on count data can be useful. Another example could be chronic obstructive pulmonary disease (COPD) exacerbation. In treating COPD patients, the number of exacerbations experienced by patients in a treatment group may serve as one of the primary efficacy endpoints. This is another application of count data. Chapter 7 of this book covers count data and confidence intervals on rate. In development of new drugs, typically many molecules are discovered as drug candidates. Only the best candidates would have the opportunity entering the clinical development stage. Clinical trials are divided into four phases—Phases I, II, III, and IV, in a sequential manner. Every phase of clinical development can be viewed as a process of weeding out unsuccessful drug candidates. In other words, only the very best drug candidate gets the chance to enter the later phase clinical development, and only the ones that can objectively and scientifically demonstrate both drug efficacy and drug safety can be approved by FDA. Hence for every drug candidate at every stage of development process, a “Go/NoGo” decision has to be made about the candidate as to whether it deserves to be progressed into the next step of development. This “Go/NoGo” decision can be thought of as a confirmatory question in drug development. However, every stage of new drug development can also be viewed as a learning process. Development team members are always asking many questions about the drug candidate. All of these questions are exploratory in nature—people are curious, and it is always important to learn the efficacy and safety associated with each drug candidate. Based on this understanding, every clinical question can be divided into a confirmatory question or an exploratory question. The confirmatory question is about the “Go/NoGo” decision, and all other questions are exploratory. In statistical applications, the method to deal with a confirmatory question is hypothesis testing, and the method to deal with an exploratory question is estimation—either point estimation or interval estimation. Another consideration is about the noninferiority decisions. This happens in active-controlled clinical trials. In other words, the Phase III clinical trial is designed to compare a test drug against another drug for the same indication, and this other drug has already been approved by regulatory agencies. Such an approved drug is known as an active control agent. Noninferiority trials are designed with a prespecified margin, which must have been agreed between the sponsor and the regulatory agency. The Phase III comparison is to demonstrate that the lower confidence limit of treatment difference (test drug mean subtract active control mean) should be above the negative value of this margin (when a positive value means better efficacy). Although the comparison is based on the one-sided confidence limit, the decision itself is of a nature as the one-sided hypothesis testing. In other words, the efficacy of the test drug has to be above a predefined quantity, in order for the test drug to be approvable. Again, this approval decision is fundamentally a “Go/NoGo” decision, the only difference is that the noninferiority decision is expressed in terms of a “one-sided confidence interval.” Because the only confirmatory question in drug development is about a “Go/NoGo” decision, most of other clinical questions are exploratory. Estimation is more useful in most of clinical applications. Also, confidence intervals are much more informative than a simple p-value. Chapter 2 of this book covers the p-value discussion from the American Statistical Association (ASA) statement. This is a very valid and relevant point, indicating that estimation, either point estimation or interval estimation, is more useful in applications than statistical hypothesis testing. This book is about confidence intervals. In fact, the emphasis on the confidence interval makes this book very useful for practitioners. There are seven chapters in this book. The first chapter is a brief overview, which covers both the frequentist approach and the Bayesian approach of constructing confidence intervals on discrete data. This chapter also mentions statistical inference, and how statistical thinking established the scientific foundation of clinical trials and modern drug development. Chapter 2 discusses various viewpoints on the “p-value,” specifically, it brings up the ASA statement on the p-value. As mentioned earlier, in clinical trial applications, most of the scientific questions are exploratory in nature. Hence confidence intervals are much more useful than p-values. It is very sensible that this book focuses on confidence interval approaches to handle clinical research questions. In statistical applications to solve real-world problems, confidence interval approaches are much more useful than p-values. The main body of this book includes Chapters 3–7. In these chapters, the authors present a wide range of statistical methods for calculating confidence intervals. For each method, this book introduces the theoretical derivation, the background, and the implementation using either SAS code or R. Many methods covered in this book have existing SAS procedures available. On the other hand, a lot of methods have to be programmed using SAS or R. For most of the applications, the book brings a numerical example (mostly based on clinical trial data) and uses SAS or R to achieve the computation. The results are illustrated to help readers in understanding how each method can be implemented in practice. Furthermore, performance of these methods is also displayed and compared using coverage probabilities obtained from the calculated confidence intervals. Chapter 3 provides methods for calculating confidence intervals on one sample binomial proportion. Broadly speaking, there are three major categories in establishing confidence intervals on one binomial proportion—asymptotic approaches, exact methods, and Bayesian intervals. Within each category, the authors introduce many methods and offer details of each method. Some of these methods are well known and available for decades, and some are relatively new. The authors provide a rather comprehensive list of methods. As mentioned above, in the development of drugs treating rare diseases, because of difficulty in patient recruitment, “single arm trial” can be found in practice. When this is the case, methods introduced in this chapter can be useful to practitioners. Chapter 4 discusses confidence intervals on differences between two independent binomial proportions. Differences between two independent proportions is frequently observed in clinical trials. As mentioned earlier, “responder analysis” is a common practice in many clinical settings. When comparing the drug efficacy of two groups (e.g., test treatment vs. control), development teams often study the differences in the proportion of responders between these two treatments. This chapter focuses on the difference p1– p2 and proposes ways to construct confidence intervals on this difference. In this chapter, the authors introduce many methods to perform the calculation. It starts with asymptotic methods and then moves into exact methods. Also introduced in Chapter 4 of this book is a Bayesian approach to compute credible intervals. Relative risk or risk ratio (RR) is another commonly used metric for comparing two independent binomial proportions in clinical trial applications. Chapter 5 covers confidence interval methods for RR. Similar to Chapter 4, the book starts with asymptotic methods and then moves into exact methods. Furthermore, Chapter 5 also introduces approaches to establish Bayesian credible intervals. Chapter 6 presents methods to calculate confidence intervals for paired binomial proportions. The book mentioned that in crossover trials there may be applications of paired proportions. However, in my personal experience, I have never seen such an application. Most of crossover clinical trials are designed for Phase I bioequivalence studies (or other Phase I trials including bioavailability, food effect, and/or drug–drug-interaction trials). In these clinical trials, the primary variables of interest are pharmacokinetic (PK) parameters. PK parameters are generally believed to follow log-normal distribution, and they tend to be continuous variables. I can imagine that in certain situations, crossover trials are applied to study drug efficacy like in Phase II or III applications and that the primary efficacy endpoint is a binary variable (such as responder status). But those applications are not common in drug development practice. On the other hand, in epidemiology research, case control is an important study type. For case-control studies, first a number of cases (patients with the disease under study, e.g., a certain type of cancer) are identified and recruited into the study. Next, for each case, find one or a few matching controls. A control subject is a person without this disease, but with similar characteristics (e.g., same gender, similar age, occupation, …) as the case. After the cases and controls are identified, data are collected from these participants and, finally, these data are analyzed. Binomial data are very common in case-control studies. Therefore, the application of this book can also include epidemiology studies, or, more generally, the book could be useful for public health applications. In Chapter 6, the authors provide a number of asymptotic approaches including both frequentist and Bayesian methods. It also talks about exact confidence intervals for paired binomial proportions. One more section of Chapter 6 discusses confidence intervals with missing data. Finally, Chapter 7 presents confidence interval methods for rate. This chapter offers a rather comprehensive treatment of count data. Again, count data are an important data type in clinical research. A reference of confidence intervals on rate parameter observed from count data is useful. The chapter starts with Poisson distribution and focuses on the confidence interval of the rate parameter λ. Then the chapter covers confidence intervals for mean and other count data models. Before wrapping up this chapter, the authors cover Bayesian credible intervals. Overall speaking, this book delivered what the authors hoped to achieve—"Confidence Intervals for Discrete Data in Clinical Research." This book provides a comprehensive review of existing methods in constructing confidence intervals for binary and count data because these are the discrete data most frequently used in clinical research. Other types of discrete data such as nominal data, multinominal data, or general categorical data do not usually appear in clinical trials. In fact, the scope of application for this book can go beyond clinical trials. This book can also be a very useful reference tool for public health applications. Generally speaking, public health covers a much wider scope and includes many more practitioners. Most of the public health studies do not use randomization. Examples of these studies are epidemiology studies (including most of observational studies), electronic medical records, registries, and medical data collected from insurance companies like claim databases, and so on. In other words, this book not only serves the readers they it intended to serve but can also help potentially a much broader readership. One strength of this book is that it covers a wide range of methods with very good reference articles (published between 1927 up to 2020, pp 215–222). This feature makes this book to be one of the very useful references on this topic. For practitioners engage in clinical research, epidemiology, or public health, this book can be a very helpful tool. *(abstract via OpenAlex)*

Biometrika Clinical Trials · 发表于 2023-03-21

Scalable subsampling: computation, aggregation and inference

Dimitris N Politis

摘要

Subsampling has seen a resurgence in the big data era where the standard, full-resample size bootstrap can be infeasible to compute. Nevertheless, even choosing a single random subsample of size b can be computationally challenging with both b and the sample size n being very large. This paper shows how a set of appropriately chosen, nonrandom subsamples can be used to conduct effective, and computationally feasible, subsampling distribution estimation. Furthermore, the same set of subsamples can be used to yield a procedure for subsampling aggregation, also known as subagging, that is scalable with big data. Interestingly, the scalable subagging estimator can be tuned to have the same, or better, rate of convergence than that of θ^n. Statistical inference could then be based on the scalable subagging estimator instead of the original θ^n.

SiM Survival Analysis Nonparametric & Semiparametric · 发表于 2023-03-21

Nonparametric estimation of marked survival data in the presence of dependent censoring

Busola Sanusi, Jianwen Cai, Michael G. Hudgens

摘要

We consider nonparametrically estimating the joint distribution of a survival time and mark variable, where the survival time is subject to right censoring and the mark variable is only observed when the survival time is not censored. The possibility of dependent censoring is allowed for using inverse probability of censoring weights. The proposed estimator is shown to be consistent and asymptotically normal. Finite sample behavior of the proposed methods are investigated via simulation study. Finally, we illustrate the nonparametric estimator from a recent HIV vaccine efficacy trial.

Biometrics Causal Inference Longitudinal Data · 发表于 2023-03-21

Longitudinal Incremental Propensity Score Interventions for Limited Resource Settings

Aaron L. Sarvet, Kerollos N. Wanis, Jessica G. Young, Roberto Hernandez-Alejandro, Mats J. Stensrud

摘要

Many real-life treatments are of limited supply and cannot be provided to all individuals in the population. For example, patients on the liver transplant waiting list usually cannot be assigned a liver transplant immediately at the time they reach highest priority because a suitable organ is not immediately available. In settings with limited supply, investigators are often interested in the effects of treatment strategies in which a limited proportion of patients receive an organ at a given time, that is, treatment regimes satisfying resource constraints. Here, we describe an estimand that allows us to define causal effects of treatment strategies that satisfy resource constraints: incremental propensity score interventions (IPSIs) for limited resources. IPSIs flexibly constrain time-varying resource utilization through proportional scaling of patients' natural propensities for treatment, thereby preserving existing propensity rank ordering compared to the status quo. We derive a simple class of inverse-probability-weighted estimators, and we apply one such estimator to evaluate the effect of restricting or expanding utilization of “increased risk” liver organs to treat patients with end-stage liver disease.

JRSSB Causal Inference · 发表于 2023-03-21

Causal inference on distribution functions

Zhenhua Lin, Dehan Kong, Linbo Wang

摘要

Understanding causal relationships is one of the most important goals of modern science. So far, the causal inference literature has focused almost exclusively on outcomes coming from the Euclidean space Rp. However, it is increasingly common that complex datasets are best summarized as data points in nonlinear spaces. In this paper, we present a novel framework of causal effects for outcomes from the Wasserstein space of cumulative distribution functions, which in contrast to the Euclidean space, is nonlinear. We develop doubly robust estimators and associated asymptotic theory for these causal effects. As an illustration, we use our framework to quantify the causal effect of marriage on physical activity patterns using wearable device data collected through the National Health and Nutrition Examination Survey.

SiM Longitudinal Data Survival Analysis Diagnostics & Prediction · 发表于 2023-03-20

A comparison of two approaches to dynamic prediction: Joint modeling and landmark modeling

Wenhao Li, Liang Li, Brad C. Astor

摘要

Summary: Joint modeling and landmark modeling are two mainstream approaches to dynamic prediction in longitudinal studies, that is, the prediction of a clinical event using longitudinally measured predictor variables available up to the time of prediction. It is an important research question to the methodological research field and also to practical users to understand which approach can produce more accurate prediction. There were few previous studies on this topic, and the majority of results seemed to favor joint modeling. However, these studies were conducted in scenarios where the data were simulated from the joint models, partly due to the widely recognized methodological difficulty on whether there exists a general joint distribution of longitudinal and survival data so that the landmark models, which consists of infinitely many working regression models for survival, hold simultaneously. As a result, the landmark models always worked under misspecification, which caused difficulty in interpreting the comparison. In this paper, we solve this problem by using a novel algorithm to generate longitudinal and survival data that satisfies the working assumptions of the landmark models. This innovation makes it possible for a “fair” comparison of joint modeling and landmark modeling in terms of model specification. Our simulation results demonstrate that the relative performance of these two modeling approaches depends on the data settings and one does not always dominate the other in terms of prediction accuracy. These findings stress the importance of methodological development for both approaches. The related methodology is illustrated with a kidney transplantation dataset.

Biometrics Survival Analysis Bayesian Methods Nonparametric & Semiparametric · 发表于 2023-03-18

Nonparametric Failure Time: Time-to-Event Machine Learning with Heteroskedastic Bayesian Additive Regression Trees and Low Information Omnibus Dirichlet Process Mixtures

Rodney A. Sparapani, Brent R. Logan, Martin J. Maiers, Purushottam W. Laud, Robert E. McCulloch

摘要

Many popular survival models rely on restrictive parametric, or semiparametric, assumptions that could provide erroneous predictions when the effects of covariates are complex. Modern advances in computational hardware have led to an increasing interest in flexible Bayesian nonparametric methods for time-to-event data such as Bayesian additive regression trees (BART). We propose a novel approach that we call nonparametric failure time (NFT) BART in order to increase the flexibility beyond accelerated failure time (AFT) and proportional hazard models. NFT BART has three key features: (1) a BART prior for the mean function of the event time logarithm; (2) a heteroskedastic BART prior to deduce a covariate-dependent variance function; and (3) a flexible nonparametric error distribution using Dirichlet process mixtures (DPM). Our proposed approach widens the scope of hazard shapes including nonproportional hazards, can be scaled up to large sample sizes, naturally provides estimates of uncertainty via the posterior and can be seamlessly employed for variable selection. We provide convenient, user-friendly, computer software that is freely available as a reference implementation. Simulations demonstrate that NFT BART maintains excellent performance for survival prediction especially when AFT assumptions are violated by heteroskedasticity. We illustrate the proposed approach on a study examining predictors for mortality risk in patients undergoing hematopoietic stem cell transplant (HSCT) for blood-borne cancer, where heteroskedasticity and nonproportional hazards are likely present.

JASA Causal Inference Genetics & Omics · 发表于 2023-03-17

Causal Inference in Transcriptome-Wide Association Studies with Invalid Instruments and GWAS Summary Data

Haoran Xue, Xiaotong Shen, Wei Pan

摘要

Transcriptome-wide association studies (TWAS) have recently emerged as a popular tool to discover (putative) causal genes by integrating an outcome GWAS dataset with another gene expression/transcriptome GWAS (called eQTL) dataset. In our motivating and target application, we'd like to identify causal genes for low-density lipoprotein cholesterol (LDL), which is crucial for developing new treatments for hyperlipidemia and cardiovascular diseases. The statistical principle underlying TWAS is (two-sample) two-stage least squares (2SLS) using multiple correlated SNPs as instrumental variables (IVs); it is closely related to typical (two-sample) Mendelian randomization (MR) using independent SNPs as IVs, which is expected to be impractical and lower-powered for TWAS (and some other) applications. However, often some of the SNPs used may not be valid IVs, e.g. due to the widespread pleiotropy of their direct effects on the outcome not mediated through the gene of interest, leading to false conclusions by TWAS (or MR). Building on recent advances in sparse regression, we propose a robust and efficient inferential method to account for both hidden confounding and some invalid IVs via two-stage constrained maximum likelihood (2ScML), an extension of 2SLS. We first develop the proposed method with individual-level data, then extend it both theoretically and computationally to GWAS summary data for the most popular two-sample TWAS design, to which almost all existing robust IV regression methods are however not applicable. We show that the proposed method achieves asymptotically valid statistical inference on causal effects, demonstrating its wider applicability and superior finite-sample performance over the standard 2SLS/TWAS (and MR). We apply the methods to identify putative causal genes for LDL by integrating large-scale lipid GWAS summary data with eQTL data. *(abstract via PubMed)*

SiM Machine Learning · 发表于 2023-03-17

Boosting multivariate structured additive distributional regression models

Annika Strömer, Nadja Klein, Christian Staerk, Hannah Klinkhammer, Andreas Mayr

摘要

We develop a model‐based boosting approach for multivariate distributional regression within the framework of generalized additive models for location, scale, and shape. Our approach enables the simultaneous modeling of all distribution parameters of an arbitrary parametric distribution of a multivariate response conditional on explanatory variables, while being applicable to potentially high‐dimensional data. Moreover, the boosting algorithm incorporates data‐driven variable selection, taking various different types of effects into account. As a special merit of our approach, it allows for modeling the association between multiple continuous or discrete outcomes through the relevant covariates. After a detailed simulation study investigating estimation and prediction performance, we demonstrate the full flexibility of our approach in three diverse biomedical applications. The first is based on high‐dimensional genomic cohort data from the UK Biobank, considering a bivariate binary response (chronic ischemic heart disease and high cholesterol). Here, we are able to identify genetic variants that are informative for the association between cholesterol and heart disease. The second application considers the demand for health care in Australia with the number of consultations and the number of prescribed medications as a bivariate count response. The third application analyses two dimensions of childhood undernutrition in Nigeria as a bivariate response and we find that the correlation between the two undernutrition scores is considerably different depending on the child's age and the region the child lives in.

SiM Bayesian Methods Nonparametric & Semiparametric Clinical Trials · 发表于 2023-03-17

A Bayesian approach to study design and analysis with type I error rate control for response variables of mixed types

Ethan M. Alt, Matthew A. Psioda, Joseph G. Ibrahim

摘要

There has been increased interest in the design and analysis of studies consisting of multiple response variables of mixed types. For example, in clinical trials, it is desirable to establish efficacy for a treatment effect in primary and secondary outcomes. In this article, we develop Bayesian approaches for hypothesis testing and study planning for data consisting of multiple response variables of mixed types with covariates. We assume that the responses are correlated via a Gaussian copula, and that the model for each response is, marginally, a generalized linear model (GLM). Taking a fully Bayesian approach, the proposed method enables inference based on the joint posterior distribution of the parameters. Under some mild conditions, we show that the joint distribution of the posterior probabilities under any Bayesian analysis converges to a Gaussian copula distribution as the sample size tends to infinity. Using this result, we develop an approach to control the type I error rate under multiple testing. Simulation results indicate that the method is more powerful than conducting marginal regression models and correcting for multiplicity using the Bonferroni‐Holm Method. We also develop a Bayesian approach to sample size determination in the presence of response variables of mixed types, extending the concept of probability of success (POS) to multiple response variables of mixed types.

SMMR Survival Analysis Longitudinal Data · 发表于 2023-03-16

Modeling unmeasured baseline information in observational time-to-event data subject to delayed study entry

Regina Stegherr, Jan Beyersmann, Peter Bramlage, Tobias Bluhmki

摘要

Unmeasured baseline information in left-truncated data situations frequently occurs in observational time-to-event analyses. For instance, a typical timescale in trials of antidiabetic treatment is “time since treatment initiation”, but individuals may have initiated treatment before the start of longitudinal data collection. When the focus is on baseline effects, one widespread approach is to fit a Cox proportional hazards model incorporating the measurements at delayed study entry. This has been criticized because of the potential time dependency of covariates. We tackle this problem by using a Bayesian joint model that combines a mixed-effects model for the longitudinal trajectory with a proportional hazards model for the event of interest incorporating the baseline covariate, possibly unmeasured in the presence of left truncation. The novelty is that our procedure is not used to account for non-continuously monitored longitudinal covariates in right-censored time-to-event studies, but to utilize these trajectories to make inferences about missing baseline measurements in left-truncated data. Simulating times-to-event depending on baseline covariates we also compared our proposal to a simpler two-stage approach which performed favorably. Our approach is illustrated by investigating the impact of baseline blood glucose levels on antidiabetic treatment failure using data from a German diabetes register.

SMMR Longitudinal Data Meta-Analysis & Data Integration Causal Inference · 发表于 2023-03-16

Linear mixed models for investigating effect modification in subgroup meta-analysis

Anne Lyngholm Sørensen, Ian C Marschner

摘要

Subgroup meta-analysis can be used for comparing treatment effects between subgroups using information from multiple trials. If the effect of treatment is differential depending on subgroup, the results could enable personalization of the treatment. We propose using linear mixed models for estimating treatment effect modification in aggregate data meta-analysis. The linear mixed models capture existing subgroup meta-analysis methods while allowing for additional features such as flexibility in modeling heterogeneity, handling studies with missing subgroups and more. Reviews and simulation studies of the best suited models for estimating possible differential effect of treatment depending on subgroups have been studied mostly within individual participant data meta-analysis. While individual participant data meta-analysis in general is recommended over aggregate data meta-analysis, conducting an aggregate data subgroup meta-analysis could be valuable for exploring treatment effect modifiers before committing to an individual participant data subgroup meta-analysis. Additionally, using solely individual participant data for subgroup meta-analysis requires collecting sufficient individual participant data which may not always be possible. In this article, we compared existing methods with linear mixed models for aggregate data subgroup meta-analysis under a broad selection of scenarios using simulation and two case studies. Both the case studies and simulation studies presented here demonstrate the advantages of the linear mixed model approach in aggregate data subgroup meta-analysis.

JASA Spatial & Time Series · 发表于 2023-03-16

Intraday Periodic Volatility Curves

Torben G. Andersen, Tao Su, Viktor Todorov, Zhiyuan Zhang

摘要

The volatility of financial asset returns displays pronounced variation over the trading day. Our goal is nonparametric inference for the average intraday volatility pattern, viewed as a function of time-of-day. The functional inference is based on a long span of high-frequency return data. Our setup allows for general forms of volatility dynamics, including time-variation in the intraday pattern. The estimation is based on forming local volatility estimates from the high-frequency returns over overlapping blocks of asymptotically shrinking size, and then averaging these estimates across days in the sample. The block-based estimation of volatility renders the error in the estimation due to the martingale return innovation asymptotically negligible. As a result, the centered and scaled calendar volatility effect estimator converges to a Gaussian process determined by the empirical process error associated with estimating average volatility across the trading day. Feasible inference is obtained by consistently estimating the limiting covariance operator. Simulation results corroborate our theoretical findings. In an application to S&P 500 futures data, we find evidence for a shift in the intraday volatility pattern over time, including a more pronounced role for volatility outside U.S. trading hours in the latter part of the sample. Supplementary materials for this article are available online. *(abstract via OpenAlex)*

SMMR Causal Inference Longitudinal Data · 发表于 2023-03-16

Estimating individualized treatment rules in longitudinal studies with covariate-driven observation times

Janie Coulombe, Erica EM Moodie, Susan M Shortreed, Christel Renoux

摘要

The sequential treatment decisions made by physicians to treat chronic diseases are formalized in the statistical literature as dynamic treatment regimes. To date, methods for dynamic treatment regimes have been developed under the assumption that observation times, that is, treatment and outcome monitoring times, are determined by study investigators. That assumption is often not satisfied in electronic health records data in which the outcome, the observation times, and the treatment mechanism are associated with patients’ characteristics. The treatment and observation processes can lead to spurious associations between the treatment of interest and the outcome to be optimized under the dynamic treatment regime if not adequately considered in the analysis. We address these associations by incorporating two inverse weights that are functions of a patient’s covariates into dynamic weighted ordinary least squares to develop optimal single stage dynamic treatment regimes, known as individualized treatment rules. We show empirically that our methodology yields consistent, multiply robust estimators. In a cohort of new users of antidepressant drugs from the United Kingdom’s Clinical Practice Research Datalink, the proposed method is used to develop an optimal treatment rule that chooses between two antidepressants to optimize a utility function related to the change in body mass index.

JASA Latent Variables & Mixtures · 发表于 2023-03-16

Crowdsourcing Utilizing Subgroup Structure of Latent Factor Modeling

Qi Xu, Yubai Yuan, Junhui Wang, Annie Qu

摘要

Crowdsourcing has emerged as an alternative solution for collecting large scale labels. However, the majority of recruited workers are not domain experts, so their contributed labels could be noisy. In this article, we propose a two-stage model to predict the true labels for multicategory classification tasks in crowdsourcing. In the first stage, we fit the observed labels with a latent factor model and incorporate subgroup structures for both tasks and workers through a multi-centroid grouping penalty. Group-specific rotations are introduced to align workers with different task categories to solve multicategory crowdsourcing tasks. In the second stage, we propose a concordance-based approach to identify high-quality worker subgroups who are relied upon to assign labels to tasks. In theory, we show the estimation consistency of the latent factors and the prediction consistency of the proposed method. The simulation studies show that the proposed method outperforms the existing competitive methods, assuming the subgroup structures within tasks and workers. We also demonstrate the application of the proposed method to real world problems and show its superiority. Supplementary materials for this article are available online. *(abstract via OpenAlex)*

Biometrika Causal Inference · 发表于 2023-03-16

Causal inference with misspecified exposure mappings: separating definitions and assumptions

F Sävje

摘要

Summary: Exposure mappings facilitate investigations of complex causal effects when units interact in experiments. Current methods require experimenters to use the same exposure mappings to define the effect of interest and to impose assumptions on the interference structure. However, the two roles rarely coincide in practice, and experimenters are forced to make the often questionable assumption that their exposures are correctly specified. This paper argues that the two roles exposure mappings currently serve can, and typically should, be separated, so that exposures are used to define effects without necessarily assuming that they are capturing the complete causal structure in the experiment. The paper shows that this approach is practically viable by providing conditions under which exposure effects can be precisely estimated when the exposures are misspecified. Some important questions remain open.

JASA Bayesian Methods · 发表于 2023-03-16

Bayesian Robustness: A Nonasymptotic Viewpoint

Kush Bhatia, Yi-An Ma, Anca D. Dragan, Peter L. Bartlett, Michael I. Jordan

摘要

We study the problem of robustly estimating the posterior distribution for the setting where observed data can be contaminated with potentially adversarial outliers. We propose Rob-ULA, a robust variant of the Unadjusted Langevin Algorithm (ULA), and provide a finite-sample analysis of its sampling distribution. In particular, we show that after T=O˜(d/εacc) iterations, we can sample from pT such that dist(pT,p*)≤εacc+O˜(ϵ), where ϵ is the fraction of corruptions and dist represents the squared 2-Wasserstein distance metric. Our results for the class of posteriors p* which satisfy log-concavity and smoothness assumptions. We corroborate our theoretical analysis with experiments on both synthetic and real-world datasets for mean estimation, regression and binary classification. Supplementary materials for this article are available online. *(abstract via OpenAlex)*

Biometrika Other · 发表于 2023-03-15

√2-estimation for smooth eigenvectors of matrix-valued functions

Giovanni Motta, Wei Biao Wu, Mohsen Pourahmadi

摘要

Summary: Modern statistical methods for multivariate time series rely on the eigendecomposition of matrix-valued functions such as time-varying covariance and spectral density matrices. The curse of indeterminacy or misidentification of smooth eigenvector functions has not received much attention. We resolve this important problem and recover smooth trajectories by examining the distance between the eigenvectors of the same matrix-valued function evaluated at two consecutive points. We change the sign of the next eigenvector if its distance with the current one is larger than the square root of 2. In the case of distinct eigenvalues, this simple method delivers smooth eigenvectors. For coalescing eigenvalues, we match the corresponding eigenvectors and apply an additional signing around the coalescing points. We establish consistency and rates of convergence for the proposed smooth eigenvector estimators. Simulation results and applications to real data confirm that our approach is needed to obtain smooth eigenvectors.

SMMR Bayesian Methods Clinical Trials Survival Analysis · 发表于 2023-03-15

Using dichotomized survival data to construct a prior distribution for a Bayesian seamless Phase II/III clinical trial

Benjamin Duputel, Nigel Stallard, François Montestruc, Sarah Zohar, Moreno Ursino

摘要

Master protocol designs allow for simultaneous comparison of multiple treatments or disease subgroups. Master protocols can also be designed as seamless studies, in which two or more clinical phases are considered within the same trial. They can be divided into two categories: operationally seamless, in which the two phases are separated into two independent studies, and inferentially seamless, in which the interim analysis is considered an adaptation of the study. Bayesian designs are scarcely studied. Our aim is to propose and compare Bayesian operationally seamless Phase II/III designs using a binary endpoint for the first stage and a time-to-event endpoint for the second stage. At the end of Phase II, arm selection is based on posterior (futility) and predictive (selection) probabilities. The results of the first phase are then incorporated into prior distributions of a time-to-event model. Simulation studies showed that Bayesian operationally seamless designs can approach the inferentially seamless counterpart, allowing for an increasing simulated power with respect to the operationally frequentist design.

Biometrika Causal Inference · 发表于 2023-03-15

Targeted optimal treatment regime learning using summary statistics

J Chu, W Lu, S Yang

SMMR Nonparametric & Semiparametric Longitudinal Data · 发表于 2023-03-15

Semiparametric generalized estimating equations for repeated measurements in cross-over designs

Nelson Alirio Cruz Gutierrez, Oscar Orlando Melo, Carlos Alberto Martinez

摘要

A model for cross-over designs with repeated measures within each period was developed. It was obtained using an extension of generalized estimating equations that includes a parametric component to model treatment effects and a non-parametric component to model time and carry-over effects; the estimation approach for the non-parametric component is based on splines. A simulation study was carried out to explore the model properties. Thus, when there is a carry-over effect or a functional temporal effect, the proposed model presents better results than the standard models. Among the theoretical properties, the solution is found to be analogous to weighted least squares. Therefore, model diagnostics can be made by adapting the results from a multiple regression. The proposed methodology was implemented in the data sets of the cross-over experiments that motivated the approach of this work: systolic blood pressure and insulin in rabbits.

JASA Causal Inference Networks & Graphs Machine Learning · 发表于 2023-03-15

Nonlinear Causal Discovery with Confounders

Chunlin Li, Xiaotong Shen, Wei Pan

摘要

This article introduces a causal discovery method to learn nonlinear relationships in a directed acyclic graph with correlated Gaussian errors due to confounding. First, we derive model identifiability under the sublinear growth assumption. Then, we propose a novel method, named the Deconfounded Functional Structure Estimation (DeFuSE), consisting of a deconfounding adjustment to remove the confounding effects and a sequential procedure to estimate the causal order of variables. We implement DeFuSE via feedforward neural networks for scalable computation. Moreover, we establish the consistency of DeFuSE under an assumption called the strong causal minimality. In simulations, DeFuSE compares favorably against state-of-the-art competitors that ignore confounding or nonlinearity. Finally, we demonstrate the utility and effectiveness of the proposed approach with an application to gene regulatory network analysis. The Python implementation is available at https://github.com/chunlinli/defuse. *(abstract via PubMed)*

SMMR Survival Analysis Nonparametric & Semiparametric Networks & Graphs · 发表于 2023-03-15

Copula graphic estimation of the survival function with dependent censoring and its application to analysis of pancreatic cancer clinical trial

Jung Hyun Jo, Zhan Gao, Inkyung Jung, Si Young Song, Geert Ridder, Hyungsik Roger Moon

摘要

In this article, we consider a survival function estimation method that may be suitable for analyses of clinical trials of cancer treatments whose prognosis is known to be poor such as pancreatic cancer treatment. Typically, these kinds of trials are not double-blind, and patients in the control group may drop out in more significant numbers than in the treatment group if their disease progresses (DP). If disease progression is associated with a higher risk of death, then censoring becomes dependent. To estimate the survival function with dependent censoring, we use copula-graphic estimation, where a parametric copula function is used to model the dependence in the joint survival function of the event and censoring time. In this article, we propose a novel method that one can use in choosing the copula parameter. As an application example, we estimate the survival function of the overall survival time of the KG4/2015 study, the phase 3 clinical trial of the efficacy of GV1001 as a treatment for pancreatic cancer. We provide both statistical and clinical pieces of evidence that support the violation of independent censoring. Applying the estimation method with dependent censoring, we obtain that the estimates of the median survival times are 339 days in the treatment group and 225.5 days in the control group. We also find that the estimated difference of the medians is 113.5 days, and the difference is statistically significant at the one-sided level with size 2.5[Formula: see text].

SMMR Testing & Uncertainty Nonparametric & Semiparametric · 发表于 2023-03-15

Confidence intervals for the length of the receiver-operating characteristic curve based on a smooth estimator

Pablo Martínez-Camblor

摘要

A good diagnostic test should show different behavior on both the positive and the negative populations. However, this is not enough for having a good classification system. The binary classification problem is a complex task, which implies to define decision criteria. The knowledge of the level of dissimilarity between the two involved distributions is not enough. We also have to know how to define those decision criteria. The length of the receiver-operating characteristic curve has been proposed as an index of the optimal discriminatory capacity of a biomarker. It is related not with the actual but with the optimal classification capacity of the considered diagnostic test. One particularity of this index is that its estimation should be based on parametric or smoothed models. We explore here the behavior of a kernel density estimator-based approximation for estimating the length of the receiver-operating characteristic curve. We prove the asymptotic distribution of the resulting statistic, propose a parametric bootstrap algorithm for confidence intervals construction, discuss the role that the bandwidth parameter plays in the quality of the provided estimations and, via Monte Carlo simulations, study its finite-sample behavior considering four different criteria for the bandwidth selection. The practical use of the length of the receiver-operating characteristic curve is illustrated through two real-world examples.

SMMR Clinical Trials Causal Inference Bayesian Methods · 发表于 2023-03-15

Bayesian order constrained adaptive design for phase II clinical trials evaluating subgroup-specific treatment effect

Mu Shan, Beibei Guo, Hao Liu, Qian Li, Yong Zang

摘要

The “one-size-fits-all’’ paradigm is inappropriate for phase II clinical trials evaluating biotherapies, which are often expected to have substantial heterogeneous treatment effects among different subgroups defined by biomarker. For these biotherapies, the objective of phase II clinical trials is often to evaluate subgroup-specific treatment effects. In this article, we propose a simple yet efficient Bayesian adaptive phase II biomarker-guided design, referred to as the Bayesian-order constrained adaptive design, to detect the subgroup-specific treatment effects of biotherapies. The Bayesian order constrained adaptive design combines the features of the enrichment design and sequential design. It starts with a “all-comers” stage, and subsequently switches to an enrichment stage for either the marker-positive subgroup or marker-negative subgroup, depending on the interim analysis results. The go/no go enrichment criteria are determined by two posterior probabilities utilizing the inherent ordering constraint between two subgroups. We also extend the Bayesian-order constrained adaptive design to handle the missing biomarker situation. We conducted comprehensive computer simulation studies to investigate the operating characteristics of the Bayesian order constrained adaptive design, and compared it with other existing and conventional designs. The results shown that the Bayesian order constrained adaptive design yielded the best overall performance in detecting the subgroup-specific treatment effects by jointly considering the efficiency and cost-effectiveness of the trials. The software for simulation and trial implementation are available for free download.

SMMR Other · 发表于 2023-03-15

A general averaging method for count data with overdispersion and/or excess zeros in biomedicine

Yin Liu, Jianhong Zhou, Zhanshou Chen, Xinyu Zhang

摘要

With the aim of providing better estimation for count data with overdispersion and/or excess zeros, we develop a novel estimation method— optimal weighting based on cross-validation—for the zero-inflated negative binomial model, where the Poisson, negative binomial, and zero-inflated Poisson models are all included as its special cases. To facilitate the selection of the optimal weight vector, a [Formula: see text]-fold cross-validation technique is adopted. Unlike the jackknife model averaging discussed in Hansen and Racine (2012), the proposed method deletes one group of observations rather than only one observation to enhance the computational efficiency. Furthermore, we also theoretically prove the asymptotic optimality of the newly developed optimal weighting based on cross-validation method. Simulation studies and three empirical applications indicate the superiority of the presented optimal weighting based on cross-validation method when compared with the three commonly used information-based model selection methods and their model averaging counterparts.

SiM Longitudinal Data Diagnostics & Prediction · 发表于 2023-03-14

Structured time‐dependent inverse regression (STIR)

Minsun Song, Efstathia Bura, Roman Parzer, Ruth M. Pfeiffer

摘要

We propose and study structured time‐dependent inverse regression (STIR), a novel sufficient dimension reduction model, to analyze longitudinally measured, correlated biomarkers in relation to an outcome. The time structure is accommodated in an inverse regression model for the markers that can be applied both to equally and unequally spaced time points for each sample. The inverse regression structure also naturally accommodates retrospectively sampled markers, that is, markers measured in case‐control studies. We estimate the corresponding linear combinations of the markers, the reduction, using least squares. We show that under additional distributional assumptions the reduction contains sufficient information about the outcome. In extensive simulations the STIR linear combinations perform well in predictive models based on samples of realistic size. A Wald‐type test for association of a particular marker with outcome at any time point based on the STIR reduction has better power overall than assessing associations based on logistic or linear regression models that include all longitudinally measured markers as independent predictors. As illustrations we estimate the STIR reductions for a cohort study of diabetes and hyperlipidemia and a case‐control study of brain cancer with multiple longitudinally measured biomarkers. We assess the STIR reductions' predictive performance and identify outcome‐associated biomarkers.

JRSSB Clinical Trials · 发表于 2023-03-14

Estimating the efficiency gain of covariate-adjusted analyses in future clinical trials using external data

Xiudi Li, Sijia Li, Alex Luedtke

摘要

We present a framework for using existing external data to identify and estimate the relative efficiency of a covariate-adjusted estimator compared to an unadjusted estimator in a future randomized trial. Under conditions, these relative efficiencies approximate the ratio of sample sizes needed to achieve a desired power. We develop semiparametrically efficient estimators of the relative efficiencies for several treatment effect estimands of interest with either fully or partially observed outcomes, allowing for the application of flexible statistical learning tools to estimate the nuisance functions. We propose an analytic Wald-type confidence interval and a double bootstrap scheme for statistical inference. We demonstrate the performance of the proposed methods through simulation studies and apply these methods to estimate the efficiency gain of covariate adjustment in Covid-19 therapeutic trials.

Biometrika Causal Inference · 发表于 2023-03-14

Ancestor regression in linear structural equation models

C Schultheiss, P Bühlmann

摘要

Summary: We present a new method for causal discovery in linear structural equation models. We propose a simple technique based on statistical testing in linear models that can distinguish between ancestors and non-ancestors of any given variable. Naturally, this approach can then be extended to estimating the causal order among all variables. We provide explicit error control for false causal discovery, at least asymptotically. This holds true even under Gaussianity, where other methods fail due to non-identifiable structures. These Type I error guarantees come at the cost of reduced power. Additionally, we provide an asymptotically valid goodness-of-fit p-value for assessing whether multivariate data stem from a linear structural equation model.

第 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 页