Search

Rathouz and Gao [2] and Luo and Tsai [3] proposed valuable extensions to the generalized linear model for modeling a nonlinear monotonic relationship between the mean response and a set of covariates. In their extensions for discrete data the baseline response distribution is unspecified and is estimated from the data. We propose to extend this model for the analysis of longitudinal data by incorporating random effects into the linear predictor, and using maximum likelihood for estimation and inference. Motivated in particular by longitudinal studies of clinical scale outcomes, we developed an estimation procedure for a finite-support response using a generalized expectation-maximization algorithm where Gauss-Hermite quadrature is employed to approximate the integrals in the E step of the algorithm. Upon convergence, the observed information matrix is estimated through second-order numerical differentiation of the log-likelihood function. Asymptotic properties of the maximum likelihood estimates follow under the usual regularity conditions. Simulation studies are conducted to assess its finite-sample properties and compare the proposed model to the generalized linear mixed model. The proposed method is illustrated in an analysis of data from a longitudinal study of Huntington's disease.

Hidden Markov and semi-Markov models (H(S)MMs) constitute useful tools for modeling observations subject to certain dependency structures. The hidden states render these models very flexible and allow them to capture many different types of latent patterns and dynamics present in the data. This has led to the increased popularity of these models, which have been applied to a variety of problems in various domains and settings, including longitudinal data. In many longitudinal studies, the response variable is categorical or count-type. Generalized linear mixed models (GLMMs) can be used to analyze a wide range of variables, including categorical and count. The present study proposes a model that combines HSMMs with GLMMs, leading to generalized linear mixed hidden semi-Markov models (GLM-HSMMs). These models can account for time-varying unobserved heterogeneity and handle different response types. Parameter estimation is achieved using a Monte Carlo Newton-Raphson (MCNR)-like algorithm. In our proposed model, the distribution of the random effects depends on hidden states. We illustrate the applicability of GLM-HSMMs with an example in the field of occupational health, where the response variable consists of count values. Furthermore, we assess the performance of our MCNR-like algorithm through a simulation study.

We propose a latent linear mixed model to analyze multivariate longitudinal data of multiple ordinal variables, which are manifestations of fewer continuous latent variables. We focus on the latent level where the effects of observed covariates on the latent variables are of interest. We incorporate serial correlation into the variance component rather than assuming independent residuals. We show that misleading inference may be drawn when misspecifying the variance component. Furthermore, we provide a graphical tool depicting latent empirical semi-variograms to detect serial correlation for latent stationary linear mixed models. We apply our proposed model to examine the treatment effect on patients having the amyotrophic lateral sclerosis disease. The result shows that the treatment can slow down progression of latent cervical and lumbar functions.

Sarcopenia is a geriatric syndrome characterized by significant loss of muscle mass. Based on a commonly used definition of the condition that involves three measurements, different subclinical and clinical states of sarcopenia are formed. These states constitute a partially ordered set (poset). This article focuses on the analysis of longitudinal poset in the context of sarcopenia. We propose an extension of the generalized linear mixed model and a recoding scheme for poset analysis such that two submodels-one for ordered categories and one for nominal categories-that include common random effects can be jointly estimated. The new poset model postulates random effects conceptualized as latent variables that represent an underlying construct of interest, that is, susceptibility to sarcopenia over time. We demonstrate how information can be gleaned from nominal sarcopenic states for strengthening statistical inference on a person's susceptibility to sarcopenia.

Stepped wedge cluster trials are an increasingly popular alternative to traditional parallel cluster randomized trials. Such trials often utilize a small number of clusters and numerous time intervals, and these components must be considered when choosing an analysis method. A generalized linear mixed model containing a random intercept and fixed time and intervention covariates is the most common analysis approach. However, the sole use of a random intercept applies a constant intraclass correlation coefficient structure, which is an assumption that is likely to be violated given stepped wedge trials (SWTs) have multiple time intervals. Alternatively, generalized estimating equations (GEE) are robust to the misspecification of the working correlation structure, although it has been shown that small-sample adjustments to standard error estimates and the use of appropriate degrees of freedom are required to maintain the validity of inference when the number of clusters is small. In this article, we show, using an extensive simulation study based on a motivating example and a more general design, the use of GEE can maintain the validity of inference in small-sample SWTs with binary outcomes. Furthermore, we show which combinations of bias corrections to standard error estimates and degrees of freedom work best in terms of attaining nominal type I error rates.

Misspecification of the covariance structure in a linear mixed model (LMM) can lead to biased population parameters' estimates under MAR drop-out. In our motivating example of modeling CD4 cell counts during untreated HIV infection, random intercept and slope LMMs are frequently used. In this article, we evaluate the performance of LMMs with specific covariance structures, in terms of bias in the fixed effects estimates, under specific MAR drop-out mechanisms, and adopt a Bayesian model comparison criterion to discriminate between the examined approaches in real-data applications. We analytically show that using a random intercept and slope structure when the true one is more complex can lead to seriously biased estimates, with the degree of bias depending on the magnitude of the MAR drop-out. Under misspecified covariance structure, we compare in terms of induced bias the approach of adding a fractional Brownian motion (BM) process on top of random intercepts and slopes with the approach of using splines for the random effects. In general, the performance of both approaches was satisfactory, with the BM model leading to smaller bias in most cases. A simulation study is carried out to evaluate the performance of the proposed Bayesian criterion in identifying the model with the correct covariance structure. Overall, the proposed method performs better than the AIC and BIC criteria under our specific simulation setting. The models under consideration are applied to real data from the CASCADE study; the most plausible model is identified by all examined criteria.

Multivariate longitudinal data usually exhibit complex features such as the presence of censored responses due to detection limits of the assay and unavoidable missing values arising when participants make irregular visits that lead to intermittently recorded characteristics. A generalization of the multivariate linear mixed model constructed by taking into account impacts of censored and intermittent missing responses simultaneously, which is named as the MLMM-CM, has been recently proposed for more precisely analyzing such kinds of data. This paper aims at presenting a fully Bayesian sampling-based approach to the MLMM-CM for addressing the uncertainties of censored and missing responses as well as unknown parameters. Two widely accepted Bayesian computational techniques based on the Markov chain Monte Carlo and the inverse Bayes formulas coupled with the Gibbs (IBF-Gibbs) schemes are developed for carrying out posterior inference of the model. The proposed methodology is illustrated through a simulation study and a real-data example from the Adult AIDS Clinical Trials Group 388 study. Numerical results show empirically that the proposed Bayesian methodology performs satisfactorily and offers reliable posterior inference.

Linear mixed models (LMMs) and their extensions have been widely used for high-dimensional genomic data analyses. While LMMs hold great promise for risk prediction research, the high dimensionality of the data and different effect sizes of genomic regions bring great analytical and computational challenges. In this work, we present a multikernel linear mixed model with adaptive lasso (KLMM-AL) to predict phenotypes using high-dimensional genomic data. We develop two algorithms for estimating parameters from our model and also establish the asymptotic properties of LMM with adaptive lasso when only one dependent observation is available. The proposed KLMM-AL can account for heterogeneous effect sizes from different genomic regions, capture both additive and nonadditive genetic effects, and adaptively and efficiently select predictive genomic regions and their corresponding effects. Through simulation studies, we demonstrate that KLMM-AL outperforms most of existing methods. Moreover, KLMM-AL achieves high sensitivity and specificity of selecting predictive genomic regions. KLMM-AL is further illustrated by an application to the sequencing dataset obtained from the Alzheimer's disease neuroimaging initiative.

In the 1990s, China experienced a high degree of antibiotics abuse, which resulted in increased drug resistance. As a result, the World Health Organization introduced a program for children under the age of 5 years who had an acute respiratory tract infection. We analyze the data pertaining to the treatment provided by doctors in several hospitals in China in order to understand the relationships in the data. The data are nested in a three-level hierarchical structure with small cluster sizes ranging from 2 to 10. While large sample theory provides a mechanism to construct confidence intervals and test hypotheses about regression coefficients, the estimation algorithms often fail to converge when they are applied to small cluster sizes. This paper presents a combination of the cluster bootstrap and primary unit splitting methods, called split bootstrap, which is a novel combination that can be used as an alternative when analyzing data pertaining to the abuse of antibiotics in China with small cluster sizes. The split bootstrap method provides accurate estimations with a minimal reduction in precision.

We present a multilevel frailty model for handling serial dependence and simultaneous heterogeneity in survival data with a multilevel structure attributed to clustering of subjects and the presence of multiple failure outcomes. One commonly observes such data, for example, in multi-institutional, randomized placebo-controlled trials in which patients suffer repeated episodes (eg, recurrent migraines) of the disease outcome being measured. The model extends the proportional hazards model by incorporating a random covariate and unobservable random institution effect to respectively account for treatment-by-institution interaction and institutional variation in the baseline risk. Moreover, a random effect term with correlation structure driven by a first-order autoregressive process is attached to the model to facilitate estimation of between patient heterogeneity and serial dependence. By means of the generalized linear mixed model methodology, the random effects distribution is assumed normal and the residual maximum likelihood and the maximum likelihood methods are extended for estimation of model parameters. Simulation studies are carried out to evaluate the performance of the residual maximum likelihood and the maximum likelihood estimators and to assess the impact of misspecifying random effects distribution on the proposed inference. We demonstrate the practical feasibility of the modeling methodology by analyzing real data from a double-blind randomized multi-institutional clinical trial, designed to examine the effect of rhDNase on the occurrence of respiratory exacerbations among patients with cystic fibrosis.

Accelerated failure time (AFT) models allowing for random effects are linear mixed models under the log-transformation of survival time with censoring and describe dependence in correlated survival data. It is well known that the AFT models are useful alternatives to frailty models. To the best of our knowledge, however, there is no literature on variable selection methods for such AFT models. In this paper, we propose a simple but unified variable-selection procedure of fixed effects in the AFT random-effect models using penalized h-likelihood (HL). We consider four penalty functions (ie, least absolute shrinkage and selection operator (LASSO), adaptive LASSO, smoothly clipped absolute deviation (SCAD), and HL). We show that the proposed method can be easily implemented via a slight modification to existing h-likelihood estimation procedures. We thus demonstrate that the proposed method can also be easily extended to AFT models with multilevel (or nested) structures. Simulation studies also show that the procedure using the adaptive LASSO, SCAD, or HL penalty performs well. In particular, we find via the simulation results that the variable selection method with HL penalty provides a higher probability of choosing the true model than other three methods. The usefulness of the new method is illustrated using two actual datasets from multicenter clinical trials.

Correlated binary responses are very common in longitudinal and repeated measures studies. Random effects models are often used to analyze such data and the h-likelihood estimating procedure provides an inferential tool. The objective of this study is to introduce goodness-of-fit score statistics for the fixed part of these models. The proposed statistics are based on processes that partition the observations into mutually exclusive groups. Weighted versions of these statistics are also introduced and are based on the correlation between an appropriately adjusted candidate covariate for entrance into the model and the model residuals. A simulation study indicates that the weighted statistics perform better than their unweighted counterparts, whereas the statistics that are based on the partitioning of the covariate space seem to perform slightly better compared with those based on other grouping procedures. The use of the proposed statistics is illustrated using a real data example.

The timing and frequency of the measurement of longitudinal outcomes in databases may be associated with the value of the outcome. Such visit processes are termed outcome dependent, and previous work showed that conducting standard analyses that ignore outcome-dependent visit times can produce highly biased estimates of the associations of covariates with outcomes. The literature contains several classes of approaches to analyze longitudinal data subject to outcome-dependent visit times, and all of these are based on simplifying assumptions about the visit process. Based on extensive discussions with subject matter investigators, we identified common characteristics of outcome-dependent visit processes that allowed us to evaluate the performance of existing methods in settings with more realistic visit processes than have been previously investigated. This paper uses the analysis of data from a study of kidney function, theory, and simulation studies to examine a range of settings that vary from those where all visits have a low degree of missingness and outcome dependence (which we call "regular" visits) to those where all visits have a high degree of missingness and outcome dependence (which we call "irregular" visits). Our results show that while all the approaches we studied can yield biased estimates of some covariate effects, other covariate effects can be estimated with little bias. In particular, mixed effects models fit by maximum likelihood yielded little bias in estimates of the effects of covariates not associated with the random effects and small bias in estimates of the effects of covariates associated with the random effects. Other approaches produced estimates with greater bias. Our results also show that the presence of some regular visits in the data set protects mixed model analyses from bias but not other methods.

Longitudinal data occur frequently in practice such as medical studies and life sciences. Generalized linear mixed models (GLMMs) are commonly used to analyze such data. It is typically assumed that the random effects covariance matrix is constant among subjects in these models. In many situations, however, the correlation structure may differ among subjects and ignoring this heterogeneity can lead to biases in model parameters estimate. Recently, Lee et al developed a heterogeneous random effects covariance matrix for GLMMs for error-free covariates. Covariates measured with error also happen frequently in the longitudinal data set-up (eg, blood pressure and cholesterol level). Ignoring this issue in the data may produce bias in model parameters estimate and lead to wrong conclusions. In this paper, we propose an approach to properly model the random effects covariance matrix based on covariates in the class of GLMMs, where we also have covariates measured with error. The resulting parameters from the decomposition of random effects covariance matrix have a sensible interpretation and can be easily modeled without the concern of positive definiteness of the resulting estimator. The performance of the proposed approach is evaluated through simulation studies, which show that the proposed method performs very well in terms of bias, mean squared error, and coverage rate. An application of the proposed method is also provided using a longitudinal data from Manitoba follow-up study.

Generalized linear mixed models have played an important role in the analysis of longitudinal data; however, traditional approaches have limited flexibility in accommodating skewness and complex correlation structures. In addition, the existing estimation approaches generally rely heavily on the specifications of random effects distributions; therefore, the corresponding inferences are sometimes sensitive to the choice of random effect distributions under certain circumstance. In this paper, we incorporate serially dependent distribution-free random effects into Tweedie generalized linear models to accommodate a wide range of skewness and covariance structures for discrete and continuous longitudinal data. An optimal estimation of our model has been developed using the orthodox best linear unbiased predictors of random effects. Our approach unifies population-averaged and subject-specific inferences. Our method is illustrated through the analyses of patient-controlled analgesia data and Framingham cholesterol data.

Longitudinal studies with multiple outcomes often pose challenges for the statistical analysis. A joint model including all outcomes has the advantage of incorporating the simultaneous behavior but is often difficult to fit due to computational challenges. We consider 2 alternative approaches to quantify and assess the loss in efficiency as compared with joint modelling when evaluating fixed effects. The first approach is pairwise fitting of pseudolikelihood functions for pairs of outcomes. The second approach recovers correlations between parameter estimates across multiple marginal linear mixed models. The methods are evaluated in terms of a data example both from a study on the effects of milk protein on health in young adolescents and in an extensive simulation study. We find that the 2 alternatives give similar results in settings where an exchangeability condition is met, but otherwise, pairwise fitting shows a larger loss in efficiency than the marginal models approach. Using an alternative to the joint modelling strategy will lead to some but not necessarily a large loss of efficiency for small sample sizes.

Meta-analysis of diagnostic studies is still field of ongoing biometrical research. Especially, clinical researchers call for methods that allow for a comparison of different diagnostic tests to a common gold standard. Focussing on two diagnostic tests, the main parameters of interest are differences of sensitivities and specificities (with their corresponding confidence intervals) between the two diagnostic tests while accounting for the various associations across the two tests and the single studies. Similar to our previous work using generalized linear mixed models to this task, we propose a model with a quadrivariate response consisting of the two sensitivities and the two specificities of both tests. This new approach uses the ideas of copula modelling, and especially a quadrivariate Gaussian copula and a quadrivariate vine copula, which is built from bivariate Plackett copulas. The different copulas are compared in a simulation study and illustrated by the application of population-based screening for type 2 diabetes.

Longitudinal measurement of biomarkers is important in determining risk factors for binary endpoints such as infection or disease. However, biomarkers are subject to measurement error, and some are also subject to left-censoring due to a lower limit of detection. Statistical methods to address these issues are few. We herein propose a generalized linear mixed model and estimate the model parameters using the Monte Carlo Newton-Raphson (MCNR) method. Inferences regarding the parameters are made by applying Louis's method and the delta method. Simulation studies were conducted to compare the proposed MCNR method with existing methods including the maximum likelihood (ML) method and the ad hoc approach of replacing the left-censored values with half of the detection limit (HDL). The results showed that the performance of the MCNR method is superior to ML and HDL with respect to the empirical standard error, as well as the coverage probability for the 95% confidence interval. The HDL method uses an incorrect imputation method, and the computation is constrained by the number of quadrature points; while the ML method also suffers from the constrain for the number of quadrature points, the MCNR method does not have this limitation and approximates the likelihood function better than the other methods. The improvement of the MCNR method is further illustrated with real-world data from a longitudinal study of local cervicovaginal HIV viral load and its effects on oncogenic HPV detection in HIV-positive women.

Widespread inconsistencies are commonly observed between physicians' ordinal classifications in screening tests results such as mammography. These discrepancies have motivated large-scale agreement studies where many raters contribute ratings. The primary goal of these studies is to identify factors related to physicians and patients' test results, which may lead to stronger consistency between raters' classifications. While ordered categorical scales are frequently used to classify screening test results, very few statistical approaches exist to model agreement between multiple raters. Here we develop a flexible and comprehensive approach to assess the influence of rater and subject characteristics on agreement between multiple raters' ordinal classifications in large-scale agreement studies. Our approach is based upon the class of generalized linear mixed models. Novel summary model-based measures are proposed to assess agreement between all, or a subgroup of raters, such as experienced physicians. Hypothesis tests are described to formally identify factors such as physicians' level of experience that play an important role in improving consistency of ratings between raters. We demonstrate how unique characteristics of individual raters can be assessed via conditional modes generated during the modeling process. Simulation studies are presented to demonstrate the performance of the proposed methods and summary measure of agreement. The methods are applied to a large-scale mammography agreement study to investigate the effects of rater and patient characteristics on the strength of agreement between radiologists. Copyright © 2017 John Wiley & Sons, Ltd.

Model-based standardization uses a statistical outcome model or exposure model to estimate a population-average association that is unconfounded by selected covariates. With it, one can compare groups using a distribution of confounders identical in each group to that of a standard population. We develop an approach based on an outcome model, in which the mean of the outcome is modeled conditional on the exposure and the confounders. In our approach, there is a confounder that clusters the observations into a very large number of categories. We treat the parameters for the clusters as random effects. We use a between-within model to account for the association of the random effects not only with the exposure but also with the cluster population sizes. We review alternative approaches presented in the literature, and we compare the outcome-modeling approach to recently proposed exposure-modeling approaches incorporating random effects. To illustrate, we use 2014 to compare proportions of acute respiratory tract infection diagnoses with an antibiotic prescription for emergency department versus outpatient visits, adjusting for confounding by unmeasured patient level variables and measured diagnosis-level variables. We also present results of a simulation study.

The literature about Prediction Interval (PI) and Tolerance Interval (TI) in linear mixed models is usually developed for specific designs, which is a main limitation to their use. This paper proposes to reformulate the two‐sided PI to be generalizable under a wide variety of designs (one random factor, nested and crossed designs for multiple random factors, and balanced or unbalanced designs). This new methodology is based on the Hessian matrix, namely, the inverse of (observed) Fisher Information matrix, and is built with a cell mean model. The degrees of freedom for the total variance are calculated with the generalized Satterthwaite method and compared to the Kenward‐Roger's degrees of freedom for fixed effects. Construction of two‐sided TIs are also detailed with one random factor, and two nested and two crossed random variables. An extensive simulation study is carried out to compare the widths and coverage probabilities of Confidence Intervals (CI), PIs, and TIs to their nominal levels. It shows excellent coverage whatever the design and the sample size are. Finally, these CIs, PIs, and TIs are applied to two real data sets: one from orthopedic surgery study (intralesional resection risk) and the other from assay validation study during vaccine development.

Longitudinal binomial data are frequently generated from multiple questionnaires and assessments in various scientific settings for which the binomial data are often overdispersed. The standard generalized linear mixed effects model may result in severe underestimation of standard errors of estimated regression parameters in such cases and hence potentially bias the statistical inference. In this paper, we propose a longitudinal beta-binomial model for overdispersed binomial data and estimate the regression parameters under a probit model using the generalized estimating equation method. A hybrid algorithm of the Fisher scoring and the method of moments is implemented for computing the method. Extensive simulation studies are conducted to justify the validity of the proposed method. Finally, the proposed method is applied to analyze functional impairment in subjects who are at risk of Huntington disease from a multisite observational study of prodromal Huntington disease. Copyright © 2016 John Wiley & Sons, Ltd.

Resecting bone tumors requires good cutting accuracy to reduce the occurrence of local recurrence. This issue is considerably reduced with a navigated technology. The estimation of extreme proportions is challenging especially with small or moderate sample sizes. When no success is observed, the commonly used binomial proportion confidence interval is not suitable while the rule of three provides a simple solution. Unfortunately, these approaches are unable to differentiate between different unobserved events. Different delta methods and bootstrap procedures are compared in univariate and linear mixed models with simulations and real data by assuming the normality. The delta method on the z-score and parametric bootstrap provide similar results but the delta method requires the estimation of the covariance matrix of the estimates. In mixed models, the observed Fisher information matrix with unbounded variance components should be preferred. The parametric bootstrap, easier to apply, outperforms the delta method for larger sample sizes but it may be time costly. Copyright © 2016 John Wiley & Sons, Ltd.

Agreement between experts’ ratings is an important prerequisite for an effective screening procedure. In clinical settings large-scale studies are often conducted to compare the agreement of experts’ ratings between new and existing medical tests, for example, digital versus film mammography. Challenges arise in these studies where many experts rate the same sample of patients undergoing two medical tests, leading to a complex correlation structure between experts’ ratings. Here we propose a novel paired kappa measure to compare the agreement between the ordinal binary ratings of many experts across two medical tests. Existing approaches can accommodate only a small number of experts, and rely heavily on Cohen’s kappa and Scott’s pi measures of agreement, thus are prone to their drawbacks. The proposed kappa appropriately accounts for correlations between ratings due to patient characteristics, corrects for agreement due to chance and is robust to disease prevalence and other flaws inherent in the use of Cohen’s kappa. It can be easily calculated in the software package R. In contrast to existing approaches, the proposed measure can flexibly incorporate large numbers of experts and patients by utilizing the generalized linear mixed models framework. It is intended to be used in population-based studies, increasing efficiency without increasing modeling complexity. Extensive simulation studies demonstrate low bias and excellent coverage probability of the proposed kappa under a broad range of conditions. Methods are applied to a recent nationwide breast cancer screening study comparing film mammography to digital mammography.

In biomedical studies and clinical trials, repeated measures are often subject to some upper and/or lower limits of detection. Hence, the responses are either left or right censored. A complication arises when more than one series of responses is repeatedly collected on each subject at irregular intervals over a period of time and the data exhibit tails heavier than the normal distribution. The multivariate censored linear mixed effect (MLMEC) model is a frequently used tool for a joint analysis of more than one series of longitudinal data. In this context, we develop a robust generalization of the MLMEC based on the scale mixtures of normal distributions. To take into account the autocorrelation existing among irregularly observed measures, a damped exponential correlation structure is considered. For this complex longitudinal structure, we propose an exact estimation procedure to obtain the maximum-likelihood estimates of the fixed effects and variance components using a stochastic approximation of the EM algorithm. This approach allows us to estimate the parameters of interest easily and quickly as well as to obtain the standard errors of the fixed effects, the predictions of unobservable values of the responses, and the log-likelihood function as a byproduct. The proposed method is applied to analyze a set of AIDS data and is examined via a simulation study.

Longitudinal data are widely analysed using linear mixed models, with ‘random slopes’ models particularly common. However, when modelling, for example, longitudinal pre‐treatment CD4 cell counts in HIV‐positive patients, the incorporation of non‐stationary stochastic processes such as Brownian motion has been shown to lead to a more biologically plausible model and a substantial improvement in model fit. In this article, we propose two further extensions. Firstly, we propose the addition of a fractional Brownian motion component, and secondly, we generalise the model to follow a multivariate‐t distribution. These extensions are biologically plausible, and each demonstrated substantially improved fit on application to example data from the Concerted Action on SeroConversion to AIDS and Death in Europe study. We also propose novel procedures for residual diagnostic plots that allow such models to be assessed. Cohorts of patients were simulated from the previously reported and newly developed models in order to evaluate differences in predictions made for the timing of treatment initiation under different clinical management strategies. A further simulation study was performed to demonstrate the substantial biases in parameter estimates of the mean slope of CD4 decline with time that can occur when random slopes models are applied in the presence of censoring because of treatment initiation, with the degree of bias found to depend strongly on the treatment initiation rule applied. Our findings indicate that researchers should consider more complex and flexible models for the analysis of longitudinal biomarker data, particularly when there are substantial missing data, and that the parameter estimates from random slopes models must be interpreted with caution. © 2015 The Authors. Statistics in Medicine Published by John Wiley & Sons Ltd.

When conducting a meta-analysis of studies with bivariate binary outcomes, challenges arise when the within-study correlation and between-study heterogeneity should be taken into account. In this paper, we propose a marginal beta-binomial model for the meta-analysis of studies with binary outcomes. This model is based on the composite likelihood approach and has several attractive features compared with the existing models such as bivariate generalized linear mixed model (Chu and Cole, 2006) and Sarmanov beta-binomial model (Chen et al., 2012). The advantages of the proposed marginal model include modeling the probabilities in the original scale, not requiring any transformation of probabilities or any link function, having closed-form expression of likelihood function, and no constraints on the correlation parameter. More importantly, because the marginal beta-binomial model is only based on the marginal distributions, it does not suffer from potential misspecification of the joint distribution of bivariate study-specific probabilities. Such misspecification is difficult to detect and can lead to biased inference using currents methods. We compare the performance of the marginal beta-binomial model with the bivariate generalized linear mixed model and the Sarmanov beta-binomial model by simulation studies. Interestingly, the results show that the marginal beta-binomial model performs better than the Sarmanov beta-binomial model, whether or not the true model is Sarmanov beta-binomial, and the marginal beta-binomial model is more robust than the bivariate generalized linear mixed model under model misspecifications. Two meta-analyses of diagnostic accuracy studies and a meta-analysis of case-control studies are conducted for illustration.

We used theoretical and simulation-based approaches to study Type I error rates for one-stage and two-stage analytic methods for cluster-randomized designs. The one-stage approach uses the observed data as outcomes and accounts for within-cluster correlation using a general linear mixed model. The two-stage model uses the cluster specific means as the outcomes in a general linear univariate model. We demonstrate analytically that both one-stage and two-stage models achieve exact Type I error rates when cluster sizes are equal. With unbalanced data, an exact size α test does not exist, and Type I error inflation may occur. Via simulation, we compare the Type I error rates for four one-stage and six two-stage hypothesis testing approaches for unbalanced data. With unbalanced data, the two-stage model, weighted by the inverse of the estimated theoretical variance of the cluster means, and with variance constrained to be positive, provided the best Type I error control for studies having at least six clusters per arm. The one-stage model with Kenward-Roger degrees of freedom and unconstrained variance performed well for studies having at least 14 clusters per arm. The popular analytic method of using a one-stage model with denominator degrees of freedom appropriate for balanced data performed poorly for small sample sizes and low intracluster correlation. Because small sample sizes and low intracluster correlation are common features of cluster-randomized trials, the Kenward-Roger method is the preferred one-stage approach.

Count data often arise in biomedical studies, while there could be a special feature with excessive zeros in the observed counts. The zero-inflated Poisson model provides a natural approach to accounting for the excessive zero counts. In the semiparametric framework, we propose a generalized partially linear single-index model for the mean of the Poisson component, the probability of zero, or both. We develop the estimation and inference procedure via a profile maximum likelihood method. Under some mild conditions, we establish the asymptotic properties of the profile likelihood estimators. The finite sample performance of the proposed method is demonstrated by simulation studies, and the new model is illustrated with a medical care dataset.

Individual participant data meta-analyses (IPD-MA) are increasingly used for developing and validating multivariable (diagnostic or prognostic) risk prediction models. Unfortunately, some predictors or even outcomes may not have been measured in each study and are thus systematically missing in some individual studies of the IPD-MA. As a consequence, it is no longer possible to evaluate between-study heterogeneity and to estimate study-specific predictor effects, or to include all individual studies, which severely hampers the development and validation of prediction models.Here, we describe a novel approach for imputing systematically missing data and adopt a generalized linear mixed model to allow for between-study heterogeneity. This approach can be viewed as an extension of Resche-Rigon's method (Stat Med 2013), relaxing their assumptions regarding variance components and allowing imputation of linear and nonlinear predictors.We illustrate our approach using a case study with IPD-MA of 13 studies to develop and validate a diagnostic prediction model for the presence of deep venous thrombosis. We compare the results after applying four methods for dealing with systematically missing predictors in one or more individual studies: complete case analysis where studies with systematically missing predictors are removed, traditional multiple imputation ignoring heterogeneity across studies, stratified multiple imputation accounting for heterogeneity in predictor prevalence, and multilevel multiple imputation (MLMI) fully accounting for between-study heterogeneity.We conclude that MLMI may substantially improve the estimation of between-study heterogeneity parameters and allow for imputation of systematically missing predictors in IPD-MA aimed at the development and validation of prediction models. Copyright (c) 2015John Wiley & Sons, Ltd.

There have been numerous attempts to extend the Wilcoxon rank-sum test to clustered data. Recently, one such rank-sum test (Dutta & Datta, 2016, Biometrics 72, 432-440) was developed to compare the group-specific marginal distributions of outcomes in clustered data where the conditional distributions of outcomes depend on the number of observations from that group in a given cluster, a phenomenon referred to as informative intra-cluster group (ICG) size. However, comparison of group-specific marginal distributions may not be sufficient in presence of some potentially useful covariables that are observed in the study. In addition, not accounting for the effect of these covariates can lead to biased and misleading inference for the group comparisons. Thus, the purpose of this article is twofold. First, we develop a method to estimate the covariate effects using rank-based weighted estimating equations that are appropriate when the ICG size is informative. Second, we construct an aligned rank-sum test based on the covariate adjusted outcomes. Asymptotic distributions of the R-estimators and the test statistic are provided. Through simulation studies, we show the importance of selecting proper weights in constructing the estimating equations when informativeness is present through the cluster or ICG sizes. We also demonstrate the superiority and the robustness of our method in comparison to regular parametric linear mixed models in clustered data. We apply our method to analyze different real-life data sets including a data on birthweights of rat pups in different litters and a dental data on tooth attachment loss.

Naturalistic driving studies provide opportunities for investigating the effects of key driving exposures on risky driving performance and accidents. New technology provides a realistic assessment of risky driving through the intensive monitoring of kinematic behavior while driving. These studies with their complex data structures provide opportunities for statisticians to develop needed modeling techniques for statistical inference. This article discusses new statistical modeling procedures that were developed to specifically answer important analytical questions for naturalistic driving studies. However, these methodologies also have important applications for the analysis of intensively collected longitudinal data, an increasingly common data structure with the advent of wearable devises. In order to examine the sources of variation between- and within-subjects in risky driving behavior, we explore the use of generalized linear mixed models with autoregressive random processes to analyzing long sequences of kinematic count data from a group of teenagers that have measurements at each trip over a 1.5 year observation period starting after receiving their license. These models provide a regression framework for examining the effects of driving conditions and exposures on risky driving behavior. Alternatively, Generalized estimating equations approaches are explored for the situation where we have intensively collected count measurements on a moderate number of participants. In addition to proposing statistical modeling for kinematic events, we explore models for relating kinematic events with crash risk. Specifically, we propose both latent variable and hidden Markov model in models for relating these two processes and for developing dynamic predictors of crash risk from longitudinal kinematic event data. These different statistical modeling techniques are all used to analyze data from the Naturalistic Teenage Driving Study (NTDS), a unique investigation into how teenagers drive after licensure.

The classic concordance correlation coefficient measures the agreement between two variables. In recent studies, concordance correlation coefficients have been generalized to deal with responses from a distribution from the exponential family using the univariate generalized linear mixed model. Multivariate data arise when responses on the same unit are measured repeatedly by several methods. The relationship among these responses is often of interest. In clustered mixed data, the correlation could be present between repeated measurements either within the same observer or between different methods on the same subjects. Indices for measuring such association are needed. This study proposes a series of indices, namely, intra-correlation, inter-correlation, and total correlation coefficients to measure the correlation under various circumstances in a multivariate generalized linear model, especially for joint modeling of clustered count and continuous outcomes. The proposed indices are natural extensions of the concordance correlation coefficient. We demonstrate the methodology with simulation studies. A case example of osteoarthritis study is provided to illustrate the use of these proposed indices. Copyright © 2014 John Wiley & Sons, Ltd.

Clustered data are common in medical research. Typically, one is interested in a regression model for the association between an outcome and covariates. Two complications that can arise when analysing clustered data are informative cluster size (ICS) and confounding by cluster (CBC). ICS and CBC mean that the outcome of a member given its covariates is associated with, respectively, the number of members in the cluster and the covariate values of other members in the cluster. Standard generalised linear mixed models for cluster-specific inference and standard generalised estimating equations for population-average inference assume, in general, the absence of ICS and CBC. Modifications of these approaches have been proposed to account for CBC or ICS. This article is a review of these methods. We express their assumptions in a common format, thus providing greater clarity about the assumptions that methods proposed for handling CBC make about ICS and vice versa, and about when different methods can be used in practice. We report relative efficiencies of methods where available, describe how methods are related, identify a previously unreported equivalence between two key methods, and propose some simple additional methods. Unnecessarily using a method that allows for ICS/CBC has an efficiency cost when ICS and CBC are absent. We review tools for identifying ICS/CBC. A strategy for analysis when CBC and ICS are suspected is demonstrated by examining the association between socio-economic deprivation and preterm neonatal death in Scotland.

In this paper, we investigate the impact of time-invariant covariates when fitting transition mixed models. This is carried out by emphasizing on the role of baseline responses on the estimation process. Transition models are allowed for two cases of exogenous and endogenous baseline responses. We illustrate these concepts in the special case of transition linear mixed models with centered time-varying covariates. Results of our simulation studies show that the omission, or the inclusion, of time-invariant covariates is not important in models with exogenous baseline responses, while it has an essential effect on fitting models with the endogenous baseline responses. It is also emphasized that the effect becomes minor when the endogeneity issue is handled. The practical consequences are illustrated in the analysis of a real data set taken from medical sciences.

Longitudinal data are often segmented by unobserved time-varying factors, which introduce latent heterogeneity at the observation level, in addition to heterogeneity across subjects. We account for this latent structure by a linear mixed hidden Markov model. It integrates subject-specific random effects and Markovian sequences of time-varying effects in the linear predictor. We propose an expectation—maximization algorithm for maximum likelihood estimation, based on data augmentation. It reduces to the iterative maximization of the expected value of a complete likelihood function, derived from an augmented dataset with case weights, alternated with weights updating. In a case study of the Survey on Stress Aging and Health in Russia, the model is exploited to estimate the influence of the observed covariates under unobserved time-varying factors, which affect the cardiovascular activity of each subject during the observation period.

We discuss the computation of randomization tests for clinical trials of two treatments when the primary outcome is based on a regression model. We begin by revisiting the seminal paper of Gail, Tan, and Piantadosi (1988), and then describe a method based on Monte Carlo generation of randomization sequences. The tests based on this Monte Carlo procedure are design-based, in that they incorporate the particular randomization procedure used. We discuss permuted block designs, complete randomization, and biased coin designs. We also use a new technique by Plamadeala and Rosenberger (2012) for simple computation of conditional randomization tests. Like Gail, Tan, and Piantadosi, we focus on residuals from generalized linear models and martingale residuals from survival models. Such techniques do not apply to longitudinal data analysis, and we introduce a method for computation of randomization tests based on the predicted rate of change from a generalized linear mixed model when outcomes are longitudinal. We show, by simulation, that these randomization tests preserve the size and power well under model misspecification.

Two main classes of methodology have been developed for addressing the analytical intractability of generalized linear mixed models: likelihood-based methods and Bayesian methods. Likelihood-based methods such as the penalized quasi-likelihood approach have been shown to produce biased estimates especially for binary clustered data with small clusters sizes. More recent methods using adaptive Gaussian quadrature perform well but can be overwhelmed by problems with large numbers of random effects, and efficient algorithms to better handle these situations have not yet been integrated in standard statistical packages. Bayesian methods, although they have good frequentist properties when the model is correct, are known to be computationally intensive and also require specialized code, limiting their use in practice. In this article, we introduce a modification of the hybrid approach of Capanu and Begg, 2011, Biometrics 67, 371–380, as a bridge between the likelihood-based and Bayesian approaches by employing Bayesian estimation for the variance components followed by Laplacian estimation for the regression coefficients. We investigate its performance as well as that of several likelihood-based methods in the setting of generalized linear mixed models with binary outcomes. We apply the methods to three datasets and conduct simulations to illustrate their properties. Simulation results indicate that for moderate to large numbers of observations per random effect, adaptive Gaussian quadrature and the Laplacian approximation are very accurate, with adaptive Gaussian quadrature preferable as the number of observations per random effect increases. The hybrid approach is overall similar to the Laplace method, and it can be superior for data with very sparse random effects. Copyright © 2013 John Wiley & Sons, Ltd.