Volume 12, Issue 2 -

Interval estimation of the proportion parameter in the analysis of binary outcome data arising in cluster studies is often an important problem in many biomedical applications. In this paper, we propose two approaches based on the profile likelihood and Wilson score. We compare them with two existing methods recommended for complex survey data and some other methods that are simple extensions of well-known methods such as the likelihood, the generalized estimating equation of Zeger and Liang and the ratio estimator approach of Rao and Scott. An extensive simulation study is conducted for a variety of parameter combinations for the purposes of evaluating and comparing the performance of these methods in terms of coverage and expected lengths. Applications to biomedical data are used to illustrate the proposed methods.

Besides being mainly used for analyzing clustered or longitudinal data, generalized linear mixed models can also be used for smoothing via restricting changes in the fit at the knots in regression splines. The resulting models are usually called semiparametric mixed models (SPMMs). We investigate the effect of smoothing using SPMMs on the correlation and variance parameter estimates for serially correlated longitudinal normal, Poisson and binary data. Through simulations, we compare the performance of SPMMs to other simpler methods for estimating the nonlinear association such as fractional polynomials, and using a parametric nonlinear function. Simulation results suggest that, in general, the SPMMs recover the true curves very well and yield reasonable estimates of the correlation and variance parameters. However, for binary outcomes, SPMMs produce biased estimates of the variance parameters for high serially correlated data. We apply these methods to a dataset investigating the association between CD4 cell count and time since seroconversion for HIV infected men enrolled in the Multicenter AIDS Cohort Study.

In phase II and/or III clinical trial study, there are several competing treatments, the goal is to assess the performances of the treatments at the end of the study, the trial design aims to minimize risks to the patients in the trial, according to some given allocation optimality criterion. Recently, a new type of clinical trial, the staggered-start trial has been proposed in some studies, in which different treatments enter the same trial at different times. Some basic questions for this trial are whether optimality can still be kept? under what conditions? and if so how to allocate the the coming patients to treatments to achieve such optimality? Here we propose and study a class of adaptive designs of staggered-start clinical trials, in which for given optimality criterion object, we show that as long as the initial sizes at the beginning of the successive trials are not too large relative to the total sample size, the proposed design can still achieve optimality criterion asymptotically for the allocation proportions as the ordinary trials; if these initial sample sizes have about the same magnitude as the total sample size, full optimality cannot be achieved. The proposed method is simple to use and is illustrated with several examples and a simulation study.

We present an integer-valued ARCH model which can be used for modeling time series of counts with under-, equi-, or overdispersion. The introduced model has a conditional binomial distribution, and it is shown to be strictly stationary and ergodic. The unknown parameters are estimated by three methods: conditional maximum likelihood, conditional least squares and maximum likelihood type penalty function estimation. The asymptotic distributions of the estimators are derived. A real application of the novel model to epidemic surveillance is briefly discussed. Finally, a generalization of the introduced model is considered by introducing an integer-valued GARCH model.

In randomized clinical trials, we often encounter ordinal categorical responses with repeated measurements. We propose a model-free approach with using the generalized odds ratio (GOR) to measure the relative treatment effect. We develop procedures for testing equality of treatment effects and derive interval estimators for the GOR. We further develop a simple procedure for testing the treatment-by-period interaction. To illustrate the use of test procedures and interval estimators developed here, we consider two real-life data sets, one studying the gender effect on pain scores on an ordinal scale after hip joint resurfacing surgeries, and the other investigating the effect of an active hypnotic drug in insomnia patients on ordinal categories of time to falling asleep.

Mendelian randomization (MR) is a technique that seeks to establish causation between an exposure and an outcome using observational data. It is an instrumental variable analysis in which genetic variants are used as the instruments. Many consortia have meta-analysed genome-wide associations between variants and specific traits and made their results publicly available. Using such data, it is possible to derive genetic risk scores for one trait and to deduce the association of that same risk score with a second trait. The properties of this approach are investigated by simulation and by evaluating the potentially causal effect of birth weight on adult glucose level. In such analyses, it is important to decide whether one is interested in the risk score based on a set of estimated regression coefficients or the score based on the true underlying coefficients. MR is primarily concerned with the latter. Methods designed for the former question will under-estimate the variance if used for MR. This variance can be corrected but it needs to be done with care to avoid introducing bias. MR based on public data sources is useful and easy to perform, but care must be taken to avoid false precision or bias.

In many scenarios, a patient in medical research is treated as a statistical unit. However, in some scenarios, we are interested in treating aggregate data as a statistical unit. In such situations, each set of aggregated data is considered to be a concept in a symbolic representation, and each concept has a hyperrectangle or multiple points in the variable space. To construct a tree-structured model from these aggregate survival data, we propose a new approach, where a datum can be included in several terminal nodes in a tree. By constructing a model under this condition, we expect to obtain a more flexible model while retaining the interpretive ease of a hierarchical structure. In this approach, the survival function of concepts that are partially included in a node is constructed using the Kaplan-Meier method, where the number of events and risks at each time point is replaced by the expectation value of the number of individual descriptions of concepts. We present an application of this proposed model using primary brain tumor patient data. As a result, we obtained a new interpretation of the data in comparison to the classical survival tree modeling methods.

Genome-wide association studies (GWAS) examine a large number of genetic variants, e. g., single nucleotide polymorphisms (SNP), and associate them with a disease of interest. Traditional statistical methods for GWASs can produce spurious associations, due to limited information from individual SNPs and confounding effects. This paper develops two statistical methods to enhance data analysis of GWASs. The first is a multiple-SNP association test, which is a weighted chi-square test derived for big contingency tables. The test assesses combinatorial effects of multiple SNPs and improves conventional methods of single SNP analysis. The second is a method that corrects for confounding effects, which may come from population stratification as well as other ambiguous (unknown) factors. The proposed method identifies a latent confounding factor, using a profile of whole genome SNPs, and eliminates confounding effects through matching or stratified statistical analysis. Simulations and a GWAS of rheumatoid arthritis demonstrate that the proposed methods dramatically remove the number of significant tests, or false positives, and outperforms other available methods.

Recurrent event data arise frequently in many longitudinal follow-up studies. Hence, evaluating covariate effects on the rates of occurrence of such events is commonly of interest. Examples include repeated hospitalizations, recurrent infections of HIV, and tumor recurrences. In this article, we consider semiparametric regression methods for the occurrence rate function of recurrent events when the covariates may be measured with errors. In contrast to the existing works, in our case the conventional assumption of independent censoring is violated since the recurrent event process is interrupted by some correlated events, which is called informative drop-out. Further, some covariates may be measured with errors. To accommodate for both informative censoring and measurement error, the occurrence of recurrent events is modelled through an unspecified frailty distribution and accompanied with a classical measurement error model. We propose two corrected approaches based on different ideas, and we show that they are numerically identical when estimating the regression parameters. The asymptotic properties of the proposed estimators are established, and the finite sample performance is examined via simulations. The proposed methods are applied to the Nutritional Prevention of Cancer trial for assessing the effect of the plasma selenium treatment on the recurrence of squamous cell carcinoma.

Abstract: Modeling hospitalization is complicated because the follow-up time can be censored due to death. In this paper, we propose a shared frailty joint model for survival time and hospitalization. A random effect semi-parametric proportional hazard model is assumed for the survival time and conditional on the follow-up time, hospital admissions or total length of stay is modeled by a generalized linear model with a nonparametric offset function of the follow-up time. We assume that the hospitalization and the survival time are correlated through a latent subject-specific random frailty. The proposed model can be implemented using existing software such as SAS Proc NLMIXED. We demonstrate the feasibility through simulations. We apply our methods to study hospital admissions and total length of stay in a cohort of patients on hemodialysis. We identify age, albumin, neutrophil to lymphocyte ratio (NLR) and vintage as significant risk factors for mortality, and age, gender, race, albumin, NLR, pre-dialysis systolic blood pressure (preSBP), interdialytic weight gain (IDWG) and equilibrated Kt/V (eKt/V) as significant risk factors for both hospital admissions and total length of stay. In addition, hospitalization admissions is positively associated with vintage.

Inverse probability of treatment weighting (IPW) and targeted maximum likelihood estimation (TMLE) are relatively new methods proposed for estimating marginal causal effects. TMLE is doubly robust, yielding consistent estimators even under misspecification of either the treatment or the outcome model. While IPW methods are known to be sensitive to near violations of the practical positivity assumption (e. g., in the case of data sparsity), the consequences of this violation in the TMLE framework for binary outcomes have been less widely investigated. As near practical positivity violations are particularly likely in high-dimensional covariate settings, a better understanding of the performance of TMLE is of particular interest for pharmcoepidemiological studies using large databases. Using plasmode and Monte-Carlo simulation studies, we evaluated the performance of TMLE compared to that of IPW estimators based on a point-exposure cohort study of the marginal causal effect of post-myocardial infarction statin use on the 1-year risk of all-cause mortality from the Clinical Practice Research Datalink. A variety of treatment model specifications were considered, inducing different degrees of near practical non-positivity. Our simulation study showed that the performance of the TMLE and IPW estimators were comparable when the dimension of the fitted treatment model was small to moderate; however, they differed when a large number of covariates was considered. When a rich outcome model was included in the TMLE, estimators were unbiased. In some cases, we found irregular bias and large standard errors with both methods even with a correctly specified high-dimensional treatment model. The IPW estimator showed a slightly better root MSE with high-dimensional treatment model specifications in our simulation setting. In conclusion, for estimation of the marginal expectation of the outcome under a fixed treatment, TMLE and IPW estimators employing the same treatment model specification may perform differently due to differential sensitivity to practical positivity violations; however, TMLE, being doubly robust, shows improved performance with richer specifications of the outcome model. Although TMLE is appealing for its double robustness property, such violations in a high-dimensional covariate setting are problematic for both methods.

Abstract: The Bland–Altman method has been widely used for assessing agreement between two methods of measurement. However, it remains unsolved about sample size estimation. We propose a new method of sample size estimation for Bland–Altman agreement assessment. According to the Bland–Altman method, the conclusion on agreement is made based on the width of the confidence interval for LOAs (limits of agreement) in comparison to predefined clinical agreement limit. Under the theory of statistical inference, the formulae of sample size estimation are derived, which depended on the pre-determined level of α , β, the mean and the standard deviation of differences between two measurements, and the predefined limits. With this new method, the sample sizes are calculated under different parameter settings which occur frequently in method comparison studies, and Monte-Carlo simulation is used to obtain the corresponding powers. The results of Monte-Carlo simulation showed that the achieved powers could coincide with the pre-determined level of powers, thus validating the correctness of the method. The method of sample size estimation can be applied in the Bland–Altman method to assess agreement between two methods of measurement.

Abstract: Schatzkin et al. and other authors demonstrated that the ratios of some conditional statistics such as the true positive fraction are equal to the ratios of unconditional statistics, such as disease detection rates, and therefore we can calculate these ratios between two screening tests on the same population even if negative test patients are not followed with a reference procedure and the true and false negative rates are unknown. We demonstrate that this same property applies to an expected utility metric. We also demonstrate that while simple estimates of relative specificities and relative areas under ROC curves (AUC) do depend on the unknown negative rates, we can write these ratios in terms of disease prevalence, and the dependence of these ratios on a posited prevalence is often weak particularly if that prevalence is small or the performance of the two screening tests is similar. Therefore we can estimate relative specificity or AUC with little loss of accuracy, if we use an approximate value of disease prevalence.

We study the problem of multiple hypothesis testing for correlated clustered data. As the existing multiple comparison procedures based on maximum likelihood estimation could be computationally intensive, we propose to construct multiple comparison procedures based on composite likelihood method. The new test statistics account for the correlation structure within the clusters and are computationally convenient to compute. Simulation studies show that the composite likelihood based procedures maintain good control of the familywise type I error rate in the presence of intra-cluster correlation, whereas ignoring the correlation leads to erratic performance.