A meta-analysis is a useful method for taking the findings of many studies and combining them in the hopes of identifying consistent patterns and sources of disagreement among those findings. While we interpret the average exposure effect, it is necessary to examine the homogeneity of the observed exposure effects across cohort, that is, exposure-by-cohort interaction. If the homogeneity is confirmed, the conclusions concerning exposure effects can be generalized to a broader population. In this paper, a Poisson mixed effects model is used to investigate the cohort effects on the exposure as well as on the baseline risk. The marginal posterior distributions are estimated by a Markov Chain Monte Carlo method, i.e. the Gibbs sampling, to overcome current computational limitations. We illustrate the methods with analyses of data from the Japan Arteriosclerosis Longitudinal Study, in which the effects of smoking on stroke events are examined based on the individual data of 23,860 subjects among 10 cohorts.
Kulldorff (1997) developed a circular spatial scan statistic for identifying the most likely cluster of disease that maximizes the likelihood ratio and his software SaTScan has been widely used for geographical disease cluster detection and disease surveillance. To detect non-circular clusters which cannot be detected by Kulldorff's circular spatial scan statistic, several non-circular spatial scan statistics have been proposed. However, it does not seem to be well recognized that these spatial scan statistics tend to detect the most likely cluster much larger than the true cluster by swallowing neighbouring regions with non-elevated risk. This paper proposes a new spatial scan statistic free from such an undesirable property by modifying the likelihood ratio so that it scans only the regions with elevated risk. Monte Carlo Simulation study shows that the proposed circular spatial scan statistic is shown to have better ability to identify the true cluster compared with Kulldorff's one in all the cluster models considered. The proposed circular spatial scan statisitc is illustrated with mortality data from cerebrovascular disease in Tokyo Metropolitan area, Japan.
In animal breeding and genetics applications, one topic is the evaluation of sexual dimorphism and genetic correlation between sexes. Also, in some cases, variances may be heterogeneous between levels of factors such as breed and herd as well as sex. Researches about these topics need the method for estimating relevant components of (co)variances. The objectives of this study are to derive a computational procedure of the average information algorithm for the restricted maximum likelihood estimation of the relevant (co)variances, and to estimate genetic correlations between sexes on beef carcass traits. In the current computational procedure, a derived expression for the average information matrix is used, whose elements are expressed using the solutions to the mixed model equations in a bivariate mixed linear model with heterogeneous variance and nil covariance of residuals assumed. For the current procedure, replacing the Hessian matrix by the derived average information matrix, a quasi-Newton type procedure is defined for the iterative estimation. Using simulated datasets, computing performance of the current procedure is investigated comparing with the expectation-maximization algorithm, the current procedure is applied to beef carcass traits data to estimate heritability for each sex and genetic correlation between sexes, and then the characteristics of the current procedure is concisely discussed.
The aim of single-arm clinical trials of a new drug is to determine whether it has sufficient promising activity to warrant its further development. For the last several years Bayesian statistical methods have been proposed and used. Bayesian approaches are ideal for earlier phase exploratory trials or proof-of-concept studies as they take into account information that accrues during a trial. Posterior and predictive probabilities are then updated and so become more accurate as the trial progresses. If the relevant external information is available, the decision will be made with a smaller sample size. The goal of this paper is to provide a review for statisticians who use Bayesian methods for the first time or investigators who have some statistical background. In addition, a clinical trial is presented as a real example to illustrate how to conduct a Bayesian approach for single-arm clinical trials with binary endpoints.
In this paper, we introduce the definition of FDR (False Discovery Rate), which gets a lot of attention as a new concept for considering multiplicity effect, and expound properties of it. Furthermore, we enumerate multiple testing procedures that control FDR; Benjamini-Hochberg procedure (linear step-up procedure), Adaptive Benjamini-Hochberg procedure, Benjamini-Yekutieli procedure, Storey's procedure, two-stage linear step-up procedure, and Student-Newman-Kuels procedure. In addition, we review comparisons of them in order to consider which procedure is best to use. As a result, we show the conservativeness of Benjamini-Hochberg procedure and the availability of two-stage linear step-up procedure by Monte-Carlo simulation, in the case of dependent test statistics.
The statistical analysis of survival time or failure time data is an important topic in many areas of research. In ordinary survival data, there is a single, possibly right-censored failure time for each individual. However, in a number of medical applications whereby data is collected an individual may experience one or more events but the first event will preclude the occurrence of another event under investigation. As a result, he/she can experience only one of several types of events. Such data are commonly referred to as competing risks data. Censoring due to such an event is generally not independent of the time to the event of interest. This paper reviews statistical methods for analyzing a competing risks model. Both conceptual considerations and common approaches to one-sample inference; two sample comparison; and covariate effect modeling are discussed. The theory for the analysis of ordinary right-censored survival data can be applied under certain circumstances. Standard statistical software package can perform the necessary analysis, although interpretation of results will vary.