Bayesian Transformation Models for Multivariate Survival Data

In this paper, we propose a general class of Gamma frailty transformation models for multivariate survival data. The transformation class includes the commonly used proportional hazards and proportional odds models. The proposed class also includes a family of cure rate models. Under an improper prior for the parameters, we establish propriety of the posterior distribution. A novel Gibbs sampling algorithm is developed for sampling from the observed data posterior distribution. A simulation study is conducted to examine the properties of the proposed methodology. An application to a data set from a cord blood transplantation study is also reported.     

A Weibull Regression Model with Gamma Frailties for Multivariate Survival Data

Frequently in the analysis of survival data, survival times within the same group are correlated due to unobserved co-variates. One way these co-variates can be included in the model is as frailties. These frailty random block effects generate dependency between the survival times of the individuals which are conditionally independent given the frailty. Using a conditional proportional hazards model, in conjunction with the frailty, a whole new family of models is introduced. By considering a gamma frailty model, often the issue is to find an appropriate model for the baseline hazard function. In this paper a flexible baseline hazard model based on a correlated prior process is proposed and is compared with a standard Weibull model. Several model diagnostics methods are developed and model comparison is made using recently developed Bayesian model selection criteria. The above methodologies are applied to the McGilchrist and Aisbett (1991) kidney infection data and the analysis is performed using Markov Chain Monte Carlo methods. This revised version was published online in July 2006 with corrections to the Cover Date.     

Survival analysis based on the proportional hazards model and survey data

The authors propose methods based on the stratified Cox proportional hazards model that account for the fact that the data have been collected according to a complex survey design. The methods they propose are based on the theory of estimating equations in conjunction with empirical process theory. The authors also discuss issues concerning ignorable sampling design, and the use of weighted and unweighted procedures. They illustrate their methodology by an analysis of jobless spells in Statistics Canada's Survey of Labour and Income Dynamics. They discuss briefly problems concerning weighting, model checking, and missing or mismeasured data. They also identify areas for further research.     

Parameter Orthogonality in Mixed Regression Models for Survival Data

The implications of parameter orthogonality for the robustness of survival regression models are considered. The question of which of the proportional hazards or the accelerated life families of models would be more appropriate for analysis is usually ignored, and the proportional hazards family is applied, particularly in medicine, for convenience. Accelerated life models have conventionally been used in reliability applications. We propose a one-parameter family mixture survival model which includes both the accelerated life and the proportional hazards models. By orthogonalizing relative to the mixture parameter, we can show that, for small effects of the covariates, the regression parameters under the alternative families agree to within a constant. This recovers a known misspecification result. We use notions of parameter orthogonality to explore robustness to other types of misspecification including misspecified base-line hazards. The results hold in the presence of censoring. We also study the important question of when proportionality matters.     

Multivariate Parametric Spatiotemporal Models for County Level Breast Cancer Survival Data

In clustered survival settings where the clusters correspond to geographic regions, biostatisticians are increasingly turning to models with spatially distributed random effects. These models begin with spatially oriented frailty terms, but may also include further region-level terms in the parametrization of the baseline hazards or various covariate effects (as in a spatially-varying coefficients model). In this paper, we propose a multivariate conditionally autoregressive (MCAR) model as a mixing distribution for these random effects, as a way of capturing correlation across both the regions and the elements of the random effect vector for any particular region. We then extend this model to permit analysis of temporal cohort effects, where we use the term temporal cohort to mean a group of subjects all of whom were diagnosed with the disease of interest (and thus, entered the study) during the same time period (say, calendar year). We show how our spatiotemporal model may be efficiently fit in a hierarchical Bayesian framework implemented using Markov chain Monte Carlo (MCMC) computational techniques. We illustrate our approach in the context of county-level breast cancer data from 22 annual cohorts of women living in the state of Iowa, as recorded by the Surveillance, Epidemiology, and End Results (SEER) database. Hierarchical model comparison using the Deviance Information Criterion (DIC), as well as maps of the fitted county-level effects, reveal the benefit of our approach.     

A note on variance estimation in the Cox proportional hazards model

The Cox proportional hazards model is widely used in clinical trials with time-to-event outcomes to compare an experimental treatment with the standard of care. At the design stage of a trial the number of events required to achieve a desired power needs to be determined, which is frequently based on estimating the variance of the maximum partial likelihood estimate of the regression parameter with a function of the number of events. Underestimating the variance at the design stage will lead to insufficiently powered studies, and overestimating the variance will lead to unnecessarily large trials. A simple approach to estimating the variance is introduced, which is compared with two widely adopted approaches in practice. Simulation results show that the proposed approach outperforms the standard ones and gives nearly unbiased estimates of the variance.     

Data-Driven Model Choice in Multivariate Nonparametric Regression

The problem of choosing between different regression models, is investigated. A completely automatic model selection procedure is defined and asymptotic optimality results are shown. These results are stated in a general framework including as well nested as unnested situations, including as well i.i.d. as dependent samples, and without specifying any class of non-parametric smoothers. Because of the well-known curse of dimensionality, model selection is of particular interest in multivariate regression settings, and it will be discussed how the previous general model choice approach can be used in several different highdimensional situations.     

Small Sample Bias in the Gamma Frailty Model for Univariate Survival

The gamma frailty model is a natural extension of the Cox proportional hazards model in survival analysis. Because the frailties are unobserved, an E-M approach is often used for estimation. Such an approach is shown to lead to finite sample underestimation of the frailty variance, with the corresponding regression parameters also being underestimated as a result. For the univariate case, we investigate the source of the bias with simulation studies and a complete enumeration. The rank-based E-M approach, we note, only identifies frailty through the order in which failures occur; additional frailty which is evident in the survival times is ignored, and as a result the frailty variance is underestimated. An adaption of the standard E-M approach is suggested, whereby the non-parametric Breslow estimate is replaced by a local likelihood formulation for the baseline hazard which allows the survival times themselves to enter the model. Simulations demonstrate that this approach substantially reduces the bias, even at small sample sizes. The method developed is applied to survival data from the North West Regional Leukaemia Register.     

A New Estimator for Cox Proportional Hazard Regression Model in Presence of Collinearity

We propose a new approach to estimate the parameters of the Cox proportional hazards model in the presence of collinearity. Generally, a maximum partial likelihood estimator is used to estimate parameters for the Cox proportional hazards model. However, the maximum partial likelihood estimators can be seriously affected by the presence of collinearity since the parameter estimates result in large variances.

In this study, we develop a Liu-type estimator for Cox proportional hazards model parameters and compare it with a ridge regression estimator based on the scalar mean squared error (MSE). Finally, we evaluate its performance through a simulation study.     

Semiparametric Efficient Estimation in the Generalized Odds-Rate Class of Regression Models for Right-Censored Time-to-Event Data

The generalized odds-rate class of regression models for time to event data is indexed by a non-negative constant and assumes thatg(S(t|Z)) = (t) + Zwhere g(s) = log(^-1(s-) for > 0, g₀(s) = log(- log s), S(t|Z) is the survival function of the time to event for an individual with qx1 covariate vector Z, is a qx1 vector of unknown regression parameters, and (t) is some arbitrary increasing function of t. When =0, this model is equivalent to the proportional hazards model and when =1, this model reduces to the proportional odds model. In the presence of right censoring, we construct estimators for and exp((t)) and show that they are consistent and asymptotically normal. In addition, we show that the estimator for is semiparametric efficient in the sense that it attains the semiparametric variance bound.     

Extended Hazard Regression Model for Reliability and Survival Analysis

We propose an extended hazard regression model which allows the spread parameter to be dependent on covariates. This allows a broad class of models which includes the most common hazard models, such as the proportional hazards model, the accelerated failure time model and a proportional hazards/accelerated failure time hybrid model with constant spread parameter. Simulations based on sub-classes of this model suggest that maximum likelihood performs well even when only small or moderate-size data sets are available and the censoring pattern is heavy. The methodology provides a broad framework for analysis of reliability and survival data. Two numerical examples illustrate the results.     

Properties of Added Variable Plots in Cox's Regression Model

The added variable plot is useful for examining the effect of a covariate in regression models. The plot provides information regarding the inclusion of a covariate, and is useful in identifying influential observations on the parameter estimates. Hall et al. (1996) proposed a plot for Cox's proportional hazards model derived by regarding the Cox model as a generalized linear model. This paper proves and discusses properties of this plot. These properties make the plot a valuable tool in model evaluation. Quantities considered include parameter estimates, residuals, leverage, case influence measures and correspondence to previously proposed residuals and diagnostics.     

A Survey of Microcomputer Survival Analysis Software: The Need for an Integrated Framework

This paper surveys commercially available MS-DOS and Microsoft Windows based microcomputer software for survival analysis, especially for Cox proportional hazards regression and parametric survival models. Emphasis is given to functionality, documentation, generality, and flexibility of software. A discussion of the need for software integration is given, which leads to the conclusion that survival analysis software not closely tied to a well-designed package will not meet an analyst's general needs. Some standalone programs are good tools for teaching the theory of some survival analysis procedures, but they may not teach the student good data analysis techniques such as critically examining regression assumptions. We contrast typical software with a general, integrated, modeling framework that is available with S-PLUS.     

Parametric Analysis for Matched Pair Survival Data

Hougaard's (1986) bivariate Weibull distribution with positive stable frailties is applied to matched pairs survival data when either or both components of the pair may be censored and covariate vectors may be of arbitrary fixed length. When there is no censoring, we quantify the corresponding gain in Fisher information over a fixed-effects analysis. With the appropriate parameterization, the results take a simple algebraic form. An alternative marginal (independence working model) approach to estimation is also considered. This method ignores the correlation between the two survival times in the derivation of the estimator, but provides a valid estimate of standard error. It is shown that when both the correlation between the two survival times is high, and the ratio of the within-pair variability to the between-pair variability of the covariates is high, the fixed-effects analysis captures most of the information about the regression coefficient but the independence working model does badly. When the correlation is low, and/or most of the variability of the covariates occurs between pairs, the reverse is true. The random effects model is applied to data on skin grafts, and on loss of visual acuity among diabetics. In conclusion some extensions of the methods are indicated and they are placed in a wider context of Generalized Estimating Equation methodology.     

Regression analysis of case-cohort studies in the presence of dependent interval censoring

The case-cohort design is widely used as a means of reducing the cost in large cohort studies, especially when the disease rate is low and covariate measurements may be expensive, and has been discussed by many authors. In this paper, we discuss regression analysis of case-cohort studies that produce interval-censored failure time with dependent censoring, a situation for which there does not seem to exist an established approach. For inference, a sieve inverse probability weighting estimation procedure is developed with the use of Bernstein polynomials to approximate the unknown baseline cumulative hazard functions. The proposed estimators are shown to be consistent and the asymptotic normality of the resulting regression parameter estimators is established. A simulation study is conducted to assess the finite sample properties of the proposed approach and indicates that it works well in practical situations. The proposed method is applied to an HIV/AIDS case-cohort study that motivated this investigation.     

Choice of Parametric Accelerated Life and Proportional Hazards Models for Survival Data: Asymptotic Results

We discuss the impact of misspecifying fully parametric proportional hazards and accelerated life models. For the uncensored case, misspecified accelerated life models give asymptotically unbiased estimates of covariate effect, but the shape and scale parameters depend on the misspecification. The covariate, shape and scale parameters differ in the censored case. Parametric proportional hazards models do not have a sound justification for general use: estimates from misspecified models can be very biased, and misleading results for the shape of the hazard function can arise. Misspecified survival functions are more biased at the extremes than the centre. Asymptotic and first order results are compared. If a model is misspecified, the size of Wald tests will be underestimated. Use of the sandwich estimator of standard error gives tests of the correct size, but misspecification leads to a loss of power. Accelerated life models are more robust to misspecification because of their log-linear form. In preliminary data analysis, practitioners should investigate proportional hazards and accelerated life models; software is readily available for several such models.     

A Comparison of Methods for Analysing a Nested Frailty Model to Child Survival in Malawi

Demographic and Health Surveys collect child survival times that are clustered at the family and community levels. It is assumed that each cluster has a specific, unobservable, random frailty that induces an association in the survival times within the cluster. The Cox proportional hazards model, with family and community random frailties acting multiplicatively on the hazard rate, is presented. The estimation of the fixed effect and the association parameters of the modified model is then examined using the Gibbs sampler and the expectation–maximization (EM) algorithm. The methods are compared using child survival data collected in the 1992 Demographic and Health Survey of Malawi. The two methods lead to very similar estimates of fixed effect parameters. However, the estimates of random effect variances from the EM algorithm are smaller than those of the Gibbs sampler. Both estimation methods reveal considerable family variation in the survival of children, and very little variability over the communities.     

Regression Analysis of Left-truncated and Case I Interval-censored Data with the Additive Hazards Model

In recent years the analysis of interval-censored failure time data has attracted a great deal of attention and such data arise in many fields including demographical studies, economic and financial studies, epidemiological studies, social sciences, and tumorigenicity experiments. This is especially the case in medical studies such as clinical trials. In this article, we discuss regression analysis of one type of such data, Case I interval-censored data, in the presence of left-truncation. For the problem, the additive hazards model is employed and the maximum likelihood method is applied for estimations of unknown parameters. In particular, we adopt the sieve estimation approach that approximates the baseline cumulative hazard function by linear functions. The resulting estimates of regression parameters are shown to be consistent and efficient and have an asymptotic normal distribution. An illustrative example is provided.     

Empirical Likelihood Local Polynomial Regression Analysis of Clustered Data

Abstract. In this article, a naive empirical likelihood ratio is constructed for a non‐parametric regression model with clustered data, by combining the empirical likelihood method and local polynomial fitting. The maximum empirical likelihood estimates for the regression functions and their derivatives are obtained. The asymptotic distributions for the proposed ratio and estimators are established. A bias‐corrected empirical likelihood approach to inference for the parameters of interest is developed, and the residual‐adjusted empirical log‐likelihood ratio is shown to be asymptotically chi‐squared. These results can be used to construct a class of approximate pointwise confidence intervals and simultaneous bands for the regression functions and their derivatives. Owing to our bias correction for the empirical likelihood ratio, the accuracy of the obtained confidence region is not only improved, but also a data‐driven algorithm can be used for selecting an optimal bandwidth to estimate the regression functions and their derivatives. A simulation study is conducted to compare the empirical likelihood method with the normal approximation‐based method in terms of coverage accuracies and average widths of the confidence intervals/bands. An application of this method is illustrated using a real data set.