A three-state continuous time Markov chain model for HIV disease burden

Plasma HIV viral load (VL) is the clinical indicator used to evaluate disease burden for HIV-infected patients. We developed a covariate-adjusted, three-state, homogenous continuous time Markov chain model for HIV/AIDS disease burden among subgroups. We defined Detectable and Undetectable HIV VL levels as two transient states and Death as the third absorbing state. We implemented the exact maximum likelihood method to estimate the parameters with related asymptotic distribution to conduct hypothesis testing. We evaluated the proposed model using HIV-infected individuals from South Carolina (SC) HIV surveillance data. Using the developed model, we estimated and compared the transition hazards, transition probabilities, and the state-specific duration for HIV-infected individuals. We examined gender, race/ethnicity, age, CD4 count, place of residence, and antiretroviral treatment regimen prescribed at the beginning of the study period. We found that patients with a higher CD4 count, increased age, heterosexual orientation, white, and single tablet regimen users were associated with reduced risk of transitioning to a Detectable VL from an Undetectable VL, whereas shorter time since diagnosis, being male, and injection drug use increased the risk of the same transition.     

Bivariate modelling of longitudinal measurements of two human immunodeficiency type 1 disease progression markers in the presence of informative drop-outs

Summary.  The main statistical problem in many epidemiological studies which involve repeated measurements of surrogate markers is the frequent occurrence of missing data. Standard likelihood-based approaches like the linear random-effects model fail to give unbiased estimates when data are non-ignorably missing. In human immunodeficiency virus (HIV) type 1 infection, two markers which have been widely used to track progression of the disease are CD4 cell counts and HIV–ribonucleic acid (RNA) viral load levels. Repeated measurements of these markers tend to be informatively censored, which is a special case of non-ignorable missingness. In such cases, we need to apply methods that jointly model the observed data and the missingness process. Despite their high correlation, longitudinal data of these markers have been analysed independently by using mainly random-effects models. Touloumi and co-workers have proposed a model termed the joint multivariate random-effects model which combines a linear random-effects model for the underlying pattern of the marker with a log-normal survival model for the drop-out process. We extend the joint multivariate random-effects model to model simultaneously the CD4 cell and viral load data while adjusting for informative drop-outs due to disease progression or death. Estimates of all the model's parameters are obtained by using the restricted iterative generalized least squares method or a modified version of it using the EM algorithm as a nested algorithm in the case of censored survival data taking also into account non-linearity in the HIV–RNA trend. The method proposed is evaluated and compared with simpler approaches in a simulation study. Finally the method is applied to a subset of the data from the 'Concerted action on seroconversion to AIDS and death in Europe' study.     

Quantile Regression Methods for Longitudinal Data with Drop-outs: Application to CD4 Cell Counts of Patients Infected with the Human Immunodeficiency Virus

Patients infected with the human immunodeficiency virus (HIV) generally experience a decline in their CD4 cell count (a count of certain white blood cells). We describe the use of quantile regression methods to analyse longitudinal data on CD4 cell counts from 1300 patients who participated in clinical trials that compared two therapeutic treatments: zidovudine and didanosine. It is of scientific interest to determine any treatment differences in the CD4 cell counts over a short treatment period. However, the analysis of the CD4 data is complicated by drop-outs: patients with lower CD4 cell counts at the base-line appear more likely to drop out at later measurement occasions. Motivated by this example, we describe the use of `weighted' estimating equations in quantile regression models for longitudinal data with drop-outs. In particular, the conventional estimating equations for the quantile regression parameters are weighted inversely proportionally to the probability of drop-out. This approach requires the process generating the missing data to be estimable but makes no assumptions about the distribution of the responses other than those imposed by the quantile regression model. This method yields consistent estimates of the quantile regression parameters provided that the model for drop-out has been correctly specified. The methodology proposed is applied to the CD4 cell count data and the results are compared with those obtained from an `unweighted' analysis. These results demonstrate how an analysis that fails to account for drop-outs can mislead.     

Two-step and likelihood methods for joint models of longitudinal and survival data

We compare the commonly used two-step methods and joint likelihood method for joint models of longitudinal and survival data via extensive simulations. The longitudinal models include LME, GLMM, and NLME models, and the survival models include Cox models and AFT models. We find that the full likelihood method outperforms the two-step methods for various joint models, but it can be computationally challenging when the dimension of the random effects in the longitudinal model is not small. We thus propose an approximate joint likelihood method which is computationally efficient. We find that the proposed approximation method performs well in the joint model context, and it performs better for more “continuous” longitudinal data. Finally, a real AIDS data example shows that patients with higher initial viral load or lower initial CD4 are more likely to drop out earlier during an anti-HIV treatment.     

A multistate approach for estimating the incidence of human immunodeficiency virus by using data from a prevalent cohort study

Summary.  Cohort studies of individuals infected with the human immunodeficiency virus (HIV) provide useful information on the past pattern of HIV diagnoses, progression of the disease and use of antiretroviral therapy. We propose a new method for using individual data from an open prevalent cohort study to estimate the incidence of HIV, by jointly modelling the HIV diagnosis, the inclusion in the cohort and the progression of the disease in a Markov model framework. The estimation procedure involves the construction of a likelihood function which takes into account the probability of observing the total number of subjects who are enrolled in the cohort and the probabilities of passage through the stages of disease for each observed subject conditionally on being included in the cohort. The estimator of the HIV infection rate is defined as the function which maximizes a penalized likelihood, and the solution of this maximization problem is approximated on a basis of cubic M -splines. The method is illustrated by using cohort data from a hospital-based surveillance system of HIV infection in Aquitaine, a region of south-western France. A simulation study is performed to study the ability of the model to reconstruct the incidence of HIV from prevalent cohort data.     

Mixture Multi-state Markov Regression Model

Although heterogeneity across individuals may be reduced when a two-state process is extended into a multi-state process, the discrepancy between the observed and the predicted for some states may still exist owing to two possibilities, unobserved mixture distribution in the initial state and the effect of measured covariates on subsequent multi-state disease progression. In the present study, we developed a mixture Markov exponential regression model to take account of the above-mentioned heterogeneity across individuals (subject-to-subject variability) with a systematic model selection based on the likelihood ratio test. The model was successfully demonstrated by an empirical example on surveillance of patients with small hepatocellular carcinoma treated by non-surgical methods. The estimated results suggested that the model with the incorporation of unobserved mixture distribution behaves better than the one without. Complete and partial effects regarding risk factors on different subsequent multi-state transitions were identified using a homogeneous Markov model. The combination of both initial mixture distribution and homogeneous Markov exponential regression model makes a significant contribution to reducing heterogeneity across individuals and over time for disease progression.     

Estimating reference ranges from stratified two-stage samples

We present methodology for estimating age-specific reference ranges by using data from two-stage samples. On the basis of the information obtained in the first stage, the initial sample is stratified and random subsamples are drawn from each stratum, where the selection probabilities in this second-stage sampling may be different across strata in the population. The variable for which the reference ranges are to be established is measured at the second phase. The approach involves maximum likelihood estimation of the parameters of the age-specific distributions and separate estimation of the population stratum probabilities. These are combined to yield estimates of the quantiles of interest. The issue of variance estimation for the estimated quantiles is also addressed. The methodology is applied to the estimation of reference ranges for a cognitive test score in a study of non-demented older Japanese-Americans.     

Regression analysis for bivariate gap time with missing first gap time data

We consider ordered bivariate gap time while data on the first gap time are unobservable. This study is motivated by the HIV infection and AIDS study, where the initial HIV contracting time is unavailable, but the diagnosis times for HIV and AIDS are available. We are interested in studying the risk factors for the gap time between initial HIV contraction and HIV diagnosis, and gap time between HIV and AIDS diagnoses. Besides, the association between the two gap times is also of interest. Accordingly, in the data analysis we are faced with two-fold complexity, namely data on the first gap time is completely missing, and the second gap time is subject to induced informative censoring due to dependence between the two gap times. We propose a modeling framework for regression analysis of bivariate gap time under the complexity of the data. The estimating equations for the covariate effects on, as well as the association between, the two gap times are derived through maximum likelihood and suitable counting processes. Large sample properties of the resulting estimators are developed by martingale theory. Simulations are performed to examine the performance of the proposed analysis procedure. An application of data from the HIV and AIDS study mentioned above is reported for illustration.     

Generalised Linear Models Incorporating Population Level Information: An Empirical Likelihood Based Approach

In many situations information from a sample of individuals can be supplemented by population level information on the relationship between a dependent variable and explanatory variables. Inclusion of the population level information can reduce bias and increase the efficiency of the parameter estimates.Population level information can be incorporated via constraints on functions of the model parameters. In general the constraints are nonlinear making the task of maximum likelihood estimation harder. In this paper we develop an alternative approach exploiting the notion of an empirical likelihood. It is shown that within the framework of generalised linear models, the population level information corresponds to linear constraints, which are comparatively easy to handle. We provide a two-step algorithm that produces parameter estimates using only unconstrained estimation. We also provide computable expressions for the standard errors. We give an application to demographic hazard modelling by combining panel survey data with birth registration data to estimate annual birth probabilities by parity.     

Clustering work and family trajectories by using a divisive algorithm

Summary.  We present an approach to the construction of clusters of life course trajectories and use it to obtain ideal types of trajectories that can be interpreted and analysed meaningfully. We represent life courses as sequences on a monthly timescale and apply optimal matching analysis to compute dissimilarities between individuals. We introduce a new divisive clustering algorithm which has features that are in common with both Ward's agglomerative algorithm and classification and regression trees. We analyse British Household Panel Survey data on the employment and family trajectories of women. Our method produces clusters of sequences for which it is straightforward to determine who belongs to each cluster, making it easier to interpret the relative importance of life course factors in distinguishing subgroups of the population. Moreover our method gives guidance on selecting the number of clusters.     

Estimates of the likelihood of extreme returns in international stock markets

This study applies extreme-value theory to daily international stock-market returns to determine (1) whether or not returns follow a heavy-tailed stable distribution, (2) the likelihood of an extreme return, such as a 20% drop in a single day, and (3) whether or not the likelihood of an extreme event has changed since October 1987. Empirical results reject a heavy-tailed stable distribution for returns. Instead, a Student-t distribution or an autoregressive conditional heteroscedastic process is better able to capture the salient features of returns. We find that the likelihood of a large single-day return diff ers widely across markets and, for the G-7 countries, the 1987 stock-market drop appears to be largely an isolated event. A drop of this magnitude, however, is not rare in the case of Hong Kong. Finally, there is only limited evidence that the chance of a large single-day decline is more likely since the October 1987 market drop; however, exceptions include stock markets in Germany, The Netherlands and the UK.     

Mixtures of regression models with incomplete and noisy data

The estimation of the mixtures of regression models is usually based on the normal assumption of components and maximum likelihood estimation of the normal components is sensitive to noise, outliers, or high-leverage points. Missing values are inevitable in many situations and parameter estimates could be biased if the missing values are not handled properly. In this article, we propose the mixtures of regression models for contaminated incomplete heterogeneous data. The proposed models provide robust estimates of regression coefficients varying across latent subgroups even under the presence of missing values. The methodology is illustrated through simulation studies and a real data analysis.     

Regression coefficient analysis for correlated binomial outcomes

At present, the generalized estimating equation (GEE) and weighted least-squares (WLS) regression methods are the most widely used methods for analyzing correlated binomial data; both are easily implemented using existing software packages. We propose an alternative technique, i.e. regression coefficient analysis (RCA), for this type of data. In RCA, a regression equation is computed for each of n individuals; regression coefficients are averaged across the n equations to produce a regression equation, which predicts marginal probabilities and which can be tested to address hypotheses of different slopes between groups, slopes different from zero, different intercepts, etc. The method is computationally simple and can be performed using standard software. Simulations and examples are used to compare the power and robustness of RCA with those of the standard GEE and WLS methods. We find that RCA is comparable with the GEE method under the conditions tested, and suggest that RCA, within specified limitations, is a viable alternative to the GEE and WLS methods in the analysis of correlated binomial data.     

The generalized log-gamma mixture model with covariates: local influence and residual analysis

In a sample of censored survival times, the presence of an immune proportion of individuals who are not subject to death, failure or relapse, may be indicated by a relatively high number of individuals with large censored survival times. In this paper the generalized log-gamma model is modified for the possibility that long-term survivors may be present in the data. The model attempts to separately estimate the effects of covariates on the surviving fraction, that is, the proportion of the population for which the event never occurs. The logistic function is used for the regression model of the surviving fraction. Inference for the model parameters is considered via maximum likelihood. Some influence methods, such as the local influence and total local influence of an individual are derived, analyzed and discussed. Finally, a data set from the medical area is analyzed under the log-gamma generalized mixture model. A residual analysis is performed in order to select an appropriate model. The authors would like to thank the editor and referees for their helpful comments. This work was supported by CNPq, Brazil.     

Accounting for Follow-up Bias in Estimation of Human Immunodeficiency Virus Incidence Rates

The objective of this paper is to describe methods for estimating current incidence rates for human immunodeficiency virus (HIV) that account for follow-up bias. Follow-up bias arises when the incidence rate among individuals in a cohort who return for follow-up is different from the incidence rate among those who do not return. The methods are based on the use of early markers of HIV infection such as p24 antigen. The first method, called the cross-sectional method, uses only data collected at an initial base-line visit. The method does not require follow-up data but does require a priori knowledge of the mean duration of the marker (μ). A confidence interval procedure is developed that accounts for uncertainty in μ. The second method combines the base-line data from all individuals together with follow-up data from those individuals who return for follow-up. This method has the distinct advantage of not requiring prior information about μ. Several confidence interval procedures for the incidence rate are compared by simulation. The methods are applied to a study in India to estimate current HIV incidence. These data suggest that the epidemic is growing rapidly in some subpopulations in India.     

A Bayesian Analysis of Correlated Interval-Censored Data

In epidemiological studies where subjects are seen periodically on follow-up visits, interval-censored data occur naturally. The exact time the change of state (such as HIV seroconversion) occurs is not known exactly, only that it occurred within some time interval. In multi-stage sampling or partner tracing studies, individuals are grouped into smaller subgroups. Individuals within a subgroup share an unobservable specific frailty which induces correlation within the subgroup. In this paper, we consider a Bayesian model for analysing correlated interval-censored data. Parameters are estimated using the Markov chain Monte Carlo methods, specifically the Gibbs sampler.     

Statistical issues in phase III human immunodeficiency virus vaccine trial design

Several scientific questions are of interest in phase III trials of prophylactic human immunodeficiency virus (HIV) vaccines. In this paper we focus on some issues related to evaluating the direct protective effects of a vaccine in reducing susceptibility, VE_S, and its effect on reducing infectiousness, VE_I. An estimation of VE_I generally requires information on contacts between the infective and susceptible individuals. By augmenting the primary participants of an HIV vaccine trial by their steady sexual partners, information can be collected that allows an estimation of VE_I as well as VE_S. Exposure to infection information, however, may be expensive and difficult to collect. A vaccine trial design can include a small validation set with good exposure to infection data to correct bias in a larger, simpler main study with only coarse exposure data. The large main study increases the efficiency of the small validation set. More research into the combination of different levels of information in vaccine trial design will yield more efficient and less biased estimates of the efficacy measures of interest.     

Inference for the random coefficients bifurcating autoregressive model for cell lineage studies

Cell lineage data consist of observations on quantitative characteristics of the descendants of an initial cell. The bifurcating autoregressive model has been previously used to model dependencies in cell lineage data by considering each line of descent to be a first order autoregressive process and allowing the environmental effects of sisters to be correlated. Here the basic bifurcating autoregressive model is modified to include random coefficients which allows for the relationship between mother and daughter cells to depend on environmental factors. Maximum likelihood inference under the assumption of multivariate normality is considered and the method is illustrated on several data sets.     

A nonparametric Bayesian model for inference in related longitudinal studies

Summary.  We discuss a method for combining different but related longitudinal studies to improve predictive precision. The motivation is to borrow strength across clinical studies in which the same measurements are collected at different frequencies. Key features of the data are heterogeneous populations and an unbalanced design across three studies of interest. The first two studies are phase I studies with very detailed observations on a relatively small number of patients. The third study is a large phase III study with over 1500 enrolled patients, but with relatively few measurements on each patient. Patients receive different doses of several drugs in the studies, with the phase III study containing significantly less toxic treatments. Thus, the main challenges for the analysis are to accommodate heterogeneous population distributions and to formalize borrowing strength across the studies and across the various treatment levels. We describe a hierarchical extension over suitable semiparametric longitudinal data models to achieve the inferential goal. A nonparametric random-effects model accommodates the heterogeneity of the population of patients. A hierarchical extension allows borrowing strength across different studies and different levels of treatment by introducing dependence across these nonparametric random-effects distributions. Dependence is introduced by building an analysis of variance (ANOVA) like structure over the random-effects distributions for different studies and treatment combinations. Model structure and parameter interpretation are similar to standard ANOVA models. Instead of the unknown normal means as in standard ANOVA models, however, the basic objects of inference are random distributions, namely the unknown population distributions under each study. The analysis is based on a mixture of Dirichlet processes model as the underlying semiparametric model.