Identifying dietary consumption patterns from survey data: a Bayesian nonparametric latent class model

Dietary assessments provide the snapshots of population-based dietary habits. Questions remain about how generalisable those snapshots are in national survey data, where certain subgroups are sampled disproportionately. We propose a Bayesian overfitted latent class model to derive dietary patterns, accounting for survey design and sampling variability. Compared to standard approaches, our model showed improved identifiability of the true population pattern and prevalence in simulation. We focus application of this model to identify the intake patterns of adults living at or below the 130% poverty income level. Five dietary patterns were identified and characterised by reproducible code/data made available to encourage further research.

Estimating poultry aspergillosis prevalence and diagnostic accuracy of histopathological and mycological culture in Côte d'Ivoire using Bayesian latent class analysis

This study aimed to estimate the prevalence of poultry aspergillosis and evaluate the accuracy of histopathology (test under evaluation) and mycological culture (an imperfect reference test). Farms raising layer and breeder or broiler birds, with suspected aspergillosis cases, clinical or subclinical, were eligible and visited for sampling. After necropsy, histopathology and mycological culture examinations were conducted by two evaluators. A Bayesian latent class model was used to estimate the accuracy of histopathology when compared to the imperfect reference test, mycological culture. A total of 142 chicken farms, 96 laying and breeding hen farms, and 46 broiler farms were used for the study. True aspergillosis median prevalence was estimated at 63.7% (95% credibility intervals, CrI: 53.8%, 73.0%) in layers and breeders and at 65.2% (95% CrI: 50.2%, 78.3%) in the broiler farms' population. The median diagnostic sensitivity of histopathology and culture were estimated at, respectively, 98.8% (95% CrI: 94.6%, 100.0%) and 90.4% (95% CrI: 83.6%, 95.3%). Tests' diagnostic specificity was estimated at, respectively, 97.3% (95% CrI: 87.7%, 99.9%) and 95.7% (95% CrI: 91.8%, 98.2%). Both tests had very high and comparable positive predictive values, but, in a population where disease prevalence was 25%, histopathology had a higher negative predictive value than culture.

Joint clustering multiple longitudinal features: A comparison of methods and software packages with practical guidance

Clustering longitudinal features is a common goal in medical studies to identify distinct disease developmental trajectories. Compared to clustering a single longitudinal feature, integrating multiple longitudinal features allows additional information to be incorporated into the clustering process, which may reveal co-existing longitudinal patterns and generate deeper biological insight. Despite its increasing importance and popularity, there is limited practical guidance for implementing cluster analysis approaches for multiple longitudinal features and evaluating their comparative performance in medical datasets. In this paper, we provide an overview of several commonly used approaches to clustering multiple longitudinal features, with an emphasis on application and implementation through R software. These methods can be broadly categorized into two categories, namely model-based (including frequentist and Bayesian) approaches and algorithm-based approaches. To evaluate their performance, we compare these approaches using real-life and simulated datasets. These results provide practical guidance to applied researchers who are interested in applying these approaches for clustering multiple longitudinal features. Recommendations for applied researchers and suggestions for future research in this area are also discussed.

A latent functional approach for modeling the effects of multidimensional exposures on disease risk

Understanding the relationships between exposure and disease incidence is an important problem in environmental epidemiology. Typically, a large number of these exposures are measured, and it is found either that a few exposures transmit risk or that each exposure transmits a small amount of risk, but, taken together, these may pose a substantial disease risk. Further, these exposure effects can be nonlinear. We develop a latent functional approach, which assumes that the individual effect of each exposure can be characterized as one of a series of unobserved functions, where the number of latent functions is less than or equal to the number of exposures. We propose Bayesian methodology to fit models with a large number of exposures and show that existing Bayesian group LASSO approaches are a special case of the proposed model. An efficient Markov chain Monte Carlo sampling algorithm is developed for carrying out Bayesian inference. The deviance information criterion is used to choose an appropriate number of nonlinear latent functions. We demonstrate the good properties of the approach using simulation studies. Further, we show that complex exposure relationships can be represented with only a few latent functional curves. The proposed methodology is illustrated with an analysis of the effect of cumulative pesticide exposure on cancer risk in a large cohort of farmers.

Global patterns of cancer transitions: A modelling study

Cancer is a major contributor to global disease burden. Many countries experienced or are experiencing the transition that non-infection-related cancers replace infection-related cancers. We aimed to characterise burden changes for major types of cancers and identify global transition patterns. We focused on 10 most common cancers worldwide and extracted age-standardised incidence and mortality in 204 countries and territories from 1990 to 2019 through the Global Burden of Disease Study. Two-stage modelling design was used. First, we applied growth mixture models (GMMs) to identify distinct trajectories for incidence and mortality of each cancer type. Next, we performed latent class analysis to detect cancer transition patterns based on the categorisation results from GMMs. Kruskal-Wallis H tests were conducted to evaluate associations between transition patterns and socioeconomic indicators. Three distinct patterns were identified as unfavourable, intermediate and favourable stages. Trajectories of lung and breast cancers had the strongest association with transition patterns among men and women. The unfavourable stage was characterised by rapid increases in lung, breast and colorectal cancers alongside stable or decreasing burden of gastric, cervical, oesophageal and liver cancers. In contrast, the favourable stage exhibited rapid declines in most cancers. The unfavourable stage was associated with lower sociodemographic index, health expenditure, gross domestic product per capita and higher maternal mortality ratio (P < .001 for all associations). Our findings suggest that unfavourable, intermediate and favourable transition patterns exist. Countries and territories in the unfavourable stage tend to be socioeconomically disadvantaged, and tailored intervention strategies are needed in these resource-limited settings.

Uncovering circadian rhythms in metabolic longitudinal data: A Bayesian latent class modeling approach

Researchers in biology and medicine have increasingly focused on characterizing circadian rhythms and their potential impact on disease. Understanding circadian variation in metabolomics, the study of chemical processes involving metabolites may provide insight into important aspects of biological mechanism. Of scientific importance is developing a statistical rigorous approach for characterizing different types of 24-hour patterns among high dimensional longitudinal metabolites. We develop a latent class approach to incorporate variation in 24-hour patterns across metabolites where profiles are modeled with finite mixtures of distinct shape-invariant circadian curves that themselves incorporate variation in amplitude and phase across metabolites. An efficient Markov chain Monte Carlo sampling is used to carry out Bayesian posterior computation. When the model was fit separately by individual to the data from a small group of participants, two distinct 24-hour rhythms were identified, with one being sinusoidal and the other being more complex with multiple peaks. Interestingly, the latent pattern associated with circadian variation (simple sinusoidal curve) had a similar phase across the three participants, while the more complex latent pattern reflecting diurnal variation differed across individual. The results suggested that this modeling framework can be used to separate 24-hour rhythms into an endogenous circadian and one or more exogenous diurnal patterns in describing human metabolism.

Bayesian Inference for an Unknown Number of Attributes in Restricted Latent Class Models

The specification of the [Formula: see text] matrix in cognitive diagnosis models is important for correct classification of attribute profiles. Researchers have proposed many methods for estimation and validation of the data-driven [Formula: see text] matrices. However, inference of the number of attributes in the general restricted latent class model remains an open question. We propose a Bayesian framework for general restricted latent class models and use the spike-and-slab prior to avoid the computation issues caused by the varying dimensions of model parameters associated with the number of attributes, K. We develop an efficient Metropolis-within-Gibbs algorithm to estimate K and the corresponding [Formula: see text] matrix simultaneously. The proposed algorithm uses the stick-breaking construction to mimic an Indian buffet process and employs a novel Metropolis-Hastings transition step to encourage exploring the sample space associated with different values of K. We evaluate the performance of the proposed method through a simulation study under different model specifications and apply the method to a real data set related to a fluid intelligence matrix reasoning test.

Consumers' Preferences for Apple Production Attributes: Results of a Choice Experiment

Various food safety and environmental problems in China have raised consumer awareness of food safety issues and negative environmental impacts in various supply chains. This research assessed consumer preferences and willingness to pay (WTP) for food safety and ecosystem delivery attributes associated with apples, demonstrated through the application of different traceability systems. Research participants were recruited in Beijing (N = 384) and Shanghai (N = 320). Choice experiment methodology was applied. The data were analyzed using conditional logit, random parameter logit, and latent class models; the results indicated significant consumer preferences for traceability information, including in relation to lower pesticide usage and application of organic fertilizer during primary production. The results also indicated that participants in this research had a significant willingness-to-pay premium for apple products that had production information traceability, had reduced pesticide use, and were grown with organic fertilizers. The models demonstrated heterogeneous preferences among participants such that consumers could be divided into three classes: non-price-sensitive (53.5%), pesticide-sensitive (21.7%), and price-sensitive (24.8%).

Many phish in the C $\mathcal{C}$ : A coexisting-choice-criteria model of security behavior

Normative decision theory proves inadequate for modeling human responses to the social-engineering campaigns of advanced persistent threat (APT) attacks. Behavioral decision theory fares better, but still falls short of capturing social-engineering attack vectors which operate through emotions and peripheral-route persuasion. We introduce a generalized decision theory, under which any decision will be made according to one of multiple coexisting choice criteria. We denote the set of possible choice criteria by C $\mathcal {C}$ . Thus, the proposed model reduces to conventional Expected Utility theory when| C EU | = 1 $|\mathcal {C}_{\text{EU}}|=1$ , while Dual-Process (thinking fast vs. thinking slow) decision making corresponds to a model with| C DP | = 2 $|\mathcal {C}_{\text{DP}}|=2$ . We consider a more general case with| C | ≥ 2 $|\mathcal {C}|\ge 2$ , which necessitates careful consideration of how, for a particular choice-task instance, one criterion comes to prevail over others. We operationalize this with a probability distribution that is conditional upon traits of the decisionmaker as well as upon the context and the framing of choice options. Whereas existing signal detection theory (SDT) models of phishing detection commingle the different peripheral-route persuasion pathways, in the present descriptive generalization the different pathways are explicitly identified and represented. A number of implications follow immediately from this formulation, ranging from the conditional nature of security-breach risk to delineation of the prerequisites for valid tests of security training. Moreover, the model explains the "stepping-stone" penetration pattern of APT attacks, which has confounded modeling approaches based on normative rationality.

The Data-Adaptive Fellegi-Sunter Model for Probabilistic Record Linkage: Algorithm Development and Validation for Incorporating Missing Data and Field Selection

Background

Quality patient care requires comprehensive health care data from a broad set of sources. However, missing data in medical records and matching field selection are 2 real-world challenges in patient-record linkage.

Objective

In this study, we aimed to evaluate the extent to which incorporating the missing at random (MAR)–assumption in the Fellegi-Sunter model and using data-driven selected fields improve patient-matching accuracy using real-world use cases.

Methods

We adapted the Fellegi-Sunter model to accommodate missing data using the MAR assumption and compared the adaptation to the common strategy of treating missing values as disagreement with matching fields specified by experts or selected by data-driven methods. We used 4 use cases, each containing a random sample of record pairs with match statuses ascertained by manual reviews. Use cases included health information exchange (HIE) record deduplication, linkage of public health registry records to HIE, linkage of Social Security Death Master File records to HIE, and deduplication of newborn screening records, which represent real-world clinical and public health scenarios. Matching performance was evaluated using the sensitivity, specificity, positive predictive value, negative predictive value, and F1-score.

Results

Incorporating the MAR assumption in the Fellegi-Sunter model maintained or improved F1-scores, regardless of whether matching fields were expert-specified or selected by data-driven methods. Combining the MAR assumption and data-driven fields optimized the F1-scores in the 4 use cases.

Conclusions

MAR is a reasonable assumption in real-world record linkage applications: it maintains or improves F1-scores regardless of whether matching fields are expert-specified or data-driven. Data-driven selection of fields coupled with MAR achieves the best overall performance, which can be especially useful in privacy-preserving record linkage.

Simplifying the estimation of diagnostic testing accuracy over time for high specificity tests in the absence of a gold standard

Many different methods for evaluating diagnostic test results in the absence of a gold standard have been proposed. In this paper, we discuss how one common method, a maximum likelihood estimate for a latent class model found via the Expectation-Maximization (EM) algorithm can be applied to longitudinal data where test sensitivity changes over time. We also propose two simplified and nonparametric methods which use data-based indicator variables for disease status and compare their accuracy to the maximum likelihood estimation (MLE) results. We find that with high specificity tests, the performance of simpler approximations may be just as high as the MLE.

Factors affecting injury severity and the number of vehicles involved in a freeway traffic accident: investigating their heterogeneous effects by facility type using a latent class approach

The number of vehicles involved in a traffic accident can be representative of the severity of the accident and provide profound insight into the diverse factors affecting severity, which cannot be identified through the victim fatality rate. This paper presents an analysis and comparison between the effects of factors affecting injury severity and the number of involved vehicles. In this study, a latent class model was used to investigate the unobserved heterogeneity of the accident factors. Freeway facility types are latent factors that affect the heterogeneity of the effects of accident factors. The class mainly including accidents at the freeway mainline sections included more injury/fatal accidents and multiple-vehicle accidents and more significant accident factor estimation results than the other class including accidents at the tollgates or ramps. Among these factors, night-time, faults made by the driver, and heavy vehicle accidents were found to increase the accident severity. Investigating accident factors affecting both the injury severity and number of involved vehicles is important as the number of people who are injured or dead is likely to increase when multiple vehicles are involved in the accident.

Estimation of diagnostic test accuracy: A "Rule of Three" for data with repeated observations but without a gold standard

This article considers how to estimate the accuracy of a diagnostic test when there are repeated observations, but without the availability of a gold standard or reference test. We identify conditions under which the structure of the observed data is rich enough to provide sufficient degrees of freedom, such that a suitable latent class model can be fitted with identifiable accuracy parameters. We show that a Rule of Three applies, specifying that accuracy can be evaluated as long as there are at least three observations per individual with the given test. This rule also applies if the three observations arise from combinations of different test methods, or from a sequential design in which individuals are tested for a maximum number of times with the same test but stopping if a positive (or negative) result occurs. The rule pertains to tests having an arbitrary number of response categories. Accuracy is evaluated by parameters reflecting rates of misclassification among the response categories, and the model also provides estimates of the underlying distribution of the true disease state. These ideas are illustrated by data from two medical studies. Issues discussed include the advantages and disadvantages of analyzing the response variable as binary or multinomial, as well as the feasibility of testing goodness of fit when the model incorporates a large number of parameters. Comparisons are possible between models that do or do not assume equal accuracy rates for the observations, and between models where certain misclassification parameters are or are not assumed to be zero.

Does Controlling for Scale Heterogeneity Better Explain Respondents' Preference Segmentation in Discrete Choice Experiments? A Case Study of US Health Insurance Demand

Analyses of preference evidence frequently confuse heterogeneity in the effects of attribute parameters (i.e., taste coefficients) and the scale parameter (i.e., variance). Standard latent class models often produce unreasonable classes with high variance and disordered coefficients because of confounding estimates of effect and scale heterogeneity. In this study, we estimated a scale-adjusted latent class model in which scale classes (heteroskedasticity) were identified using respondents' randomness in choice behavior on the internet panel (e.g., time to completion and time of day). Hence, the model distinctly explained the taste/preference variation among classes associated with individual socioeconomic characters, in which scales are adjusted. Using data from a discrete-choice experiment on US health insurance demand among single employees, the results demonstrated how incorporating behavioral data enhances the interpretation of heterogeneous effects. Once scale heterogeneity was controlled, we found substantial heterogeneity with 4 taste classes. Two of the taste classes were highly premium sensitive (economy class), coming mostly from the low-income group, and the class associated with better educational backgrounds preferred to have a better quality of coverage of health insurance plans. The third class was a highly quality-sensitive class, with a higher SES background and lower self-stated health condition. The last class was identified as stayers, who were not premium or quality sensitive. This case study demonstrates that one size does not fit all in the analysis of preference heterogeneity. The novel use of behavioral data in the latent class analysis is generalizable to a wide range of health preference studies.

Effect of decision rules in choice experiments on hunting and bushmeat trade

Providing insight on decisions to hunt and trade bushmeat can facilitate improved management interventions that typically include enforcement, alternative employment, and donation of livestock. Conservation interventions to regulate bushmeat hunting and trade have hitherto been based on assumptions of utility- (i.e., personal benefits) maximizing behavior, which influences the types of incentives designed. However, if individuals instead strive to minimize regret, interventions may be misguided. We tested support for 3 hypotheses regarding decision rules through a choice experiment in Tanzania. We estimated models based on the assumptions of random utility maximization (RUM) and pure random regret maximization (P-RRM) and combinations thereof. One of these models had an attribute-specific decision rule and another had a class-specific decision rule. The RUM model outperformed the P-RRM model, but the attribute-specific model performed better. Allowing respondents with different decision rules and preference heterogeneity within each decision rule in a class-specific model performed best, revealing that 55% of the sample used a P-RRM decision rule. Individuals using a P-RRM decision rule responded less to enforcement, salary, and livestock donation than did individuals using the RUM decision rule. Hence, 3 common strategies, enforcement, alternative income-generating activities, and providing livestock as a substitute protein, are likely less effective in changing the behavior of more than half of respondents. Only salary elicited a large (i.e. elastic) response, and only for one RUM class. Policies to regulate the bushmeat trade based solely on the assumption of individuals maximizing utility, may fail for a significant proportion of the sample. Despite the superior performance of models that allow both RUM and P-RRM decision rules there are drawbacks that must be considered before use in the Global South, where very little is known about the social-psychology of decision making.

Internet, gaming, and smartphone usage patterns of children and adolescents in Korea: A c-CURE clinical cohort study

BACKGROUND AND AIM

Whereas many studies on Internet gaming disorder (IGD) have used self-report questionnaires, only a few have adopted clinical interviews and samples. The current study aimed at using data from face-to-face diagnostic interviews, based on the criteria for IGD in the DSM-5, to determine the Internet, gaming, and smartphone usage patterns of children and adolescents.

METHODS

A latent class analysis was conducted using data collected through diagnostic interviews for Internet, gaming, and smartphone addiction with 190 participants (M = 13.14 years, SD = 2.46; 143 boys, 47 girls) who were part of a multicenter clinical cohort study.

RESULTS

Participants were classified into four groups: pleasure-seeking (Class 1), internal-use (Class 2), problematic-use (Class 3), and pathological-use (Class 4). The pleasure-seeking group (8.11%) showed low tendencies in general and proper control. The internal-use group (17.63%) showed significant increases in "cognitive salience" and "craving," with strong internal desires. The problematic-use group (37.28%) had no "interference with role performance"; however, they displayed "difficulty regulating use" and "persistent use despite negative consequences," with a slight functional impairment. The pathological-use group (36.98%) scored the highest on all these items, revealing a severe functional impairment. Compared to the other groups, the pathological-use group had the highest depression and daily stress levels and displayed the lowest levels of happiness.

CONCLUSIONS

This study provides basic data to elucidate Internet, gaming, and smartphone overuse patterns among children and adolescents, which could be used to develop differentiated intervention strategies for each group.

Integrating latent classes in the Bayesian shared parameter joint model of longitudinal and survival outcomes

Cystic fibrosis is a chronic lung disease requiring frequent lung-function monitoring to track acute respiratory events (pulmonary exacerbations). The association between lung-function trajectory and time-to-first exacerbation can be characterized using joint longitudinal-survival modeling. Joint models specified through the shared parameter framework quantify the strength of association between such outcomes but do not incorporate latent sub-populations reflective of heterogeneous disease progression. Conversely, latent class joint models explicitly postulate the existence of sub-populations but do not directly quantify the strength of association. Furthermore, choosing the optimal number of classes using established metrics like deviance information criterion is computationally intensive in complex models. To overcome these limitations, we integrate latent classes in the shared parameter joint model through a fully Bayesian approach. To choose the optimal number of classes, we construct a mixture model assuming more latent classes than present in the data, thereby asymptotically "emptying" superfluous latent classes, provided the Dirichlet prior on class proportions is sufficiently uninformative. Model properties are evaluated in simulation studies. Application to data from the US Cystic Fibrosis Registry supports the existence of three sub-populations corresponding to lung-function trajectories with high initial forced expiratory volume in 1 s (FEV₁), rapid FEV₁ decline, and low but steady FEV₁ progression. The association between FEV₁ and hazard of exacerbation was negative in each class, but magnitude varied.

A pseudolikelihood approach for assessing genetic association in case-control studies with unmeasured population structure

The case-control study design is one of the main tools for detecting associations between genetic markers and diseases. It is well known that population substructure can lead to spurious association between disease status and a genetic marker if the prevalence of disease and the marker allele frequency vary across subpopulations. In this paper, we propose a novel statistical method to estimate the association in case-control studies with unmeasured population substructure. The proposed method takes two steps. First, the information on genomic markers and disease status is used to infer the population substructure; second, the association between the disease and the test marker adjusting for the population substructure is modeled and estimated parametrically through polytomous logistic regression. The performance of the proposed method, relative to the existing methods, on bias, coverage probability and computational time, is assessed through simulations. The method is applied to an end-stage renal disease study in African Americans population.

Test characteristics of the tuberculin skin test and post-mortem examination for bovine tuberculosis diagnosis in cattle in Northern Ireland estimated by Bayesian latent class analysis with adjustments for covariates

The single intradermal comparative cervical tuberculin (SICCT) test and post-mortem examination are the main diagnostic tools for bovine tuberculosis (bTB) in cattle in the British Isles. Latent class modelling is often used to estimate the bTB test characteristics due to the absence of a gold standard. However, the reported sensitivity of especially the SICCT test has shown a lot of variation. We applied both the Hui–Walter latent class model under the Bayesian framework and the Bayesian model specified at the animal level, including various risk factors as predictors, to estimate the SICCT test and post-mortem test characteristics. Data were collected from all cattle slaughtered in abattoirs in Northern Ireland in 2015. Both models showed comparable posterior median estimation for the sensitivity of the SICCT test (88.61% and 90.56%, respectively) using standard interpretation and for post-mortem examination (53.65% and 53.79%, respectively). Both models showed almost identical posterior median estimates for the specificity (99.99% vs. 99.80% for SICCT test at standard interpretation and 99.66% vs. 99.86% for post-mortem examination). The animal-level model showed slightly narrower posterior 95% credible intervals. Notably, this study was carried out in slaughtered cattle which may not be representative for the general cattle population.

Bayesian hierarchical latent class models for estimating diagnostic accuracy

The diagnostic accuracy of a test or rater has a crucial impact on clinical decision making. The assessment of diagnostic accuracy for multiple tests or raters also merits much attention. A Bayesian hierarchical conditional independence latent class model for estimating sensitivities and specificities for a large group of tests or raters is proposed, which is applicable to both with-gold-standard and without-gold-standard situations. Through the hierarchical structure, not only are the sensitivities and specificities of individual tests estimated, but also the diagnostic performance of the whole group of tests. For a small group of tests or raters, the proposed model is further extended by introducing pairwise covariances between tests to improve the fitting and to allow for more modeling flexibility. Correlation residual analysis is applied to detect any significant covariance between multiple tests. Just Another Gibbs Sampler (JAGS) implementation is efficiently adopted for both models. Three real data sets from literature are analyzed to explicitly illustrate the proposed methods.

A Bayesian multi-dimensional couple-based latent risk model with an application to infertility

Motivated by the Longitudinal Investigation of Fertility and the Environment (LIFE) Study that investigated the association between exposure to a large number of environmental pollutants and human reproductive outcomes, we propose a joint latent risk class modeling framework with an interaction between female and male partners of a couple. This formulation introduces a dependence structure between the chemical patterns within a couple and between the chemical patterns and the risk of infertility. The specification of an interaction enables the interplay between the female and male's chemical patterns on the risk of infertility in a parsimonious way. We took a Bayesian perspective to inference and used Markov chain Monte Carlo algorithms to obtain posterior estimates of model parameters. We conducted simulations to examine the performance of the estimation approach. Using the LIFE Study dataset, we found that in addition to the effect of PCB exposures on females, the male partners' PCB exposures play an important role in determining risk of infertility. Further, this risk is subadditive in the sense that there is likely a ceiling effect which limits the probability of infertility when both partners of the couple are at high risk.

To "vape" or smoke? Experimental evidence on adult smokers

A growing share of the United States population uses e-cigarettes but the optimal regulation of these controversial products remains an open question. We conduct a discrete choice experiment to investigate how adult tobacco cigarette smokers' demand for e-cigarettes and tobacco cigarettes varies by four attributes: (i) whether e-cigarettes are considered healthier than tobacco cigarettes, (ii) the effectiveness of e-cigarettes as a cessation device, (iii) bans on use in public places, and (iv) price. We find that adult smokers' demand for e-cigarettes is motivated more by health concerns than by the desire to avoid smoking bans or higher prices.

A shared-parameter continuous-time hidden Markov and survival model for longitudinal data with informative dropout

A shared-parameter approach for jointly modeling longitudinal and survival data is proposed. With respect to available approaches, it allows for time-varying random effects that affect both the longitudinal and the survival processes. The distribution of these random effects is modeled according to a continuous-time hidden Markov chain so that transitions may occur at any time point. For maximum likelihood estimation, we propose an algorithm based on a discretization of time until censoring in an arbitrary number of time windows. The observed information matrix is used to obtain standard errors. We illustrate the approach by simulation, even with respect to the effect of the number of time windows on the precision of the estimates, and by an application to data about patients suffering from mildly dilated cardiomyopathy.

Different latent class models were used and evaluated for assessing the accuracy of campylobacter diagnostic tests: overcoming imperfect reference standards?

In the absence of perfect reference standard, classical techniques result in biased diagnostic accuracy and prevalence estimates. By statistically defining the true disease status, latent class models (LCM) constitute a promising alternative. However, LCM is a complex method which relies on parametric assumptions, including usually a conditional independence between tests and might suffer from data sparseness. We carefully applied LCMs to assess new campylobacter infection detection tests for which bacteriological culture is an imperfect reference standard. Five diagnostic tests (culture, polymerase chain reaction and three immunoenzymatic tests) of campylobacter infection were collected in 623 patients from Bordeaux and Lyon Hospitals, France. Their diagnostic accuracy were estimated with standard and extended LCMs with a thorough examination of models goodness-of-fit. The model including a residual dependence specific to the immunoenzymatic tests best complied with LCM assumptions. Asymptotic results of goodness-of-fit statistics were substantially impaired by data sparseness and empirical distributions were preferred. Results confirmed moderate sensitivity of the culture and high performances of immunoenzymatic tests. LCMs can be used to estimate diagnostic tests accuracy in the absence of perfect reference standard. However, their implementation and assessment require specific attention due to data sparseness and limitations of existing software.

People's Risk Recognition Preceding Evacuation and Its Role in Demand Modeling and Planning

Evacuation planning and management involves estimating the travel demand in the event that such action is required. This is usually done as a function of people's decision to evacuate, which we show is strongly linked to their risk awareness. We use an empirical data set, which shows tsunami evacuation behavior, to demonstrate that risk recognition is not synonymous with objective risk, but is instead determined by a combination of factors including risk education, information, and sociodemographics, and that it changes dynamically over time. Based on these findings, we formulate an ordered logit model to describe risk recognition combined with a latent class model to describe evacuation choices. Our proposed evacuation choice model along with a risk recognition class can evaluate quantitatively the influence of disaster mitigation measures, risk education, and risk information. The results obtained from the risk recognition model show that risk information has a greater impact in the sense that people recognize their high risk. The results of the evacuation choice model show that people who are unaware of their risk take a longer time to evacuate.

Two-Stage maximum likelihood estimation in the misspecified restricted latent class model

The maximum likelihood classification rule is a standard method to classify examinee attribute profiles in cognitive diagnosis models (CDMs). Its asymptotic behaviour is well understood when the model is assumed to be correct, but has not been explored in the case of misspecified latent class models. This paper investigates the asymptotic behaviour of a two-stage maximum likelihood classifier under a misspecified CDM. The analysis is conducted in a general restricted latent class model framework addressing all types of CDMs. Sufficient conditions are proposed under which a consistent classification can be obtained by using a misspecified model. Discussions are also provided on the inconsistency of classification under certain model misspecification scenarios. Simulation studies and a real data application are conducted to illustrate these results. Our findings can provide some guidelines as to when a misspecified simple model or a general model can be used to provide a good classification result.

Which attributes of whole genome sequencing tests are most important to the general population? Results from a German preference study

Objective

The aim of this study was to identify the preferences for whole genome sequencing (WGS) tests without genetic counseling.

Methods

A discrete choice experiment was conducted where participants chose between two hypothetical alternatives consisting of the following attributes: test accuracy, test costs, identified diseases, probability of disease occurrence, and data access. People from the general German population aged ≥18 years were eligible to participate in the survey. We estimated generalized linear mixed effects models, latent class mixed-logit models, and the marginal willingness to pay.

Results

Three hundred and one participants were included in the final analysis. Overall, the most favored WGS testing attributes were 95% test accuracy, report of severe hereditary diseases and 40% probability of disease development, test costs of €1,000, and access to test results for researchers. Subgroup analysis, however, showed differences in these preferences between males and females. For example, males preferred reporting of results at a 10% probability of disease development and females preferred reporting of results at a 40% probability. The test cost, participant’s educational level, and access to data influenced the willingness to participate in WGS testing in reality.

Conclusion

The German general population was aware of the importance of genetic research and preferred to provide their own genetic data for researchers. However, among others, the reporting of results with a comparatively relatively low probability of disease development at a level of 40%, and the test accuracy of 95% had a high preference. This shows that the results and consequences of WGS testing without genetic counseling are hard to assess for individuals. Therefore, WGS testing should be supported by qualified genetic counseling, where the attributes and consequences are explained.

Joint assessment of handedness and footedness through latent class factor analysis

Recently, some studies have demonstrated the effectiveness of two latent variable approaches in which hand preferences are analysed using either latent class methods or latent class factor (LCF) methods. The main aims of this study are: (i) to establish whether these approaches are adequate for assessing footedness, (ii) to evaluate their appropriateness when hand and foot preferences are jointly analysed, and (iii) to measure the association between handedness and footedness based on the examined latent variable models. To this end, a dataset providing information about the limb used to perform ten hand actions and three foot movements by 2236 young Italian sportspeople is analysed. The first aim is pursued through an exploratory analysis of the observed foot preferences; according to this analysis, footedness patterns are adequately described by two latent levels of footedness. As far as the second aim is concerned, a confirmatory analysis of foot and hand preferences is carried out; the best fit to the dataset is obtained using a two-dimensional LCF model with four latent levels of handedness and two latent levels of footedness. Finally, the association between handedness and footedness resulting from the employed methods is remarkably lower than that registered in other studies.

Evaluation of Student Performance through a Multidimensional Finite Mixture IRT Model

In the Italian academic system, a student can enroll for an exam immediately after the end of the teaching period or can postpone it; in this second case the exam result is missing. We propose an approach for the evaluation of a student performance throughout the course of study, accounting also for nonattempted exams. The approach is based on an item response theory model that includes two discrete latent variables representing student performance and priority in selecting the exams to take. We explicitly account for nonignorable missing observations as the indicators of attempted exams also contribute to measure the performance (within-item multidimensionality). The model also allows for individual covariates in its structural part.

Modeling conditional dependence among multiple diagnostic tests

When multiple imperfect dichotomous diagnostic tests are applied to an individual, it is possible that some or all of their results remain dependent even after conditioning on the true disease status. The estimates could be biased if this conditional dependence is ignored when using the test results to infer about the prevalence of a disease or the accuracies of the diagnostic tests. However, statistical methods correcting for this bias by modelling higher-order conditional dependence terms between multiple diagnostic tests are not well addressed in the literature. This paper extends a Bayesian fixed effects model for 2 diagnostic tests with pairwise correlation to cases with 3 or more diagnostic tests with higher order correlations. Simulation results show that the proposed fixed effects model works well both in the case when the tests are highly correlated and in the case when the tests are truly conditionally independent, provided adequate external information is available in the form of fixed constraints or prior distributions. A data set on the diagnosis of childhood pulmonary tuberculosis is used to illustrate the proposed model.

A method to compare the performance of two molecular diagnostic tools in the absence of a gold standard

The paper is motivated by the problem of comparing the accuracy of two molecular tests in detecting genetic mutations in tumor samples when there is no gold standard test. Commonly used sequencing methods require a large number of tumor cells in the tumor sample and the proportion of tumor cells with mutation positivity to be above a threshold level whereas new tests aim to reduce the requirement for number of tumor cells and the threshold level. A new latent class model is proposed to compare these two tests in which a random variable is used to represent the unobserved proportion of mutation positivity so that these two tests are conditionally dependent; furthermore, an independent random variable is included to address measurement error associated with the reading from each test, while existing latent class models often assume conditional independence and do not allow measurement error. In addition, methods for calculating the sample size for a study that is sufficiently powered to compare the accuracy of two molecular tests are proposed and compared. The proposed methods are then applied to a study which aims to compare two molecular tests for detecting EGFR mutations in lung cancer patients.

A Bayesian approach to model the conditional correlation between several diagnostic tests and various replicated subjects measurements

Two key aims of diagnostic research are to accurately and precisely estimate disease prevalence and test sensitivity and specificity. Latent class models have been proposed that consider the correlation between subject measures determined by different tests in order to diagnose diseases for which gold standard tests are not available. In some clinical studies, several measures of the same subject are made with the same test under the same conditions (replicated measurements), and thus, replicated measurements for each subject are not independent. In the present study, we propose an extension of the Bayesian latent class Gaussian random effects model to fit the data with binary outcomes for tests with replicated subject measures. We describe an application using data collected on hookworm infection carried out in the municipality of Presidente Figueiredo, Amazonas State, Brazil. In addition, the performance of the proposed model was compared with that of current models (the subject random effects model and the conditional (in)dependent model) through a simulation study. As expected, the proposed model presented better accuracy and precision in the estimations of prevalence, sensitivity and specificity. Copyright © 2017 John Wiley & Sons, Ltd.

Study protocol for a transversal study to develop a screening model for excessive gambling behaviours on a representative sample of users of French authorised gambling websites

INTRODUCTION

Since the legalisation of online gambling in France in 2010, gambling operators must implement responsible gambling measures to prevent excessive gambling practices. However, actually there is no screening procedure for identifying problematic gamblers. Although several studies have already been performed using several data sets from online gambling operators, the authors deplored several methodological and clinical limits that prevent scientifically validating the existence of problematic gambling behaviour. The aim of this study is to develop a model for screening excessive gambling practices based on the gambling behaviours observed on French gambling websites, coupled with a clinical validation.

METHODS AND ANALYSIS

The research is divided into three successive stages. All analyses will be performed for each major type of authorised online gambling in France. The first stage aims at defining a typology of users of French authorised gambling websites based on their gambling behaviour. This analysis will be based on data from the Authority for Regulating Online Gambling (ARJEL) and the Française Des Jeux (FDJ). For the second stage aiming at determining a score to predict whether a gambler is problematic or not, we will cross answers from the Canadian Problem Gambling Index with real gambling data. The objective of the third stage is to clinically validate the score previously developed. Results from the screening model will be compared (using sensitivity, specificity, area under the curve, and positive and negative predictive values) with the diagnosis obtained with a telephone clinical interview, including diagnostic criteria for gambling addiction.

ETHICS AND DISSEMINATION

This study was approved by the local Research Ethics Committee (GNEDS) on 25 March 2015. Results will be presented in national and international conferences, submitted to peer-reviewed journals and will be part of a PhD thesis. A final report with the study results will be presented to the ARJEL, especially the final screening model.

TRIAL REGISTRATION NUMBER

NCT02415296.

Impact of cause of death adjudication on the results of the European prostate cancer screening trial

BACKGROUND

The European Randomised Study of Prostate Cancer Screening has shown a 21% relative reduction in prostate cancer mortality at 13 years. The causes of death can be misattributed, particularly in elderly men with multiple comorbidities, and therefore accurate assessment of the underlying cause of death is crucial for valid results. To address potential unreliability of end-point assessment, and its possible impact on mortality results, we analysed the study outcome adjudication data in six countries.

METHODS

Latent class statistical models were formulated to compare the accuracy of individual adjudicators, and to assess whether accuracy differed between the trial arms. We used the model to assess whether correcting for adjudication inaccuracies might modify the study results.

RESULTS

There was some heterogeneity in adjudication accuracy of causes of death, but no consistent differential accuracy by trial arm. Correcting the estimated screening effect for misclassification did not alter the estimated mortality effect of screening.

CONCLUSIONS

Our findings were consistent with earlier reports on the European screening trial. Observer variation, while demonstrably present, is unlikely to have materially biased the main study results. A bias in assigning causes of death that might have explained the mortality reduction by screening can be effectively ruled out.

Therapy preferences of patients with lung and colon cancer: a discrete choice experiment

OBJECTIVES

There is increasing interest in studies that examine patient preferences to measure health-related outcomes. Understanding patients' preferences can improve the treatment process and is particularly relevant for oncology. In this study, we aimed to identify the subgroup-specific treatment preferences of German patients with lung cancer (LC) or colorectal cancer (CRC).

METHODS

Six discrete choice experiment (DCE) attributes were established on the basis of a systematic literature review and qualitative interviews. The DCE analyses comprised generalized linear mixed-effects model and latent class mixed logit model.

RESULTS

The study cohort comprised 310 patients (194 with LC, 108 with CRC, 8 with both types of cancer) with a median age of 63 (SD =10.66) years. The generalized linear mixed-effects model showed a significant (P<0.05) degree of association for all of the tested attributes. "Strongly increased life expectancy" was the attribute given the greatest weight by all patient groups. Using latent class mixed logit model analysis, we identified three classes of patients. Patients who were better informed tended to prefer a more balanced relationship between length and health-related quality of life (HRQoL) than those who were less informed. Class 2 (LC patients with low HRQoL who had undergone surgery) gave a very strong weighting to increased length of life. We deduced from Class 3 patients that those with a relatively good life expectancy (CRC compared with LC) gave a greater weight to moderate effects on HRQoL than to a longer life.

CONCLUSION

Overall survival was the most important attribute of therapy for patients with LC or CRC. Differences in treatment preferences between subgroups should be considered in regard to treatment and development of guidelines. Patients' preferences were not affected by sex or age, but were affected by the cancer type, HRQoL, surgery status, and the main source of information on the disease.

Automated linkage of patient records from disparate sources

We introduce an automated method of record linkage that has two key features, automated selection of match field interactions to include in the model for estimation and automated threshold determination for classifying record pairs to matches or non-matches. We applied our method to two real-world examples. The first example demonstrated results consistent with our earlier work: When data quality is adequate and the match field discriminating power is high, matching algorithms exhibit similar performance. The second example demonstrated that our method yields a lower false positive rate and higher positive predictive value than the Fellegi-Sunter model in the face of low data quality. When compared to the Fellegi-Sunter model, simulation studies suggest that our method exhibits better overall performance as indicated by higher area under the curve, and less biased estimates for both the match prevalence rate and the m- and u-probabilities over a range of data scenarios, especially when the match prevalence is extreme. Computationally, our method is as efficient as the Fellegi-Sunter model. We recommend this method in situations that an unsupervised linking algorithm is needed.

Understanding the effects of conditional dependence in research studies involving imperfect diagnostic tests

When two imperfect diagnostic tests are carried out on the same subject, their results may be correlated even after conditioning on the true disease status. While past work has focused on the consequences of ignoring conditional dependence, the degree to which conditional dependence can be induced has not been systematically studied. We examine this issue in detail by introducing a hypothetical missing covariate that affects the sensitivities of two imperfect dichotomous tests. We consider four forms for this covariate, normal, uniform, dichotomous and trichotomous. In the case of a dichotomous covariate, we derive an expression showing that the conditional covariance is a function of the product of the changes in test sensitivities (or specificities) between the subgroups defined by the covariate. The maximum possible covariance is induced by a dichotomous covariate with a very strong effect on both tests. Through simulations, we evaluate the extent to which fitting a latent class model ignoring each type of covariate but including a general covariance term can adjust for the correlation induced by the covariate. We compare the results to when the conditional dependence is ignored. We find that the bias because of ignoring conditional dependence is generally small even for moderate covariate effects, and when bias is present, a model including a covariance term works well. We illustrate our methods by analyzing data from a childhood tuberculosis study. Copyright © 2016 John Wiley & Sons, Ltd.

Estimating diagnostic accuracy without a gold standard: A continued controversy

There is a wide body of literature in biostatistics and epidemiology literature about estimating diagnostic accuracy, such as sensitivity and specificity of a binary test, without a gold standard. This methodology is very attractive since obtaining gold standard information is impossible, difficult, or very expensive in some situations. Although there are many proponents of these approaches, there have also been some serious criticisms. We review important methodological developments as well as discuss problems with the approaches. We propose alternative designs that may be less controversial and present ideas for future research. Lastly, we provide recommendations about how these methods should be used in practice.

Negative Treatment by Family as a Predictor of Depressive Symptoms, Life Satisfaction, Suicidality, and Tobacco/Alcohol Use in Vietnamese Sexual Minority Women

PURPOSE

Research linking family rejection and health outcomes in sexual minority people is mostly limited to North America. We assessed the associations between negative treatment by family members and depressive symptoms, life satisfaction, suicidality, and tobacco/alcohol use in sexual minority women (SMW) in Viet Nam.

METHODS

Data were from an anonymous internet survey (n = 1936). Latent class analysis characterized patterns of negative treatment by family members experienced by respondents. Latent class with distal outcome modeling was used to regress depressive symptoms, life satisfaction, suicidality, and tobacco/alcohol use on family treatment class, controlling for predictors of family treatment and for two other types of sexual prejudice.

RESULTS

Five latent family treatment classes were extracted, including four negative classes representing varying patterns of negative family treatment. Overall, more than one negative class predicted lower life satisfaction, more depressive symptoms, and higher odds of attempted suicide (relative to the non-negative class), supporting the minority stress hypothesis that negative family treatment is predictive of poorer outcomes. Only the most negative class had elevated alcohol use. The association between family treatment and smoking status was not statistically significant. The most negative class, unexpectedly, did not have the highest odds of having attempted suicide, raising a question about survivor bias.

CONCLUSION

This population requires public health attention, with emphasis placed on interventions targeting the family to promote acceptance and to prevent negative treatment, and interventions supporting those SMW who encounter the worst types of negative family treatment.

Binomial regression with a misclassified covariate and outcome

Misclassification occurring in either outcome variables or categorical covariates or both is a common issue in medical science. It leads to biased results and distorted disease-exposure relationships. Moreover, it is often of clinical interest to obtain the estimates of sensitivity and specificity of some diagnostic methods even when neither gold standard nor prior knowledge about the parameters exists. We present a novel Bayesian approach in binomial regression when both the outcome variable and one binary covariate are subject to misclassification. Extensive simulation results under various scenarios and a real clinical example are given to illustrate the proposed approach. This approach is motivated and applied to a dataset from the Baylor Alzheimer's Disease and Memory Disorders Center.

A joint latent class analysis for adjusting survival bias with application to a trauma transfusion study

There is no clear classification rule to rapidly identify trauma patients who are severely hemorrhaging and may need substantial blood transfusions. Massive transfusion (MT), defined as the transfusion of at least 10 units of red blood cells within 24 h of hospital admission, has served as a conventional surrogate that has been used to develop early predictive algorithms and establish criteria for ordering an MT protocol from the blood bank. However, the conventional MT rule is a poor proxy, because it is likely to misclassify many severely hemorrhaging trauma patients as they could die before receiving the 10th red blood cells transfusion. In this article, we propose to use a latent class model to obtain a more accurate and complete metric in the presence of early death. Our new approach incorporates baseline patient information from the time of hospital admission, by combining respective models for survival time and usage of blood products transfused within the framework of latent class analysis. To account for statistical challenges, caused by induced dependent censoring inherent in 24-h sums of transfusions, we propose to estimate an improved standard via a pseudo-likelihood function using an expectation-maximization algorithm with the inverse weighting principle. We evaluated the performance of our new standard in simulation studies and compared with the conventional MT definition using actual patient data from the Prospective Observational Multicenter Major Trauma Transfusion study. Copyright © 2015 John Wiley & Sons, Ltd.

On the Link between Cognitive Diagnostic Models and Knowledge Space Theory

The present work explores the connections between cognitive diagnostic models (CDM) and knowledge space theory (KST) and shows that these two quite distinct approaches overlap. It is proved that in fact the Multiple Strategy DINA (Deterministic Input Noisy AND-gate) model and the CBLIM, a competence-based extension of the basic local independence model (BLIM), are equivalent. To demonstrate the benefits that arise from integrating the two theoretical perspectives, it is shown that a fairly complete picture on the identifiability of these models emerges by combining results from both camps. The impact of the results is illustrated by an empirical example, and topics for further research are pointed out.

A general framework for comparative Bayesian meta-analysis of diagnostic studies

BACKGROUND

Selecting the most effective diagnostic method is essential for patient management and public health interventions. This requires evidence of the relative performance of alternative tests or diagnostic algorithms. Consequently, there is a need for diagnostic test accuracy meta-analyses allowing the comparison of the accuracy of two or more competing tests. The meta-analyses are however complicated by the paucity of studies that directly compare the performance of diagnostic tests. A second complication is that the diagnostic accuracy of the tests is usually determined through the comparison of the index test results with those of a reference standard. These reference standards are presumed to be perfect, i.e. allowing the classification of diseased and non-diseased subjects without error. In practice, this assumption is however rarely valid and most reference standards show false positive or false negative results. When an imperfect reference standard is used, the estimated accuracy of the tests of interest may be biased, as well as the comparisons between these tests.

METHODS

We propose a model that allows for the comparison of the accuracy of two diagnostic tests using direct (head-to-head) comparisons as well as indirect comparisons through a third test. In addition, the model allows and corrects for imperfect reference tests. The model is inspired by mixed-treatment comparison meta-analyses that have been developed for the meta-analysis of randomized controlled trials. As the model is estimated using Bayesian methods, it can incorporate prior knowledge on the diagnostic accuracy of the reference tests used.

RESULTS

We show the bias that can result from using inappropriate methods in the meta-analysis of diagnostic tests and how our method provides more correct estimates of the difference in diagnostic accuracy between two tests. As an illustration, we apply this model to a dataset on visceral leishmaniasis diagnostic tests, comparing the accuracy of the RK39 dipstick with that of the direct agglutination test.

CONCLUSIONS

Our proposed meta-analytic model can improve the comparison of the diagnostic accuracy of competing tests in a systematic review. This is however only true if the studies and especially information on the reference tests used are sufficiently detailed. More specifically, the type and exact procedures used as reference tests are needed, including any cut-offs used and the number of subjects excluded from full reference test assessment. If this information is lacking, it may be better to limit the meta-analysis to direct comparisons.

Estimation of diagnostic test accuracy without full verification: a review of latent class methods

The performance of a diagnostic test is best evaluated against a reference test that is without error. For many diseases, this is not possible, and an imperfect reference test must be used. However, diagnostic accuracy estimates may be biased if inaccurately verified status is used as the truth. Statistical models have been developed to handle this situation by treating disease as a latent variable. In this paper, we conduct a systematized review of statistical methods using latent class models for estimating test accuracy and disease prevalence in the absence of complete verification.

Joint model for a diagnostic test without a gold standard in the presence of a dependent terminal event

Breast cancer patients after breast conservation therapy often develop ipsilateral breast tumor relapse (IBTR), whose classification (true local recurrence versus new ipsilateral primary tumor) is subject to error, and there is no available gold standard. Some patients may die because of breast cancer before IBTR develops. Because this terminal event may be related to the individual patient's unobserved disease status and time to IBTR, the terminal mechanism is non-ignorable. This article presents a joint analysis framework to model the binomial regression with misclassified binary outcome and the correlated time to IBTR, subject to a dependent terminal event and in the absence of a gold standard. Shared random effects are used to link together two survival times. The proposed approach is evaluated by a simulation study and is applied to a breast cancer data set consisting of 4477 breast cancer patients. The proposed joint model can be conveniently fit using adaptive Gaussian quadrature tools implemented in SAS 9.3 (SAS Institute Inc., Cary, NC, USA) procedure NLMIXED.

A fast Monte Carlo EM algorithm for estimation in latent class model analysis with an application to assess diagnostic accuracy for cervical neoplasia in women with AGC

In this article we use a latent class model (LCM) with prevalence modeled as a function of covariates to assess diagnostic test accuracy in situations where the true disease status is not observed, but observations on three or more conditionally independent diagnostic tests are available. A fast Monte Carlo EM (MCEM) algorithm with binary (disease) diagnostic data is implemented to estimate parameters of interest; namely, sensitivity, specificity, and prevalence of the disease as a function of covariates. To obtain standard errors for confidence interval construction of estimated parameters, the missing information principle is applied to adjust information matrix estimates. We compare the adjusted information matrix based standard error estimates with the bootstrap standard error estimates both obtained using the fast MCEM algorithm through an extensive Monte Carlo study. Simulation demonstrates that the adjusted information matrix approach estimates the standard error similarly with the bootstrap methods under certain scenarios. The bootstrap percentile intervals have satisfactory coverage probabilities. We then apply the LCM analysis to a real data set of 122 subjects from a Gynecologic Oncology Group (GOG) study of significant cervical lesion (S-CL) diagnosis in women with atypical glandular cells of undetermined significance (AGC) to compare the diagnostic accuracy of a histology-based evaluation, a CA-IX biomarker-based test and a human papillomavirus (HPV) DNA test.

The effect of job stress on smoking and alcohol consumption

This paper examines the effect of job stress on two key health risk-behaviors: smoking and alcohol consumption, using data from the Canadian National Population Health Survey. Findings in the extant literature are inconclusive and are mainly based on standard models which can model differential responses to job stress only by observed characteristics. However, the effect of job stress on smoking and drinking may largely depend on unobserved characteristics such as: self control, stress-coping ability, personality traits and health preferences. Accordingly, we use a latent class model to capture heterogeneous responses to job stress. Our results suggest that the effects of job stress on smoking and alcohol consumption differ substantially for at least two "types" of individuals, light and heavy users. In particular, we find that job stress has a positive and statistically significant impact on smoking intensity, but only for light smokers, while it has a positive and significant impact on alcohol consumption mainly for heavy drinkers. These results provide suggestive evidence that the mixed findings in previous studies may partly be due to unobserved individual heterogeneity which is not captured by standard models.

Latent class joint model of ovarian function suppression and DFS for premenopausal breast cancer patients

Breast cancer is the leading cancer in women of reproductive age; more than a quarter of women diagnosed with breast cancer in the US are premenopausal. A common adjuvant treatment for this patient population is chemotherapy, which has been shown to cause premature menopause and infertility with serious consequences to quality of life. Luteinizing-hormone-releasing hormone (LHRH) agonists, which induce temporary ovarian function suppression (OFS), has been shown to be a useful alternative to chemotherapy in the adjuvant setting for estrogen-receptor-positive breast cancer patients. LHRH agonists have the potential to preserve fertility after treatment, thus, reducing the negative effects on a patient's reproductive health. However, little is known about the association between a patient's underlying degree of OFS and disease-free survival (DFS) after receiving LHRH agonists. Specifically, we are interested in whether patients with lower underlying degrees of OFS (i.e. higher estrogen production) after taking LHRH agonists are at a higher risk for late breast cancer events. In this paper, we propose a latent class joint model (LCJM) to analyze a data set from International Breast Cancer Study Group (IBCSG) Trial VIII to investigate the association between OFS and DFS. Analysis of this data set is challenging due to the fact that the main outcome of interest, OFS, is unobservable and the available surrogates for this latent variable involve masked event and cured proportions. We employ a likelihood approach and the EM algorithm to obtain parameter estimates and present results from the IBCSG data analysis.

New insights into the efficacy of naltrexone based on trajectory-based reanalyses of two negative clinical trials

BACKGROUND

The heterogeneity of clinical findings in studies evaluating the efficacy of naltrexone in the treatment of alcohol dependence has led to growing efforts to explore novel approaches to data analysis. The objective of this study was to identify distinct trajectories of daily drinking over time in two negative clinical trials and to determine whether naltrexone affected the probability to follow a particular trajectory.

METHODS

The Veterans Affairs (VA) Cooperative Study #425 and the Women's Naltrexone Study failed to demonstrate efficacy on primary outcome variables. Separately for each study, we analyzed daily indicators of any drinking and heavy drinking using a semiparametric group-based approach.

RESULTS

We estimated three distinct trajectories of daily drinking (both any and heavy drinking) which we described as "abstainer," "sporadic drinker," and "consistent drinker." Naltrexone doubled the odds of following the abstainer trajectory instead of the consistent drinker trajectory but did not significantly change the odds of following the abstainer trajectory as contrasted with the sporadic drinker trajectory.

CONCLUSIONS

Naltrexone may have a clinically meaningful effect for alcohol-dependent patients with a high chance of consistent drinking, even in studies where it failed to show efficacy in planned analyses.