A multivariate frailty model for disease recurrences and survival

A multivariate frailty hazard model is developed for joint-modeling of three correlated time-to-event outcomes: (1) local recurrence, (2) distant recurrence, and (3) overall survival. The term frailty is introduced to model population heterogeneity. The dependence is modeled by conditioning on a shared frailty that is included in the three hazard functions. Independent variables can be included in the model as covariates. The Markov chain Monte Carlo methods are used to estimate the posterior distributions of model parameters. The algorithm used in present application is the hybrid Metropolis-Hastings algorithm, which simultaneously updates all parameters with evaluations of gradient of log posterior density. The performance of this approach is examined based on simulation studies using Exponential and Weibull distributions. We apply the proposed methods to a study of patients with soft tissue sarcoma, which motivated this research. Our results indicate that patients with chemotherapy had better overall survival with hazard ratio of 0.242 (95% CI: 0.094 - 0.564) and lower risk of distant recurrence with hazard ratio of 0.636 (95% CI: 0.487 - 0.860), but not significantly better in local recurrence with hazard ratio of 0.799 (95% CI: 0.575 - 1.054). The advantages and limitations of the proposed models, and future research directions are discussed. ^

Stages of change and carotene intake in the Paso del Norte region: Que Sabrosa Vida

Understanding a population's dietary behavior is important to promote behaviors which have the most beneficial impact on health. The most recent Dietary Guidelines for Americans (2005) identifies carotenoids as a key nutrient to be consumed through increased intake of fruits and vegetables (FV). While some studies have included or focused on the Hispanic population, few have focused only on Mexican-American populations and staged its intake of FV. Stage of change behavior theory has been used to understand the adoption and promotion of healthy behaviors such as increased intake of FV. It has been shown to effectively aid interventionists' understanding of dietary behavior. Intake patterns of FV of older women, rural residents, and adolescents of Mexican American descent have been conducted but not by stages of change. This study aimed to determine the relationship between stages of change for fruits and vegetables (SOC-FV) and total carotene intake to assess the quality of SOC-FV as a surrogate measure of total carotene. ^ Data from the 2000 Qué Sabrosa Vida Community Nutrition Survey (QSV-CNS) were analyzed to identify the SOC-FV and sources of carotenes in a Mexican American population 18-60 yrs. of the Paso del Norte region. A 107 item interviewer administered food frequency questionnaire (FFQ) specifically calibrated for a Mexican American population was used to collect usual intake of total carotene. The QSV survey study population included 963 participants, 590 (61.3%) women and 373 (38.7%) men. A statistically significant mean difference in caloric intake between men and women was found (p-value = <0.01). When total carotene intake was adjusted for energy, there were significant differences between men and women (p-value = <0.0001) with women consuming a higher amount of total carotene (406 RE/kcal 1,000) than men (332 RE/kcal 1000). The food sources of total carotene for both genders included many items found in a traditional Mexican American diet. Chile, after carrots, was the highest contributor of dietary carotene. Total carotene intake was not associated with stages of change among women or men and their distributions were not linear. Mean differences of total carotene by stages of change were significant for women for pre-contemplation/contemplation (p-value = 0.04) and preparation (p-value = 0.0004) but not for men. ^ SOC-FV may serve as a surrogate measure for dietary carotene intake. This study's Mexican American population had a high carotene quality diet derived from traditional food items irrespective of their stage of change for fruits and vegetables. To better understand this population's dietary intake a measure for acculturation should be included. Interventions aimed at Mexican American populations should aim to promote traditional diets consistent with cultural practices.^ ^

When Policy Opportunity is not Enough: College Access and Enrollment Patterns among Texas Percent Plan Eligible Students

In 1998, Texas initiated a bold new statewide university admission policy aimed at increasing college access for traditionally underserved students in the state. House Bill 588 (known as the Texas Top 10 Percent Plan (TTPP)) guaranteed automatic admission to the college or university of their choice for all top performing students in Texas public high schools. Fourteen years after the plan’s implementation, we see great strides and complexities in understanding student outcomes as a result of the percent plan. However, the legal controversy over the percent plan both in Texas and other states incorporating similar yet distinctly motivated alternative admissions plans continues to play out from institutional decision boards to the highest court in the nation. This study seeks to add to that discussion by exploring two questions. Descriptively, what are the admission and enrollment patterns within racial/ethnic groups of percent plan eligible students, over time, for Texas elite, emergent elite, and remaining public institutions? Given that all eligible percent plan students may enter the institution of choice in Texas, does which type of institution a TTPP student chooses relate to their race/ethnicity? The descriptive story told by the admission and enrollment distributions of equally eligible TTPP students is a complex but compelling one. Fundamentally, it identifies that statistically different application and enrollment patterns exist for Hispanic and especially African American TTPP beneficiaries relative to their White and Asian American counterparts.

Bayesian generalized linear models for meta-analysis of diagnostic tests

With the recognition of the importance of evidence-based medicine, there is an emerging need for methods to systematically synthesize available data. Specifically, methods to provide accurate estimates of test characteristics for diagnostic tests are needed to help physicians make better clinical decisions. To provide more flexible approaches for meta-analysis of diagnostic tests, we developed three Bayesian generalized linear models. Two of these models, a bivariate normal and a binomial model, analyzed pairs of sensitivity and specificity values while incorporating the correlation between these two outcome variables. Noninformative independent uniform priors were used for the variance of sensitivity, specificity and correlation. We also applied an inverse Wishart prior to check the sensitivity of the results. The third model was a multinomial model where the test results were modeled as multinomial random variables. All three models can include specific imaging techniques as covariates in order to compare performance. Vague normal priors were assigned to the coefficients of the covariates. The computations were carried out using the 'Bayesian inference using Gibbs sampling' implementation of Markov chain Monte Carlo techniques. We investigated the properties of the three proposed models through extensive simulation studies. We also applied these models to a previously published meta-analysis dataset on cervical cancer as well as to an unpublished melanoma dataset. In general, our findings show that the point estimates of sensitivity and specificity were consistent among Bayesian and frequentist bivariate normal and binomial models. However, in the simulation studies, the estimates of the correlation coefficient from Bayesian bivariate models are not as good as those obtained from frequentist estimation regardless of which prior distribution was used for the covariance matrix. The Bayesian multinomial model consistently underestimated the sensitivity and specificity regardless of the sample size and correlation coefficient. In conclusion, the Bayesian bivariate binomial model provides the most flexible framework for future applications because of its following strengths: (1) it facilitates direct comparison between different tests; (2) it captures the variability in both sensitivity and specificity simultaneously as well as the intercorrelation between the two; and (3) it can be directly applied to sparse data without ad hoc correction. ^