Perceived discrimination, trust in physicians, and prolonged symptom duration before ovarian cancer diagnosis in the African American Cancer Epidemiology Study

BACKGROUND

Discrimination and trust are known barriers to accessing health care. Despite well-documented racial disparities in the ovarian cancer care continuum, the role of these barriers has not been examined. This study evaluated the association of everyday discrimination and trust in physicians with a prolonged interval between symptom onset and ovarian cancer diagnosis (hereafter referred to as prolonged symptom duration).

METHODS

Subjects included cases enrolled in the African American Cancer Epidemiology Study, a multisite case-control study of epithelial ovarian cancer among black women. Logistic regression was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for associations of everyday discrimination and trust in physicians with a prolonged symptom duration (1 or more symptoms lasting longer than the median symptom-specific duration), and it controlled for access-to-care covariates and potential confounders.

RESULTS

Among the 486 cases in this analysis, 302 women had prolonged symptom duration. In the fully adjusted model, a 1-unit increase in the frequency of everyday discrimination increased the odds of prolonged symptom duration 74% (OR, 1.74; 95% CI, 1.22-2.49), but trust in physicians was not associated with prolonged symptom duration (OR, 0.86; 95% CI, 0.66-1.11).

CONCLUSIONS

Perceived everyday discrimination was associated with prolonged symptom duration, whereas more commonly evaluated determinants of access to care and trust in physicians were not. These results suggest that more research on the effects of interpersonal barriers affecting ovarian cancer care is warranted.

Analysis of racial differences in hospital stays in the presence of geographic confounding

Using recent methods for spatial propensity score modeling, we examine differences in hospital stays between non-Hispanic black and non-Hispanic white veterans with type 2 diabetes. We augment a traditional patient-level propensity score model with a spatial random effect to create a matched sample based on the estimated propensity score. We then use a spatial negative binomial hurdle model to estimate differences in both hospital admissions and inpatient days. We demonstrate that in the presence of unmeasured geographic confounding, spatial propensity score matching in addition to the spatial negative binomial hurdle outcome model yields improved performance compared to the outcome model alone. In the motivating application, we construct three estimates of racial differences in hospitalizations: the risk difference in admission, the mean difference in number of inpatient days among those hospitalized, and the mean difference in number of inpatient days across all patients (hospitalized and non-hospitalized). Results indicate that non-Hispanic black veterans with type 2 diabetes have a lower risk of hospital admission and a greater number of inpatient days on average. The latter result is especially important considering that we observed much smaller effect sizes in analyses that did not incorporate spatial matching. These results emphasize the need to address geographic confounding in health disparity studies.

Interval Estimation for Age-Adjusted Rate Ratios Using Bayesian Convolution Model

Spatial correlation raises challenges in estimating confidence intervals for region specific event rates and rate ratios between geographic units that are nested. Methods have been proposed to incorporate spatial correlation by assuming various distributions for the structure of autocorrelation patterns. However, the derivation of these statistics based on approximation may have to condition on the distributional assumption underlying the data generating process, which may not hold for certain situations. This paper explores the feasibility of utilizing a Bayesian convolution model (BCM), which includes an uncorrelated heterogeneity (UH) and a conditional autoregression (CAR) component to accommodate both uncorrelated and correlated spatial heterogeneity, to estimate the 95% confidence intervals for age-adjusted rate ratios among geographic regions with existing spatial correlations. A simulation study is conducted and a BCM method is applied to two cancer incidence datasets to calculate age-adjusted rate/ratio for the counties in the State of Kentucky relative to the entire state. In comparison to three existing methods, without and with spatial correlation, the Bayesian convolution model-based estimation provides moderate shrinkage effect for the point estimates based on the neighbor structure across regions and produces a wider interval due to the inclusion of uncertainty in the spatial autocorrelation parameters. The overall spatial pattern of region incidence rate from BCM approach appears to be like the direct estimates and other methods for both datasets, even though "smoothing" occurs in some local regions. The Bayesian Convolution Model allows flexibility in the specification of risk components and can improve the accuracy of interval estimates of age-adjusted rate ratios among geographical regions as it considers spatial correlation.

A data-driven approach for estimating the change-points and impact of major events on disease risk

Considering the impact of events on disease risk is important. Here, a Bayesian spatio-temporal accelerated failure time model furnished an ideal situation for modeling events that could impact survival experience via spatial and temporal frailty estimates. Through a hierarchical structure, this model allowed the data to detect the change-point(s) in addition to generating the event-related estimates. Both a real data case study and a simulation study were employed for testing these methods. The results suggested that meaningful and accurate change-points could be detected. Further, accurate event-related estimates for individuals in relation to those change-points could be obtained. By allowing the data to drive the change-point choices, the models were better fitting and the inference was more accurate.

Temporal Trends in Cardiovascular Hospital Discharges Following a Mass Chlorine Exposure Event in Graniteville, South Carolina

Background: On January 6, 2005, a train derailed in Graniteville, South Carolina, releasing nearly 60,000 kg of toxic chlorine gas. The disaster left nine people dead and was responsible for hundreds of hospitalizations and outpatient visits in the subsequent weeks. While chlorine gas primarily affects the respiratory tract, a growing body of evidence suggests that acute exposure may also cause vascular injury and cardiac toxicity. Here, we describe the incidence of cardiovascular hospitalizations among residents of the zip codes most affected by the chlorine gas plume, and compare the incidence of cardiovascular discharges in the years leading up to the event (2000–2004) to the incidence in the years following the event (2005–2014).

Methods: De-identified hospital discharge information was collected from the South Carolina Revenue and Fiscal Affairs Office for individuals residing in the selected zip codes for the years 2000 to 2014. A quasi-experimental study design was utilized with a population-level interrupted time series model to examine hospital discharge rates for Graniteville-area residents for three cardiovascular diagnoses: hypertension (HTN), acute myocardial infarction (AMI), and coronary heart disease (CHD). We used linear regression with autoregressive error correction to compare slopes for pre- and post-spill time periods. Data from the 2000 and 2010 censuses were used to calculate rates and to provide information on potential demographic shifts over the course of the study.

Results: A significant increase in hypertension-related hospital discharge rates was observed for the years following the Graniteville chlorine spill (slope 8.2, p < 0.001). Concurrent changes to CHD and AMI hospital discharge rates were in the opposite direction (slopes −3.2 and −0.3, p < 0.01 and 0.14, respectively). Importantly, the observed trend cannot be attributed to an aging population.

Conclusions: An unusual increase in hypertension-related hospital discharge rates in the area affected by the Graniteville chlorine spill contrasts with national and state-level trends. A number of factors related to the spill may be contributing the observation: disaster-induced hypertension, healthcare services access issues, and, possibly, chlorine-induced susceptibility to vascular pathologies. Due to the limitations of our data, we cannot determine whether the individuals who visited the hospital were the ones exposed to chlorine gas, however, the finding warrants additional research. Future studies are needed to determine the etiology of the increase and whether individuals exposed to chlorine are at a heightened risk for hypertensive heart disease.

Bayesian cure-rate survival model with spatially structured censoring

We propose a Bayesian spatial model for time-to-event data in which we allow the censoring mechanism to depend on covariates and have a spatial structure. The survival model incorporates a cure rate fraction and assumes that the time-to-event follows a Weibull distribution, with covariates such as race, stage, grade, marital status and age at diagnosis being linked to its scale parameter. With right censoring being a primary concern, we consider a joint logistic regression model for the death versus censoring indicator, allowing dependence on covariates and including a spatial structure via the use of random effects. We apply the models to examine prostate cancer data from the Surveillance, Epidemiology, and End Results (SEER) registry, which displays marked spatial variation.

A Bayesian Quantile Modeling for Spatiotemporal Relative Risk: An Application to Adverse Risk Detection of Respiratory Diseases in South Carolina, USA

Quantile modeling has been seen as an alternative and useful complement to ordinary regression mainly focusing on the mean. To directly apply quantile modeling to areal data the discrete conditional quantile function of the data can be an issue. Although jittering by adding a small number from a uniform distribution to impose pseudo-continuity has been proposed, the approach can have a great influence on responses with small values. Thus we proposed an alternative to model the quantiles of relative risk for spatiotemporal areal health data within a Bayesian framework using the log-Laplace distribution. A simulation study was conducted to assess the performance of the proposed method and examine whether the model could robustly estimate quantiles of spatiotemporal count data. To perform a test with a real data example, we evaluated the potential application of clustering under the proposed log-Laplace and mean regression. The data were obtained from the total number of emergency room discharges for respiratory conditions, both infectious and non-infectious diseases, in the U.S. state of South Carolina in 2009. From both simulation and case studies, the proposed quantile modeling demonstrated potential for broad applicability in various areas of spatial health studies including anomaly detection.

Multiscale measurement error models for aggregated small area health data

Spatial data are often aggregated from a finer (smaller) to a coarser (larger) geographical level. The process of data aggregation induces a scaling effect which smoothes the variation in the data. To address the scaling problem, multiscale models that link the convolution models at different scale levels via the shared random effect have been proposed. One of the main goals in aggregated health data is to investigate the relationship between predictors and an outcome at different geographical levels. In this paper, we extend multiscale models to examine whether a predictor effect at a finer level hold true at a coarser level. To adjust for predictor uncertainty due to aggregation, we applied measurement error models in the framework of multiscale approach. To assess the benefit of using multiscale measurement error models, we compare the performance of multiscale models with and without measurement error in both real and simulated data. We found that ignoring the measurement error in multiscale models underestimates the regression coefficient, while it overestimates the variance of the spatially structured random effect. On the other hand, accounting for the measurement error in multiscale models provides a better model fit and unbiased parameter estimates.

A Bayesian two-stage spatially dependent variable selection model for space-time health data

In space-time epidemiological modeling, most studies have considered the overall variations in relative risk to better estimate the effects of risk factors on health outcomes. However, the associations between risk factors and health outcomes may vary across space and time. Especially, the temporal patterns of the covariate effects may depend on space. Thus, we propose a Bayesian two-stage spatially dependent variable selection approach for space-time health data to determine the spatially varying subsets of regression coefficients with common temporal dependence. The two-stage structure allows reduction of the spatial confounding bias in the estimates of the regression coefficients. A simulation study is conducted to examine the performance of the proposed two-stage model. We apply the proposed model to the number of inpatients with lung cancer in 159 counties of Georgia, USA.

Zero-inflated multiscale models for aggregated small area health data

It is our primary focus to study the spatial distribution of disease incidence at different geographical levels. Often, spatial data are available in the form of aggregation at multiple scale levels such as census tract, county, state, and so on. When data are aggregated from a fine (e.g. county) to a coarse (e.g. state) geographical level, there will be loss of information. The problem is more challenging when excessive zeros are available at the fine level. After data aggregation, the excessive zeros at the fine level will be reduced at the coarse level. If we ignore the zero inflation and the aggregation effect, we could get inconsistent risk estimates at the fine and coarse levels. Hence, in this paper, we address those problems using zero inflated multiscale models that jointly describe the risk variations at different geographical levels. For the excessive zeros at the fine level, we use a zero inflated convolution model, whereas we consider a regular convolution model for the smoothed data at the coarse level. These methods provide a consistent risk estimate at the fine and coarse levels when high percentages of structural zeros are present in the data.

Spatiotemporal multivariate mixture models for Bayesian model selection in disease mapping

It is often the case that researchers wish to simultaneously explore the behavior of and estimate overall risk for multiple, related diseases with varying rarity while accounting for potential spatial and/or temporal correlation. In this paper, we propose a flexible class of multivariate spatio-temporal mixture models to fill this role. Further, these models offer flexibility with the potential for model selection as well as the ability to accommodate lifestyle, socio-economic, and physical environmental variables with spatial, temporal, or both structures. Here, we explore the capability of this approach via a large scale simulation study and examine a motivating data example involving three cancers in South Carolina. The results which are focused on four model variants suggest that all models possess the ability to recover simulation ground truth and display improved model fit over two baseline Knorr-Held spatio-temporal interaction model variants in a real data application.

Addressing geographic confounding through spatial propensity scores: a study of racial disparities in diabetes

Motivated by a study exploring differences in glycemic control between non-Hispanic black and non-Hispanic white veterans with type 2 diabetes, we aim to address a type of confounding that arises in spatially referenced observational studies. Specifically, we develop a spatial doubly robust propensity score estimator to reduce bias associated with geographic confounding, which occurs when measured or unmeasured confounding factors vary by geographic location, leading to imbalanced group comparisons. We augment the doubly robust estimator with spatial random effects, which are assigned conditionally autoregressive priors to improve inferences by borrowing information across neighboring geographic regions. Through a series of simulations, we show that ignoring spatial variation results in increased absolute bias and mean squared error, while the spatial doubly robust estimator performs well under various levels of spatial heterogeneity and moderate sample sizes. In the motivating application, we construct three global estimates of the risk difference between race groups: an unadjusted estimate, a doubly robust estimate that adjusts only for patient-level information, and a hierarchical spatial doubly robust estimate. Results indicate a gradual reduction in the risk difference at each stage, with the inclusion of spatial random effects providing a 20% reduction compared to an estimate that ignores spatial heterogeneity. Smoothed maps indicate poor glycemic control across Alabama and southern Georgia, areas comprising the so-called "stroke belt." These results suggest the need for community-specific interventions to target diabetes in geographic areas of greatest need.

Assessment of spatial variation in breast cancer-specific mortality using Louisiana SEER data

BACKGROUND

Previous studies suggest spatial differences in mortality for many types of cancer, including breast cancer. Identifying explanations for these spatial differences results in a better understanding of what leads to longer survival time.

METHODS

We used a Bayesian accelerated failure time model with spatial frailty terms to investigate potential spatial differences in breast cancer mortality following breast cancer diagnosis using 2000-2013 Louisiana SEER data.

RESULTS

There are meaningful spatial differences in breast cancer mortality across the parishes of Louisiana, even after adjusting for known demographic and clinical risk factors. For example, the average survival time of a woman diagnosed in Orleans parish was 1.51 times longer than that of a woman diagnosed in Terrebonne parish. Additionally, there is evidence to suggest shorter survival times in lower income parishes along the Red and Mississippi Rivers, as well as parishes with lower socioeconomic status, less access to care and fresh food, worse quality of care, and more workers in certain industries.

CONCLUSION

The addition of spatial frailties to account for an individual's geographic location is useful when analyzing breast cancer mortality data. Our findings suggest that survival following breast cancer diagnosis could potentially be improved if socioeconomic status differences were addressed, healthcare improved in quality and became more accessible, and certain industrial situations were improved for individuals diagnosed in parishes identified as having shorter average survival times.

Comparing multilevel and multiscale convolution models for small area aggregated health data

In spatial epidemiology, data are often arrayed hierarchically. The classification of individuals into smaller units, which in turn are grouped into larger units, can induce contextual effects. On the other hand, a scaling effect can occur due to the aggregation of data from smaller units into larger units. In this paper, we propose a shared multilevel model to address the contextual effects. In addition, we consider a shared multiscale model to adjust for both scale and contextual effects simultaneously. We also study convolution and independent multiscale models, which are special cases of shared multilevel and shared multiscale models, respectively. We compare the performance of the models by applying them to real and simulated data sets. We found that the shared multiscale model was the best model across a range of simulated and real scenarios as measured by the deviance information criterion (DIC) and the Watanabe Akaike information criterion (WAIC).

Space-time variation of respiratory cancers in South Carolina: a flexible multivariate mixture modeling approach to risk estimation

PURPOSE

Many types of cancer have an underlying spatiotemporal distribution. Spatiotemporal mixture modeling can offer a flexible approach to risk estimation via the inclusion of latent variables.

METHODS

In this article, we examine the application and benefits of using four different spatiotemporal mixture modeling methods in the modeling of cancer of the lung and bronchus as well as "other" respiratory cancer incidences in the state of South Carolina.

RESULTS

Of the methods tested, no single method outperforms the other methods; which method is best depends on the cancer under consideration. The lung and bronchus cancer incidence outcome is best described by the univariate modeling formulation, whereas the "other" respiratory cancer incidence outcome is best described by the multivariate modeling formulation.

CONCLUSIONS

Spatiotemporal multivariate mixture methods can aid in the modeling of cancers with small and sparse incidences when including information from a related, more common type of cancer.

Spatially-dependent Bayesian model selection for disease mapping

In disease mapping where predictor effects are to be modeled, it is often the case that sets of predictors are fixed, and the aim is to choose between fixed model sets. Model selection methods, both Bayesian model selection and Bayesian model averaging, are approaches within the Bayesian paradigm for achieving this aim. In the spatial context, model selection could have a spatial component in the sense that some models may be more appropriate for certain areas of a study region than others. In this work, we examine the use of spatially referenced Bayesian model averaging and Bayesian model selection via a large-scale simulation study accompanied by a small-scale case study. Our results suggest that BMS performs well when a strong regression signature is found.

Bayesian latent structure modeling of walking behavior in a physical activity intervention

The analysis of walking behavior in a physical activity intervention is considered. A Bayesian latent structure modeling approach is proposed whereby the ability and willingness of participants is modeled via latent effects. The dropout process is jointly modeled via a linked survival model. Computational issues are addressed via posterior sampling and a simulated evaluation of the longitudinal model's ability to recover latent structure and predictor effects is considered. We evaluate the effect of a variety of socio-psychological and spatial neighborhood predictors on the propensity to walk and the estimation of latent ability and willingness in the full study.

Mind the Scales: Harnessing Spatial Big Data for Infectious Disease Surveillance and Inference

Spatial big data have the velocity, volume, and variety of big data sources and contain additional geographic information. Digital data sources, such as medical claims, mobile phone call data records, and geographically tagged tweets, have entered infectious diseases epidemiology as novel sources of data to complement traditional infectious disease surveillance. In this work, we provide examples of how spatial big data have been used thus far in epidemiological analyses and describe opportunities for these sources to improve disease-mitigation strategies and public health coordination. In addition, we consider the technical, practical, and ethical challenges with the use of spatial big data in infectious disease surveillance and inference. Finally, we discuss the implications of the rising use of spatial big data in epidemiology to health risk communication, and public health policy recommendations and coordination across scales.

Spatio-temporal Bayesian model selection for disease mapping

Spatio-temporal analysis of small area health data often involves choosing a fixed set of predictors prior to the final model fit. In this paper, we propose a spatio-temporal approach of Bayesian model selection to implement model selection for certain areas of the study region as well as certain years in the study time line. Here, we examine the usefulness of this approach by way of a large-scale simulation study accompanied by a case study. Our results suggest that a special case of the model selection methods, a mixture model allowing a weight parameter to indicate if the appropriate linear predictor is spatial, spatio-temporal, or a mixture of the two, offers the best option to fitting these spatio-temporal models. In addition, the case study illustrates the effectiveness of this mixture model within the model selection setting by easily accommodating lifestyle, socio-economic, and physical environmental variables to select a predominantly spatio-temporal linear predictor.