Effect of primary care availability and health insurance coverage on risk and cost of Preventable Hospitalizations in Harris County

Preventable Hospitalizations (PHs) are hospitalizations that can be avoided with appropriate and timely care in the ambulatory setting and hence are closely associated with primary care access in a community. Increased primary care availability and health insurance coverage may increase primary care access, and consequently may be significantly associated with risks and costs of PHs. Objective. To estimate the risk and cost of preventable hospitalizations (PHs); to determine the association of primary care availability and health insurance coverage with the risk and costs of PHs, first alone and then simultaneously; and finally, to estimate the impact of expansions in primary care availability and health insurance coverage on the burden of PHs among non-elderly adult residents of Harris County. Methods. The study population was residents of Harris County, age 18 to 64, who had at least one hospital discharge in a Texas hospital in 2008. The primary independent variables were availability of primary care physicians, availability of primary care safety net clinics and health insurance coverage. The primary dependent variables were PHs and associated hospitalization costs. The Texas Health Care Information Collection (THCIC) Inpatient Discharge data was used to obtain information on the number and costs of PHs in the study population. Risk of PHs in the study population, as well as average and total costs of PHs were calculated. Multivariable logistic regression models and two-step Heckman regression models with log-transformed costs were used to determine the association of primary care availability and health insurance coverage with the risk and costs of PHs respectively, while controlling for individual predisposing, enabling and need characteristics. Predicted PH risk and cost were used to calculate the predicted burden of PHs in the study population and the impact of expansions in primary care availability and health insurance coverage on the predicted burden. Results. In 2008, hospitalized non-elderly adults in Harris County had 11,313 PHs and a corresponding PH risk of 8.02%. Congestive heart failure was the most common PH. PHs imposed a total economic burden of $84 billion at an average of $7,449 per PH. Higher primary care safety net availability was significantly associated with the lower risk of PHs in the final risk model, but only in the uninsured. A unit increase in safety net availability led to a 23% decline in PH odds in the uninsured, compared to only a 4% decline in the insured. Higher primary care physician availability was associated with increased PH costs in the final cost model (β=0.0020; p<0.05). Lack of health insurance coverage increased the risk of PH, with the uninsured having 30% higher odds of PHs (OR=1.299; p<0.05), but reduced the cost of a PH by 7% (β=-0.0668; p<0.05). Expansions in primary care availability and health insurance coverage were associated with a reduction of about $1.6 million in PH burden at the highest level of expansion. Conclusions. Availability of primary care resources and health insurance coverage in hospitalized non-elderly adults in Harris County are significantly associated with the risk and costs of PHs. Expansions in these primary care access factors can be expected to produce significant reductions in the burden of PHs in Harris County.^

TIME SERIES ANALYSIS AS INPUT FOR PREDICTIVE MODELING: PREDICTING CARDIAC ARREST IN A PEDIATRIC INTENSIVE CARE UNIT

The first manuscript, entitled "Time-Series Analysis as Input for Clinical Predictive Modeling: Modeling Cardiac Arrest in a Pediatric ICU" lays out the theoretical background for the project. There are several core concepts presented in this paper. First, traditional multivariate models (where each variable is represented by only one value) provide single point-in-time snapshots of patient status: they are incapable of characterizing deterioration. Since deterioration is consistently identified as a precursor to cardiac arrests, we maintain that the traditional multivariate paradigm is insufficient for predicting arrests. We identify time series analysis as a method capable of characterizing deterioration in an objective, mathematical fashion, and describe how to build a general foundation for predictive modeling using time series analysis results as latent variables. Building a solid foundation for any given modeling task involves addressing a number of issues during the design phase. These include selecting the proper candidate features on which to base the model, and selecting the most appropriate tool to measure them. We also identified several unique design issues that are introduced when time series data elements are added to the set of candidate features. One such issue is in defining the duration and resolution of time series elements required to sufficiently characterize the time series phenomena being considered as candidate features for the predictive model. Once the duration and resolution are established, there must also be explicit mathematical or statistical operations that produce the time series analysis result to be used as a latent candidate feature. In synthesizing the comprehensive framework for building a predictive model based on time series data elements, we identified at least four classes of data that can be used in the model design. The first two classes are shared with traditional multivariate models: multivariate data and clinical latent features. Multivariate data is represented by the standard one value per variable paradigm and is widely employed in a host of clinical models and tools. These are often represented by a number present in a given cell of a table. Clinical latent features derived, rather than directly measured, data elements that more accurately represent a particular clinical phenomenon than any of the directly measured data elements in isolation. The second two classes are unique to the time series data elements. The first of these is the raw data elements. These are represented by multiple values per variable, and constitute the measured observations that are typically available to end users when they review time series data. These are often represented as dots on a graph. The final class of data results from performing time series analysis. This class of data represents the fundamental concept on which our hypothesis is based. The specific statistical or mathematical operations are up to the modeler to determine, but we generally recommend that a variety of analyses be performed in order to maximize the likelihood that a representation of the time series data elements is produced that is able to distinguish between two or more classes of outcomes. The second manuscript, entitled "Building Clinical Prediction Models Using Time Series Data: Modeling Cardiac Arrest in a Pediatric ICU" provides a detailed description, start to finish, of the methods required to prepare the data, build, and validate a predictive model that uses the time series data elements determined in the first paper. One of the fundamental tenets of the second paper is that manual implementations of time series based models are unfeasible due to the relatively large number of data elements and the complexity of preprocessing that must occur before data can be presented to the model. Each of the seventeen steps is analyzed from the perspective of how it may be automated, when necessary. We identify the general objectives and available strategies of each of the steps, and we present our rationale for choosing a specific strategy for each step in the case of predicting cardiac arrest in a pediatric intensive care unit. Another issue brought to light by the second paper is that the individual steps required to use time series data for predictive modeling are more numerous and more complex than those used for modeling with traditional multivariate data. Even after complexities attributable to the design phase (addressed in our first paper) have been accounted for, the management and manipulation of the time series elements (the preprocessing steps in particular) are issues that are not present in a traditional multivariate modeling paradigm. In our methods, we present the issues that arise from the time series data elements: defining a reference time; imputing and reducing time series data in order to conform to a predefined structure that was specified during the design phase; and normalizing variable families rather than individual variable instances. The final manuscript, entitled: "Using Time-Series Analysis to Predict Cardiac Arrest in a Pediatric Intensive Care Unit" presents the results that were obtained by applying the theoretical construct and its associated methods (detailed in the first two papers) to the case of cardiac arrest prediction in a pediatric intensive care unit. Our results showed that utilizing the trend analysis from the time series data elements reduced the number of classification errors by 73%. The area under the Receiver Operating Characteristic curve increased from a baseline of 87% to 98% by including the trend analysis. In addition to the performance measures, we were also able to demonstrate that adding raw time series data elements without their associated trend analyses improved classification accuracy as compared to the baseline multivariate model, but diminished classification accuracy as compared to when just the trend analysis features were added (ie, without adding the raw time series data elements). We believe this phenomenon was largely attributable to overfitting, which is known to increase as the ratio of candidate features to class examples rises. Furthermore, although we employed several feature reduction strategies to counteract the overfitting problem, they failed to improve the performance beyond that which was achieved by exclusion of the raw time series elements. Finally, our data demonstrated that pulse oximetry and systolic blood pressure readings tend to start diminishing about 10-20 minutes before an arrest, whereas heart rates tend to diminish rapidly less than 5 minutes before an arrest.

Intensive care unit utilization among patients with cancer at the end of life

Over the last 2 decades, survival rates in critically ill cancer patients have improved. Despite the increase in survival, the intensive care unit (ICU) continues to be a location where end-of-life care takes place. More than 20% of deaths in the United States occur after admission to an ICU, and as baby boomers reach the seventh and eighth decades of their lives, the volume of patients in the ICU is predicted to rise. The aim of this study was to evaluate intensive care unit utilization among patients with cancer who were at the end of life. End of life was defined using decedent and high-risk cohort study designs. The decedent study evaluated characteristics and ICU utilization during the terminal hospital stay among patients who died at The University of Texas MD Anderson Cancer Center during 2003-2007. The high-risk cohort study evaluated characteristics and ICU utilization during the index hospital stay among patients admitted to MD Anderson during 2003-2007 with a high risk of in-hospital mortality. Factors associated with higher ICU utilization in the decedent study included non-local residence, hematologic and non-metastatic solid tumor malignancies, malignancy diagnosed within 2 months, and elective admission to surgical or pediatric services. Having a palliative care consultation on admission was associated with dying in the hospital without ICU services. In the cohort of patients with high risk of in-hospital mortality, patients who went to the ICU were more likely to be younger, male, with newly diagnosed non-metastatic solid tumor or hematologic malignancy, and admitted from the emergency center to one of the surgical services. A palliative care consultation on admission was associated with a decreased likelihood of having an ICU stay. There were no differences in ethnicity, marital status, comorbidities, or insurance status between patients who did and did not utilize ICU services. Inpatient mortality probability models developed for the general population are inadequate in predicting in-hospital mortality for patients with cancer. The following characteristics that differed between the decedent study and high-risk cohort study can be considered in future research to predict risk of in-hospital mortality for patients with cancer: ethnicity, type and stage of malignancy, time since diagnosis, and having advance directives. Identifying those at risk can precipitate discussions in advance to ensure care remains appropriate and in accordance with the wishes of the patient and family.^