No description available.

No description available.

No description available.

<p>Despite the proven clinical benefits of use of adjuvant hormone therapy with tamoxifen or aromatase inhibitors for breast cancer, adherence to and persistence with adjuvant hormone therapy are suboptimal. It is critical to understand the clinical and economic impacts of low adherence to and low persistence with adjuvant hormone therapy in breast cancer. The overall objective was to assess associations between adherence to and persistence with adjuvant hormone therapy and mortality, healthcare utilization, and healthcare costs among older women with breast cancer. A retrospective longitudinal analysis of the Surveillance, Epidemiology, and End Results (SEER) registry linked with Medicare claims was conducted. This study included 25,796 older women diagnosed with hormone receptor-positive stage I-III breast cancer from 2009 through 2017. Adherence was defined as having proportion of days covered (PDC) of 0.80 or more. Persistence was defined as having no hormone therapy discontinuation, i.e., a break of at least 180 continuous days. Length of persistence was calculated as time from therapy initiation to discontinuation. All analyses were conducted using SAS 9.4 and RStudio for Linux environment. An <em>a priori</em> alpha level of 0.05 was used to determine significance for all the analyses. Time-dependent Cox models were used to assess associations between adherence to and persistence with adjuvant hormone therapy and mortality. Hurdle generalized linear mixed models were used to assess associations between adherence and persistence with annual number of hospitalizations, hospital days, hospital outpatient visits, inpatient costs, and outpatient costs across five years to account for excess zeroes.  Generalized linear mixed models were used for other types of healthcare utilization and costs. Annual adherence rates were 78.1 percent, 75.2 percent, 72.4 percent, 70.0 percent, and 61.5 percent from year-one to year-five after hormone therapy initiation. Persistence rates were 87.5 percent, 81.7 percent, 77.1 percent, 72.9 percent, and 68.9 percent through cumulative intervals of one year up to five years after hormone therapy initiation. Adherence was associated with lower risk of all-cause mortality, but was not significantly associated with breast cancer-specific mortality. Both being persistent and longer persistence were associated with lower risk of all-cause mortality and lower risk of breast cancer-specific mortality. Being adherent was associated with fewer hospitalizations, fewer hospital days, fewer emergency room visits, and fewer hospital outpatient visits, but was not associated with physician office visits. Being persistent was associated with fewer hospital days, fewer emergency room visits, and fewer hospital outpatient visits, but was associated with more physician office visits. Longer persistence was associated with fewer hospital days, fewer emergency room visits, and fewer hospital outpatient visits, but was not significantly associated with physician office visits. Adherent women had lower inpatient costs, lower outpatient costs, lower medical costs, and lower total healthcare costs despite higher prescription drug costs. Both being persistent and longer persistence were associated with lower inpatient costs, lower outpatient costs, lower medical costs, and lower total healthcare costs despite higher prescription drug costs. </p>

<p>Predictive modeling in healthcare involves the development of data-driven and computational models which can predict what will happen, be it for a single individual or for an entire system. The adoption of predictive models can guide various stakeholders’ decision-making in the healthcare sector, and consequently improve individual outcomes and the cost-effectiveness of care. With the rapid development in healthcare of big data and the Internet of Things technologies, research in healthcare decision-making has grown in both importance and complexity. One of the complexities facing those who would build predictive models is heterogeneity of patient populations, clinical practices, and intervention outcomes, as well as from diverse health systems. There are many sub-domains in healthcare for which predictive modeling is useful such as disease risk modeling, clinical intelligence, pharmacovigilance, precision medicine, hospitalization process optimization, digital health, and preventive care. In my dissertation, I focus on predictive modeling for applications that fit into three broad and important domains of healthcare, namely clinical practice, public health, and healthcare system. In this dissertation, I present three papers that present a collection of predictive modeling studies to address the challenge of modeling heterogeneity in health care. The first paper presents a decision-tree model to address clinicians’ need to decide among various liver cirrhosis diagnosis strategies. The second paper presents a micro-simulation model to assess the impact on cardiovascular disease (CVD) to help decision makers at government agencies develop cost-effective food policies to prevent cardiovascular diseases, a public-health domain application. The third paper compares a set of data-driven prediction models, the best performing of which is paired together with interpretable machine learning to facilitate the coordination of optimization for hospital-discharged patients choosing skilled nursing facilities. This collection of studies addresses important modeling challenges in specific healthcare domains, and also broadly contribute to research in medical decision-making, public health policy and healthcare systems.</p>

<p dir="ltr">Background. Although depression is a risk factor for cardiovascular disease (CVD), few clinical trials in people without CVD have examined the effect of depression treatment on CVD-related outcomes. It’s unknown if successful depression treatment improves indicators of CVD risk, such as CVD-relevant health behaviors, traditional CVD risk factors, and CVD events. </p><p dir="ltr">Methods. We examined data from eIMPACT trial, a phase II randomized controlled trial conducted from 2015-2020. Depressive symptoms, CVD-relevant health behaviors (self-reported CVD prevention medication adherence, sedentary behavior, and sleep quality) and traditional CVD risk factors (blood pressure and lipid fractions) were assessed. Incident CVD events over four years were identified using a statewide health information exchange. </p><p dir="ltr">Results. The intervention group exhibited greater improvement in depressive symptoms (p < 0.01) and sleep quality (p < 0.01) than the usual care group, but there was no intervention effect on systolic blood pressure (p = 0.36), low-density lipoprotein cholesterol (p = 0.38), high-density lipoprotein cholesterol (p = 0.79), triglycerides (p = 0.76), CVD prevention medication adherence (p = 0.64), or sedentary behavior (p = 0.57). There was an intervention effect on diastolic blood pressure that favored the usual care group (p = 0.02). CVD-relevant health behaviors did not mediate any intervention effects on traditional CVD risk factors. Twenty-two participants (10%) experienced an incident CVD event. The likelihood of an CVD event did not differ between the intervention group (12.1%) and the usual care group (8.3%; HR = 1.45, 95% CI: 0.62-3.40, p = 0.39). </p><p dir="ltr">Conclusions. Successful depression treatment alone improves self-reported sleep quality but is not sufficient to lower CVD risk of people with depression. Alternative approaches may be needed reduce CVD risk in depression. </p><p dir="ltr">Trial Registration: ClinicalTrials.gov Identifier: NCT02458690 </p><p dir="ltr">Keywords: depression, cardiovascular disease, blood pressure, lipids, medication adherence, sedentary behavior, sleep quality, collaborative care, internet interventions, clinical trial</p>