The fragmented approaches to delivering health care services in the United States, along with the associated structural inefficiencies and unsustainable increases in health care costs affecting all payers, compel the need for reform. Various federal and state-level delivery system reform models have emerged in response. The Medical Home (MH) is one of such reform models. In 2004 a national initiative entitled "The Future for Family Medicine Project" identified the lack of emphasis on comprehensive primary care, especially for chronic care patients, and proposed the introduction of MHs to improve comprehensive primary care delivery for every patient. Oregon's MH variant, the Patient-Centered Primary Care Home (PCPCH), was introduced in 2009 as part of a state-wide health reform initiative ushered in by the passage of House Bill 2009 to promote the Triple Aim. Since 2011, over 600 primary care clinics have been recognized as PCPCHs. Proponents of the model argued that it will help improve comprehensive primary care services upstream and reduce inappropriate utilization of Emergency Department (ED) and Inpatient (IP) care and expenditures downstream. Evidence on the model's application to reduce ED and IP utilization and expenditures have so far been mixed. Based on growing interests in the effects of the model's application to provide care for different types of patients, this research was designed to evaluate the policy effects of the application of PCPCHs, with a focus on PCPCHs that treat greater proportions of chronic care patients, to answer the following questions: (1) What is the average chronic disease burden of PCPCHs, and how does their average chronic disease burden compare to the communities PCPCHs are in pre-post PCPCH recognition? (2) How do primary care expenditures change based on the chronic disease burden of PCPCHs? (3) Do PCPCHs that engage more high chronic disease burden patients have more reductions in ED and IP expenditures? For this research, a chronic disease burden measure was developed from 10 markers of chronic conditions. This measure was then used to stratify PCPCH clinics and their comparators into high and low chronic disease burden clinics. The research was designed as a natural experiment, utilizing difference-in-difference methods to measure outcome differences pre-post PCPCH policy implementation and comparing outcome differences between PCPCHs and their control groups. The unit of analysis was PCPCH clinics. The theoretical perspectives that informed this research were Risk Selection and Complex Adaptive Systems (CAS). Data from Oregon's All Payer All Claims (APAC) data system, which included 16 quarters of claims and eligibility data from fourth quarter 2010 to third quarter 2014, as well as PCPCH attestation data on 525 clinics were utilized for this research. The results suggest that the chronic disease burden for PCPCHs was significantly lower than their comparator groups before clinics recognition as PCPCHs, but the chronic disease burden did not change after clinics recognition as PCPCHs. Average primary care expenditures did not change after PCPCH recognition. Average ED and average IP expenditures for high chronic disease burden PCPCHs did not change but rather decreased significantly for low chronic disease burden PCPCHs. The results imply that the distribution of chronic disease burden in PCPCHs is important and related to ED and IP expenditures, but in a different direction than expected. The results also suggest that focusing on low chronic disease burden patients in PCPCHs could help reduce ED and IP expenditures in the short and medium terms. Policies to engage a broader mix of chronic disease burden patients in PCPCHs could help increase savings from ED and IP utilization. The results also suggest the need for more research to improve current understanding of how PCPCHs are impacting health care trajectories in the current delivery system environment.

Managed care plans purport to improve the health of their members with chronic diseases. How has the growing adoption of Medicare Advantage (MA), the managed care program for Medicare beneficiaries, affected the progression of chronic disease? The literature is rich with articles focusing on managed care organizations' impacts on quality of care, access, patient satisfaction, and costs. However, few studies have analyzed these impacts with respect to market penetration of Medicare managed care. The objective of this research has been to analyze the relationships between the market penetration of MA plans and the progression of chronic diseases among Medicare beneficiaries. The Chronic Disease Severity Index scale (CDSI) was constructed to represent beneficiaries' overall chronic disease states for survey or claims-based data, when more direct clinical measures of disease progression are not available. Using the CDSI on the MEPS survey dataset from AHRQ, we sought to assess the impacts of MA market penetration and other covariates on the overall chronic disease state of Medicare beneficiaries from 2004 through 2008. Though the model explains much of the variation in CDSI change, the author expected the multilevel model would show that MA penetration explains a significant level of variation in CDSI change. However, this hypothesis was not substantiated, and the findings suggest that unmeasured factors may be contributing to additional unexplained heterogeneity. Policymakers should explore opportunities to refine the current MA program. The MA program costs the federal government more than the Traditional Fee-for-Service Medicare program, and there is no definitive evidence that outcomes differ. Within both programs, there is opportunity to experiment with different models of payment, healthcare service delivery and care coordination. The Patient Protection and Affordable Care Act (ACA) contains provisions for innovative demonstration projects in delivery and payment. The effectiveness of these ACA initiatives must be monitored, both for impacts on health outcomes and for economic effects. This research can inform future approaches to outcomes assessment using the CDSI, and multilevel modeling methodologies similar to those employed here. Firms offering MA health plans would be prudent to proactively demonstrate their value to beneficiaries and taxpayers. They should explore means of better monitoring and reporting the longitudinal outcomes of their enrolled beneficiaries. Demonstrating that they can bring value in terms of improved health outcomes will help insure their long-term survival, both in the marketplace and in the political arena. / Graduation date: 2012