Patients With Chronic Illness in the Patient-Centered Medical Home: Costs, Use, Quality and Morbidity-Based Variation

Abstract

Originally described in 1967 as a central source of a medical record for children with special health care needs, the "medical home" is now being tested on a national scale as a model of primary care delivery for children and adults. The recent rise in prominence of the patient-centered medical home (PCMH) coincides with increasing acknowledgement that primary care needs to be redesigned to address the needs of chronically ill individuals, whose care accounts for 85 percent of American health care costs. The PCMH unites the core attributes of primary care with the chronic care model, and has been implemented in the context of individual chronic illnesses, but many questions regarding the PCMH's impacts on chronically ill individuals remain unanswered. In this dissertation, I address gaps in the evidence base through three studies of patients with chronic illnesses in two sequential PCMH redesigns in an integrated health care delivery system. The first study investigated outcomes for patients with three common chronic illnesses in a 2007-08 PCMH prototype redesign at one clinic, compared to a control group with the same chronic illnesses at 19 non-intervention control sites. In the second study, I examined whether secure electronic messaging and telephone encounters substituted for, or complemented, primary care office visits among patients with diabetes in a 2009-10 system-wide PCMH redesign. The third study described changes in outpatient specialty care utilization, and variation according to overall morbidity burden, among patients with treated hypertension in the system-wide PCMH redesign. I observed modestly improved quality of care at the PCMH prototype clinic during 2007-08. Compared to controls, PCMH patients had seven percent lower total health care costs over two years, largely driven by lower utilization, and associated costs, of inpatient and emergency/urgent care. Results from the second study suggested that telephone encounters and, to a lesser extent, secure electronic message threads served as complements to office visits for individuals with diabetes. In the final study, I observed, on average, small decreases in total specialty visits for patients with hypertension in the two years during and immediately following system-wide PCMH implementation. In low morbidity patients this decrease was rapid and sustained over three years. Dissertation findings improve our understanding of the PCMH's impacts on costs, quality and health care use in chronically ill individuals, and can be applied to the planning, implementation and evaluation stages of future PCMH redesigns.