Understanding the impact of medical homes on the health care utilization and expenditures of chronically ill adults

Growing health system quality challenges and financial constraints warrant a paradigm shift in the organization and delivery of the health care system. There are both state and national level debates for the establishment of a Patient Centered Medical Home (PCMH) as the centerpiece of primary care reform. While most states and the Centers for Medicare and Medicaid Services (CMS) have implemented demonstrations to test this model of care, limited empirical evidence is available to date. PCMH is characterized as accessible, family centered, coordinated, continuous, culturally competent and compassionate. This research seeks to estimate the impact of PCMH on health care utilization and expenditure for chronically ill adults. This is a cross-sectional design that identifies a sub-population of chronically ill adults living with asthma, diabetes, chronic obstructive pulmonary disorder (COPD), hypertension depression, arthritis, coronary heart disease and stroke adults (<18 years) ages old in the Medical Expenditure Panel Survey (2004 -2007). This analysis was restricted to individuals who had a usual source of care. To ascertain the construct of a PCMH, an exploratory factor analysis was conducted on nineteen theoretically based items. A Principle Component Analysis was used to extract a sixteen item measure with a Crohnbach alpha of 0.64. Multiple regression models assessed the association between evidence of “medical homeness” and healthcare outcome (total expenditure, physician office visits, emergency room visits, prescription drugs, inpatient discharges, and nights in hospital) among 3,125 chronically ill adults. The primary independent variable of interest was whether an individual's care showed evidence of medical homeness. Individuals who reported having more chronic conditions had a greater odds (OR = 1.07, 95% Confidence Interval CI: 1.02 -1.13) of experiencing care that showed evidence of medical homeness. Respondents who reported being in poor health (OR = 1.43, CI: 1.02 - 2.07), fair health (OR = 1.71, CI: 1.32 - 2.98) and good health (OR = 2.06, CI: 1.37 - 4.02) had greater odds of care that showed evidence of medical homeness. Conversely, individuals who reported a higher physical functioning score of SF-12 had a lower odds (OR = 0.99, CI: 0.98 - 0.99) of having care that showed evidence of medical homeness. Evidence of medical homeness was significantly associated for one of the utilization models, prescription drug (IRR = 0.32, p-value = 0.00). These findings suggest that medical homes could impact healthcare expenditures and utilization.