The Impact of Patient-Centered Medical Home Certification on Healthcare Utilization and Health Outcomes for Patients with Diabetes

The purpose of this work is to combine two panel datasets capturing statewide quality reporting about care for patients with diabetes, and health care claims data, with information about certification of primary care clinics as patient-centered medical homes (PCMH). This work leverages legislative changes that created incentives for transforming primary care practices to PCMHs and required participation in quality reporting in Minnesota. The quality of care measures were based on electronic health record data capturing five markers of optimal diabetes care, including blood sugar control, cholesterol control and blood pressure control. We used the health plan data to develop measures of resource use in total and by place of service, computed using a standardized fee schedule to neutralize differences in market power among the provider groups. We used a difference-in-differences framework to identify the impact of PCMH certification using an 8-value categorical variable that categorized the patient-year observation as (1) being from a clinic that has no history of PCMH certification, (2) is captured 3 or more years prior to the clinics certification, (3) 2 years prior to certification, (4) 1 year prior to certification, (5) from the certification year, (6) 1 year after certification, (7) 2 years after certification, or (8) 3 or more years after certification.

We found the first wave of practices that achieved PCMH certification (observed 2010-2013, certified 2010-2015) effected only small changes in quality of care prior to certification, but the latter wave of practices (observed 2015-2017, certified 2013-2017) achieved clinically meaningful increases in quality of care as they prepared for and achieved PCMH certification. Using claims data from 2009-2014, certification was found to be associated with significant increases in office-based care in the years surrounding certification, with accompanying decreases in care delivered in outpatient hospital settings.

Our results suggest that clinics that initially pursue certification are already performing at a high level and make minimal changes. Clinics not initially pursuing certification make significant changes in care delivery during the preparatory period leading up the certification process, and continue to make additional improvements after certification. The changes in resource use we identified suggest that increased patient contact in office-based settings may be providing the pathway to strengthen patient-provider relationships and provide the coaching and chronic care management needed to improve patient outcomes.