1009 — Higher Level of Patient-Centered Medical Home Implementation Associated with Larger Improvements in Clinical Quality in the PACT Initiative

Rosland AM, Center for Clinical Management Research, VA Ann Arbor; Wong ES, VA Puget Sound Health Care System; Zulman D, VA Palo Alto Health Care System; Piegari R, Department of Veterans Affairs, Canandaigua, NY; Prenovost K, Center for Clinical Management Research, VA Ann Arbor; Fihn SD, VA Office of Analytics and Business Intelligence; Nelson KM, VA Puget Sound Health Care System;

Objectives:
Starting in 2010, the VHA Patient Aligned Care Teams (PACT) initiative has focused on whole-clinic improvements in patient-centered care delivery, such as increased continuity and access, and multi-disciplinary team-based care. PACT does not explicitly focus on care for specific medical conditions or patient populations. However, effective implementation of PACT may have beneficial 'downstream' effects on chronic illness care. This study examined whether extent of PACT implementation at individual VHA primary care (PC) clinics by 2012 was associated with changes in chronic illness care between 2009 (pre-PACT) and 2013.

Methods:
In a pre-post observational study of VHA's 955 PC clinics we used multivariable linear regression to examine the associations between clinic PACT implementation (PI-Squared score) and clinic-level change in 15 individual External Peer Review Program (EPRP) clinical quality indicators. EPRP indicators are expressed as the percent of qualifying patients at a clinic who received guideline adherent care. The PI-Squared score is a 53 variable index representing 8 PCMH domains. All models were adjusted for clinics' baseline value of the EPRP measure, CBOC vs. hospital-based, rural vs. urban, and area unemployment levels. Results are reported as model-based predictions for 2009-2013 change in proportion of patients receiving guideline adherent care, for clinics with the least PACT implementation (PI-Squared < -4) vs. those with the most (PI-Squared > +4).

Results:
Overall, clinical quality was high. However, more extensive PACT implementation was significantly associated with larger improvements in five quality measures and a trend in two additional measures. Specifically, clinics with the most PACT implementation had significantly greater improvements than clinics with the least PACT implementation in: annual LDL measurement in cardiovascular disease (CVD, +2.4 percent vs. 0 percent, p < 0.01), LDL < 100 in CVD (+7.8 percent vs. +2.7 percent, p < 0.01) and diabetes (+3.1 percent vs. +0.09 percent, p = 0.06), ACE-inhibitor/ARB prescription in CVD with EF < 40 percent (+0.03 percent vs. -0.02 percent, p < 0.01), blood pressure < 160/100 in hypertension (+0.8 percent vs. -0.9 percent, p < 0.001) and diabetes (+0.2 percent vs. -1.1 percent, p = 0.03), and annual HbA1c measurement in diabetes (+1.3 percent vs. +0.5 percent, p = 0.06). Measures for which change over time was unassociated wit PI-Squared score were: aspirin prescription in CVD or diabetes; HbA1c < 9 in diabetes; regular foot, retinal or renal function checks in diabetes; and heart failure with ACE-inhibitor/ARB prescription.

Implications:
Clinics with PACT most extensively in place by 2012 had significantly larger improvements in almost half of the chronic disease process and outcome quality measures examined.

Impacts:
Focusing resources on PCMH-aligned changes in care delivery across all patients may result in downstream improvements in quality of care for patients with specific chronic diseases.