1017 — Patient Racial/Ethnic Composition and Other Facility Characteristics Associated with Delivery of Patient-Centered Care in VHA Outpatient Facilities

Hausmann LR, COIN - Pittsburgh/Philadelphia (CHERP); Gao S, COIN - Pittsburgh/Philadelphia (CHERP); Canamucio A, VISN 4 Center for Evaluation of PACT (CEPACT);

Objectives:
Despite national initiatives to increase patient-centered care through Patient Aligned Care Teams (PACTs), delivery of patient-centered care varies across VHA facilities. Moreover, facilities serving high proportions of racial/ethnic minority patients often rank lowest in patient-centeredness. Understanding why this variation exists is essential to delivering equitable care across VHA. Therefore, we examined the association of patient racial/ethnic composition and other facility characteristics with patient-centered care as measured by patient-centered components of the PACT Implementation Progress Index (Pi2).

Methods:
We analyzed facility-level administrative data on VHA outpatient facilities from fiscal year 2013 (FY13). Our outcome was a modified patient-centered Pi2 score based on care comprehensiveness, self-management support, communication, and shared decision-making (range: -4 to 4, higher = more patient-centeredness). Our independent variables were the %Black, %Hispanic, and %Other racial/ethnic minority patients served by each facility, and other facility characteristics possibly associated with patient-centered care (i.e., patient volume and clinical severity; staff ratio, turnover, racial composition, and satisfaction). We used mixed effects linear regression to examine unadjusted associations between each characteristic and patient-centered Pi2, accounting for clustering of VHA substations within parent stations. We examined a base model containing only the patient racial composition variables, then adjusted for other characteristics associated with patient-centered Pi2 at p < 0.10, and retained those with p < 0.05 in a final model.

Results:
The analyses included 908 facilities. Among the patient racial composition variables, only %Hispanic patients was associated with patient-centered Pi2 scores (Estimate [Est] = -2.78, Standard Error [SE] = 1.03, p < 0.007). This negative association remained significant after adjusting for additional facility characteristics associated with patient-centered Pi2 (Est = -2.16, SE = 1.04, p < .038). In the final model, patient-centered Pi2 was negatively associated with panel size and primary care volume, and positively associated with patient clinical severity, %White medical doctors, and employee satisfaction.

Implications:
VHA facilities serving higher proportions of Hispanic patients have lower levels of patient-centered care, even after controlling for other facility characteristics associated with patient-centeredness.

Impacts:
This analysis identifies organizational factors associated with patient-centered care at VHA outpatient facilities. Efforts are needed to improve care at facilities serving Hispanic Veterans. Larger facilities and facilities with low employee satisfaction are additional areas of opportunity for improving patient-centered care.