3020 — Changes in Utilization and Costs after PACT Implementation through Evidence-Based Quality Improvement

Yoon J, VA Palo Alto; Chow A, VA Palo Alto; Rubenstein LV, VA Greater Los Angeles;

Objectives:
Several VA primary care practices emphasized an evidence-based quality improvement (EBQI) approach to facilitate and support their transformation into Patient Aligned Care Team (PACT) models. We assessed changes in VA health care utilization and costs for patients receiving care from practices using EBQI methods and comparison practices over a five-year period.

Methods:
We conducted a longitudinal study of 136,856 patients from 6 practices that were selected for EBQI and 28 practices that were not for FY2009-2013. Information on practice characteristics, patient demographics, chronic conditions, and VA utilization and costs was obtained from administrative data for one year prior to and four years following PACT implementation. We measured ambulatory care encounters, ED visits, all-cause admissions, and total health care costs per patient. Time trends in utilization and costs were compared between practices that did and did not use EBQI in bivariate analyses, and multilevel regressions were conducted adjusting for EBQI status, year, patient and clinic factors, and patient random effects. We used count data models for utilization outcomes and a linear regression for log-transformed costs.

Results:
EBQI practices were more likely to be VAMC sites and larger than non-EBQI practices (all P < 0.05). Patients in all practices had lower utilization for primary care, specialty care, and mental health/substance abuse care and higher utilization of telephone care after PACT implementation (all P < 0.001). They also had fewer hospitalizations and lower costs per patient over the study period. For patients in EBQI practices, the magnitude of utilization changes was larger than for other patients. In adjusted analyses, EBQI practice was independently associated with a lower rate of primary care (IRR = 0.85), specialty care (IRR = 0.83), and mental health care encounters (IRR = 0.69). There was no independent effect of EBQI on prescription drug use, acute care, or health care costs.

Implications:
EBQI methods appeared to enhance the effects of PACT implementation in several practices by leading to lower outpatient care use for certain types of care. However, acute care use and health care costs overall were not affected by EBQI.

Impacts:
Given the challenges practices face in attempting to implement PACT, support through an EBQI framework may provide the means to fully engage in practice transformation.