3002 — Does the Adoption of Patient Aligned Care Team (PACT) Structure Improve Health Care Delivery?

Crawford ER, Center for Comprehensive Access & Delivery Research and Evaluation, Iowa City; Tippie College of Business, University of Iowa; Reeves CJ, Center for Comprehensive Access & Delivery Research and Evaluation, Iowa City; Tippie College of Business, University of Iowa; Stewart GL, Center for Comprehensive Access & Delivery Research and Evaluation, Iowa City; Tippie College of Business, University of Iowa; Astrove SL, Center for Comprehensive Access & Delivery Research and Evaluation, Iowa City; Tippie College of Business, University of Iowa;

Objectives:
Use network analysis to visualize and quantify the extent to which PACT structure has been adopted in VHA, and to link differences in structure with primary care performance measures.

Methods:
Network analysis procedures graphically portray differences in the PACT structure in 864 unique VHA health care facilities. On average, each facility consists of 8.74 primary care teamlets responsible for a mean of 5,755 patients each. We quantitatively captured differences in structure using network measures of degree (number of teamlet memberships), betweenness (bridging between teamlets), Blau's index (role coverage), and centralization (one highly central person assigned to many teamlets). We used Beta Regression to link these differences to performance measures.

Results:
There is wide variation in the degree to which PACT structure has been implemented. Only 19% of PACT teamlets strictly conform to recommended PACT implementation principles. One major contributor to deviation occurs when a teamlet is missing a key role. Clinical Associate is missing for 20.1% of teamlets, Administrative Associate for 11.4%, and Nurse Care Manager for 5.6%. Another contributor to deviation occurs through sharing of members across teamlets. We found that 35.9% of Administrative Associates, 32.2% of Nurse Care Managers, and 25.8% of Clinical Associates are shared. Beta Regression results reveal that teamlets with fewer average teamlet assignments per person and complete coverage of all four core roles exhibited better outcomes in terms of ER/Urgent Care Utilization, 2-Day Post Discharge Contact, Percentage of Telephone Encounters, and Percentage of Patients seen within 1-day of Desired Date. We found no relationships with Continuity of Care.

Implications:
Implementation of PACT structure varies greatly in VHA and is predictive of critical outcomes. In particular, having staff assigned to one unique teamlet and having balanced coverage of the four roles is associated with better access and coordination.

Impacts:
Our results suggest that failure to fully adopt team-based structure is a possible explanation why PACT implementation has been slower than desired. We also show that recommended PACT structure does indeed correspond with better access and coordination of care.