Return toAll Abstracts
HSR&D 2004 National Meeting Abstracts


2029. VA Primary Care Organizational Characteristics Associated with Lower HbA1c
George L Jackson, MHA, Center for Health Services Research in Primary Care, Durham VAMC and Duke University, EM Yano, Center for the Study of Healthcare Provider Behavior, Sepulveda VAMC and University of California at Los Angeles, SL Krein, Center for Practice Management and Outcomes Research, Ann Abor Healthcare System and University of Michigan, D Edelman, Center for Health Services Research in Primary Care, Durham VAMC and Duke University, MA Ibrahim, University of North Carolina at Chapel Hill (emeritus) and John Hopkins University, TK Dudley, Center for Health Services Research in Primary Care, Durham VAMC, M Weinberger, Center for Health Services Research in Primary Care, Durham VAMC and University of North Carolina at Chapel Hill

Objectives: Examine organizational features of VA primary care (PC) clinics having potential associations with better diabetes outcomes.

Methods: We established a cohort of 224,221 diabetic patients (alive October 1, 1999) using VA Diabetes Registry and VHA Medical SAS Datasets. 1999 VHA Survey of Primary Care Practices results were combined with individual patient data. A two-level hierarchical model determined percentage of variance in HbA1c accounted for by facility differences. Elements of care structure were assessed for potential association with HbA1c levels (follow-up period FY 2000-FY 2001).

Results: Variance in HbA1c resulting from facility level differences was lower in FY 1999 (5.9%) and FY 2000 (3.9%) than reported by others for FY 1998 (12%). It increased in FY 2001 to 8.4%. Attributes associated with lower HbA1c (p < 0.1 for all) were: greater staffing authority (lower by 0.8354 [per 10 points above scale average]); computerized reminders (0.1567); disease management teams (0.1370); large academic practice (0.1623); and notifying all patients of PC provider (0.1766). Associated with greater HbA1c were: more diffuse guideline support (greater by 0.03115 [per 10 points above scale average]); greater authority to establish interorganizational relationships (0.7682 [per 10 points above scale average]); program reporting patients always see assigned provider (0.1711); and TQI program involving all nurses without all physicians (0.3434). Results were risk adjusted by patient comorbidities, demographics, and medication.

Conclusions: Systems that simultaneously lead to using integrated teams involving physicians, actively supporting evidence-based guidelines, and engaging patients may lead to clinically important outcome improvement.

Impact: Enhancing team-focused PC systems may improve VA care.