American Public Health Association (APHA) Annual Meeting, November 10-14, 2018, San Diego, California
Annemarie Hirsch, PhD, MPH, Melissa Poulsen, PhD, MPH, Karen Bandeen-Roche, PhD, MS, Gary Wand, MD, Sherita Golden, MD, MHS, FAHA, Lorraine Dean, ScD, Cara Nordberg, MPH, Dione Mercer, BS, Brian Schwartz, MD, MS and Allison Bleistein, BS,
BACKGROUND: Type 2 diabetes (T2D) ranges in prevalence from under 4% to almost 18% in U.S. counties. Contextual factors that explain most of the variation in T2D prevalence in the Midwest, explain little variation in Pennsylvania. Identifying the contextual factors that influence T2Dl in Pennsylvania is critical to developing targeted prevention strategies.
OBJECTIVE: As a The Diabetes LEAD (Location, Environmental Attributes & Disparities) Network member, we are studying T2D in Pennsylvania, focusing on: chronic environmental contamination, and social, and physical activity environments. The study setting is the large region of Pennsylvania served by Geisinger, a health system, allowing us to link contextual factors to Geisinger’s electronic health records (EHR). Here, we describe county-level variation in Pennsylvania for T2D prevalence and control.
METHODS: We calculated age-adjusted prevalence (2014-2015) for each of 38 counties in Pennsylvania among adult Geisinger primary care patients. We applied a previously validated EHR-based algorithm to identify patients with T2D. Poor T2D was defined as having fewer than half of HbA1c values ≤ 7.0%.
RESULTS: The mean age-adjusted prevalence of T2D was 13.9 (range, 4.8 to 22.3.) The mean prevalence of poor HbA1c control was 36.5 (range, 16.5 to 65.8.) In 7 of the 15 counties with T2D prevalence below the mean, more than 40% of patients had poor control.
CONCLUSION: The variation and discordance in T2D prevalence and control across counties suggests differential influence of community contextual factors on these outcomes, underscoring the importance of identifying relevant contextual factors and the pathways by which they operate.