摘要：State efforts to identify subpopulations at higher risk for inadequate diabetes maintenance are sometimes hampered by small sample size. We provide a model of a cross-state collaboration that might provide the foundation for identifying political and economic forces underlying inter- and intra-state variability in chronic disease care. We collected Behavioral Risk Factor Surveillance System data directly from 5 of 6 New England states and ran multivariate logistic regressions on 5 exposures: race/ethnicity, federal poverty level (FPL) bracket, insurance status (yes or no), insurance type (public or private), and state of residence. Our sample consisted of adults aged 35 or older diagnosed with diabetes. Outcomes included whether respondents with diabetes received complete annual diabetes care (≥2 hemoglobin A 1c tests, eye examination, foot examination), had ever taken a diabetes self-management class, or reported diabetes-related retinopathy. Half (50.4%) of our sample had incomplete annual diabetes care. In multivariate logistic regressions, race/ethnicity and FPL bracket were not major drivers of outcomes, although Hispanic/Latino adults had significantly higher risk than non-Hispanic white adults of not knowing how many hemoglobin A 1c tests they had had in the past year or what such a test is (adjusted odds ratio = 2.74 [95% confidence interval, 1.15-6.56]) and of diabetes-related retinopathy (adjusted odds ratio = 3.13 [95% confidence interval, 1.61-6.10]). With few exceptions, higher FPL bracket, insurance status, insurance type, and state of residence were not associated with diabetes maintenance. Inadequate annual diabetes care among adults with diagnosed diabetes was endemic even in this relatively advantaged US census division, and traditional disparities (eg, race/ethnicity, FPL bracket) only partially explained patterns in diabetes maintenance activities. Interstate analyses can create the foundation for active partnerships to identify and address the causes of lapses in care.
