Insurance coverage is associated with receipt of diabetes preventive care in linked safety net practice data / Medicaid claims data, 2005-2007

Background: Insurance coverage is associated with improved care in the general population, but its impact among safety net clinic patients is poorly understood. Methods: We linked Oregon Medicaid claims with safety net clinics' practice management data, and evaluated the association between percent time insured and receipt of diabetes preventive care services. Our study population included 4240 adults with ≥2 diabetes-associated clinic visits in 2004-2005, ≥1 clinic visit in 2006, and ≥1 in 2007. Results: During the study period (2005-2007) 24% of the patient population had no insurance coverage, 47% continuous coverage, and 29% partial coverage. Among those partially covered, approximately one-third had coverage for 1-59% of the study period, and two-thirds for 60-99%. Those continuously covered were more likely to receive recommended care than were those partially covered or not covered. Partial coverage was more favorable than no coverage; for example, 85% of those continuously covered received ≥1 LDL screen in 2005-2007, versus 78% of those partially covered and 70% of those with no coverage. We found similar patterns for flu vaccination, HbA1c screening, and microalbumin screening, in 2005-2007 and each year alone. For example, in 2007 alone, 33% of those continuously covered received microalbumin screening, versus 29% with partial coverage and 23% with none. Conclusions: Continuous insurance coverage (versus partial or no coverage) is associated with improved receipt of preventive diabetes care, even among persons receiving care in the safety net. These results illustrate the potential for using linked safety net practice data for policy impact evaluations.