Opting out of an integrated healthcare system: Dual-system use is associated with poorer glycemic control in veterans with diabetes

Abstract

To test for an association between quality of care and patient choice to obtain care outside an integrated healthcare delivery system. We used administrative data to define dual-system use (Veterans Health Administration (VHA) and Medicare) in 1999 for VHA users with diabetes over 65 years old. Quality of diabetes care was determined by the last hemoglobin A1c (HA1c) value in 2000. The distance to nearest VHA facility minus the distance to nearest non-federal hospital was the instrumental variable in a two-part regression model which controlled for observed and unobserved factors. In 1999, 57.4% of subjects received care from both VHA and Medicare providers; their mean proportion of visits to Medicare providers was 0.41 (median 0.38). After controlling for observed and unobserved factors, higher proportions of Medicare visits were significantly associated with higher HA1c values; a 40% increase in the proportion of Medicare visits by those who did not use Medicare was associated with a 0.23% point increase in HA1c value. Dual-system use was associated with higher HA1c values, suggesting that veterans who chose to receive care outside the integrated VHA may have worse intermediate clinical outcomes than those who received care exclusively within the system.