Increased incidence and severity of diabetic ketoacidosis among uninsured children with newly diagnosed type 1 diabetes mellitus

OBJECTIVES: (a) To determine the incidence and severity of diabetic ketoacidosis (DKA) and (b) to stratify according to insurance status at the initial diagnosis of type 1 diabetes (T1DM). RESEARCH DESIGN AND METHODS: Subjects included children <18 yr who presented with new-onset T1DM from January 2002 to December 2003 and were subsequently followed at the Barbara Davis Center. Insurance status and initial venous pH were obtained. RESULTS: Overall, 383 subjects presented with new-onset T1DM and 359 (93.7%) were enrolled. Forty-three (12.0%) of these children were uninsured and 40 (11.1%) had Medicaid. One hundred and two (28.4%) subjects presented with DKA. When compared to the insured subjects, uninsured subjects had a significantly increased risk of presenting with DKA odds ratios (OR): 6.19, 95% CI 3.04-12.60, p < 0.0001], as well as presenting with severe DKA, defined as venous pH <7.10 (OR: 6.09, 95% CI 3.21-11.56, p < 0.0001). There were no differences, however, between the insured and Medicaid subjects in their probability of presenting with DKA or severe DKA. The risk of presenting with DKA (as well as with severe DKA) was the highest among patients <4 yr old. CONCLUSIONS: At the time of initial diagnosis, uninsured patients were more likely to present with DKA than insured patients. Furthermore, when the uninsured subjects presented with DKA, the condition tended to be more severe and life-threatening. A potential explanation is that uninsured subjects may delay seeking timely medical care, thereby presenting more critically ill, whereas insured subjects may have their T1DM diagnosed earlier.