Factors associated with discontinuing insulin therapy after diabetic ketoacidosis in adult diabetic patients.

AIMS: To assess the factors associated with successful discontinuation of insulin therapy after diabetic ketoacidosis (DKA) in adult patients. METHODS: Patients (>or= 18 years) attending the Endocrine and Metabolism Clinic at a major hospital in southern Taiwan were recruited. After recovery from the acute stage, those with no contraindications to oral antidiabetic agents, with adequate beta cell reserve, and with no antiglutamic acid decarboxylase (GAD) antibody were treated with oral agents. RESULTS: Sixty-six patients (38 males, 28 females, aged 18-76 years) were included, and 21 qualified for treatment with oral agents. These 21 patients were older at diagnosis of diabetes (45.5 +/- 14.0 vs. 40.0 +/- 13.8 years, P = 0.047), had shorter diabetes duration (median 0 vs. 5.5 months, P = 0.040), higher BMI (median 23.4 vs. 19.5 kg/m2, P < 0.001), higher serum osmolality during DKA (352.1 +/- 40.7 vs. 318.0 +/- 16.4 mmol/kg, P = 0.005), and lower insulin dose following recovery (median 0.49 vs. 0.83 unit/kg/d, P < 0.001) than those patients that had to continue insulin therapy. Thirteen patients (8 males, 5 females; 62%) successfully discontinued insulin for at least one year without recurrence of DKA. Multiple logistic regression analyses showed that BMI >or= 25 kg/m2 (adjusted relative risk (ARR) 8.85, 95% CI 1.05, 8.39), diabetes onset age >or= 40 years (ARR 8.08, 95% CI 1.16, 6.95), and undiagnosed diabetes before DKA (ARR 8.90, 95% CI 1.19, 7.51) were significant factors associated with successful discontinuation of insulin therapy. CONCLUSION: We identified three independent clinical factors associated with successful discontinuation of insulin therapy after DKA.