Active use of cocaine: an independent risk factor for recurrent diabetic ketoacidosis in a city hospital.

OBJECTIVE To identify the risk factors for recurrent diabetic ketoacidosis (DKA) in a city hospital. METHODS We performed a retrospective analysis of sequential adult admissions for DKA at Bronx Lebanon Hospital Center in New York between July 1, 2001, and June 30, 2004. The patients were divided into cohorts, which were compared with use of analysis of variance and X2 tests. Multivariate logistic regression analysis was performed where indicated. RESULTS In 168 patients (96 men and 72 women), 219 episodes of DKA occurred. The mean age (+/- SD) of the overall study group was 38.6 +/- 14.8 years. Fifty-four patients (32%) had type 2 diabetes, and 44 patients (26%) had new-onset diabetes. The recurrence rate of DKA was 169% in cocaine users and 39% in nonusers (P<0.0001). Active use of cocaine, noncompliance, and Hispanic ethnicity emerged as independent risk factors for recurrent DKA-odds ratio (OR) = 4.38, P = 0.001; OR = 1.96, P = 0.05; and OR = 0.40, P = 0.005, respectively. The commonest precipitants of DKA were noncompliance (44%) and infection (26%). Noncompliance was associated with use of cocaine, use of cannabis, and cigarette smoking (P = 0.008, 0.04, and 0.01, respectively). In 91 of the hospital admissions for DKA (42%), the patients were active smokers. CONCLUSION Active use of cocaine is an independent risk factor for recurrent DKA, as are noncompliance and Hispanic ethnicity. Of these 3 factors, cocaine showed the strongest association with DKA. Therefore, toxicology screening in patients with recurrent DKA may be prudent and worthwhile.