Abdominal Pathology in Patients With Diabetes Ketoacidosis

Introduction: The objective is to describe the incidence and nature of significant abdominal pathologies in patients with diabetes ketoacidosis (DKA) and abdominal pain. Methods: Retrospective chart review (N = 86) of patients with DKA from January 1, 2005, to January 31, 2010, was performed. Data included demographics, co-morbidities, compliance, chief complaints and physical findings, blood count, metabolic profile, lactic acid, glycosylated hemoglobin (HbA1C), amylase, lipase, anion gap, arterial gases, imaging and final diagnosis. Continuous variables were described as mean +/- standard deviation and compared with the Student's t test. Categorical variables were expressed as percentages (%) and compared with the Mantel-Haenszel chi(2) test. Univariate analysis was conducted among patients with and without significant abdominal pain and also with and without significant abdominal pathology. Two lipase strata were created at 400 U. Multivariate model to identify limits (confidence interval) of the estimated risk imposed by the predictor found significant in univariate analysis. A P value of <= 0.05 was considered significant. Stat View 5.0 (SAS Institute, Cary, NC) was used for the statistical analysis. Results: In patients with abdominal pain, 17% had significant abdominal pathology mainly acute pancreatitis (AP). Serum amylase and lipase level were found to be an indicator of significant underlying pathology (both P values <= 0.001). The logistic model created showed that patients with lipase level >= 400 U have a 7% increased risk of having AP with confidence interval of 0.01 to 0.6. Conclusion: Patients with DKA and abdominal pain with lipase >400 U have an increased risk of significant underlying abdominal pathology (AP).