SUN-136 Know Your Patients: Our Experience with DKA in the Suburbs

Background: A variety of precipitating factors and clinical presentations have been identified for Diabetic Ketoacidosis (DKA). Most of the available data is from inner city hospitals which cater to a specific demographic much different from suburban medical centers. This study was performed at a suburban community hospital to analyze disease and demographic characteristics of patients admitted with DKA. Methods: Data was retrospectively collected from electronic medical record review for patients admitted with DKA between January 1, 2017 and January 1, 2018. Data included demographics, medical history, hemoglobin A1C, ancillary laboratory tests and length of hospital stay. Patients who left against medical advice were excluded. For patients with multiple DKA admissions, only the first admission was analyzed. Per hospital policy, all DKA patients were admitted to the medical intensive care unit. Results: 55 hospitalizations met the inclusion criteria involving 50 patients. 5 out of 55 patients had readmissions for DKA. The mean age was 47(range 17-85) years. 52% (n=26) of the population were males. The patient population comprised of 52% (n=26) African Americans, 48 % (n=24) Caucasians and 2%(n=1) other ethnic groups. The most common presenting symptom was nausea/vomiting reported by 48% (n=24) of the study subjects. Lethargy/fatigue, abdominal pain was the second and third most common reported symptom by 30% (n=15) and 26% (n=13) subjects respectively. Other presenting symptoms were polyuria 20%(n=10), polydipsia 18%(n=9), confusion 18%(n=9), weakness 12%(n=6), headache 4%(n=2) and blurring of vision 2%(n=1). The above reported symptoms were not mutually exclusive. The most common precipitant for DKA was medication noncompliance 44%(n=22) which was comparable in African American (11/26) and Caucasian (11/24) subgroups. This was followed by infectious etiologies in 36%(n=18) of the cases. DKA was the first presentation of undiagnosed DM in 10%(n=5) of the subjects. Steroid use was implicated in 4%(n=2) patients. Insulin pump failure was implicated in one subject. No definite cause was identified in 4%(n=2) of the patients. Mean hemoglobin A1C was 12.5% (range 8.2-12.5). The mean acute physiology and chronic health evaluation score (APACHE II) at admission was 15 (range 3-28). There was no mortality in the study population. The average length of hospital stay was found to be 3.7 days. Conclusions: (1) Suburban patients with DKA have subtle differences in characteristics as compared to inner city residents and more studies are required to explore these characteristics. (2) Elevated hemoglobin A1C reflects the need for better long-term glycemic control even in patients reporting good compliance. (3) Medication noncompliance rates appear to be independent of ethnicity. (4) Despite high APACHE scores, DKA remains a low mortality condition.