Using a severity of illness scoring system to assess intensive care unit admissions for diabetic ketoacidosis.

OBJECTIVE: To determine the appropriateness of intensive care unit (ICU) admissions for patients with the diagnosis of diabetic ketoacidosis (DKA) at our institution.

DESIGN: Retrospective chart review.

SETTING: Tertiary care inner-city hospital.

SUBJECTS: All subjects admitted to the medical intensive care unit (MICU) from September 1996 to June 1997 with a diagnosis of DKA were included.

INTERVENTIONS: None.

MEASUREMENTS AND MAIN RESULTS: A grading system for the severity of DKA (grades 0-IV) from a previously published work was modified. Admissions to the MICU that were deemed appropriate a priori included patients with grade III or IV DKA, patients with grade II DKA who were older than 65 yrs of age, or patients with special situations normally warranting intensive care.

MAIN RESULTS: Sixty-seven cases of DKA were reviewed. Two thirds of the patients had type I diabetes mellitus, and approximately 50% were men. No deaths occurred. Forty-four patients (66%) met the a priori ICU admission criteria. The average hospital stay for all patients was 4.2 (+/-3.6) days. The mean ICU stay was significantly longer in those with DKA grade III or IV, although the total hospital stay did not differ by severity of illness score.

CONCLUSIONS: One third of the patients admitted to our MICU to receive treatment for DKA did not warrant ICU treatment based on the admission criteria. These individuals had an approximate MICU length of stay of 1 day. A prospective study of the severity of illness score will be undertaken to evaluate the safety, validity, and potential resource savings of applying these DKA ICU admission criteria within our institution.