Introduction: Intensive Care Units (ICUs) are among the most expensive components of hospital care. Experts believe that ICUs are overused; however, hospitals vary in their ICU admission rate. Our hypothesis is based on clinical observations that many patients with diabetic ketoacidosis (DKA), stroke, and gastrointestinal (GI) bleeding admitted to the ICU don't really need it and could be managed safely in a non-ICU level of care. Reducing inappropriate admissions would reduce healthcare costs and improve outcomes. Our primary objective was to determine the frequency of inappropriate ICU admissions. Secondary objectives were to evaluate which diagnoses were more unnecessarily admitted to the ICU, evaluate different variables and comorbidities, and determine the mortality rates during ICU admissions.
Methods: Patients admitted to the ICU, from the Emergency Department (ED) or transferred from the floor, during a one-year period were evaluated in this retrospective study. Patients 18-years old and above who had an admitting diagnosis of DKA, GI bleed, ischemic stroke, or hemorrhagic stroke were included. Patients in a comatose state, intubated, on vasopressors, hemodynamically unstable or had an unstable comorbid disease, subarachnoid hemorrhage, surgery during hospitalization prior to the ICU admission were excluded. Patients were categorized as having an appropriate or inappropriate ICU admission based on our institutional ICU admission criteria and data from available literature and guidelines.
Results: A total of 95 patients were included in our cohort. Seventy-two out of 95 (76%) were considered as inappropriate ICU admissions. When comparing each of the four admitting diagnoses, a significantly higher proportion of DKA patients were considered inappropriate ICU admissions when compared to the other diagnoses (P = 0.001). The overall mortality rate of ICU admissions was 16%, 15 patients out of 95 study population. When comparing each of the four admitting diagnoses, there was a significant difference in mortality rate with DKA having the lowest mortality (3%) and GI bleed having the highest mortality (43%). Out of the 15 patients who died, only 1 patient was categorized as an inappropriate ICU admission.
Conclusions: More than three-quarters of our study population was admitted to the ICU inappropriately. Incorporating severity scores in ICU admission criteria could improve the appropriateness of ICU admission and financial feasibility. This article is based on a poster: Alsamman S, Alsamman MA, Castro M, Koselka H, Steinbrunner J: ICU admission patterns in patients with DKA, stroke and GI bleed: do they all need ICU? J Hosp Med. March 2015.
Keywords: health care cost; icu admission criteria; icu stay.