To determine the diagnostic yield and safety of abdominal computerized tomography (CT) in the medical intensive care unit (MICU).
This retrospective study included 95 MICU patients who received CT for suspected infection. We collected demographics, symptoms, APACHE III score and predicted mortality, selected laboratory data and mortality. CT related findings, their impact on patient management and CT related complications were noted.
Mean patient age 64.4 years. Mean admission and CT day APACHE III scores were 82 and 75 with predicted mortality rates of 38.5% and 45%, respectively. Fever was present in 34%, abdominal pain in 59%, nausea/vomiting in 8% and diarrhea 30%. Leukocyte count was elevated in 67%, alkaline phosphatase in 84%, total bilirubin in 57% and lipase in 22%. The main CT findings included colitis 8, gall bladder disease 4, diverticulitis 3, incarcerated hernia 3, pancreatitis 2, bowel perforation 2, bladder distention/obstruction 2. There were management changes based on the CT in 16%. No airway or vascular access loss was documented. The mean PaO2 to FIO2 ratio was 267 and 242, before and after CT respectively (p = 0.08). The serum creatinine increased over a 3-day period by more than 2 mg/dL in 1 patient and between 1-2 mg/dL in 3 patients. The hospital mortality rate was 35%: 33% in patients who had management change compared to 36% in the others (p=0.8492). There was no independent association between management change and hospital outcome when adjustment was made for APACHE III predicted mortality.
Abdominal CT changed patient management in 16%. Management change did not have significant impact on hospital mortality. CT scanning was not associated with any significant adverse patient outcomes including vascular access loss, airway loss, post contrast renal dysfunction or change in arterial oxygenation.
Abdominal CT in selected ICU patients is a safe diagnostic procedure. However, it leads to management changes only in the minority of patients. When it leads to management changes mortality outcomes are not significantly effected.
A.J. Stiehm, None.