(524) Antibiotic Prescribing Patterns in a Medical Critical Care Population at an Academic Medical Center

Introduction: Antimicrobial stewardship in the critically ill population is challenging due to high acuity, rapid progression of illness, and limited data to guide therapy. Still, stewardship is crucial given concerns for resistance, side effects, and cost. There are currently no descriptions of antibiotic prescribing patterns in critically ill patients to aid in identification of potential stewardship targets. The purpose of this report was to characterize antibiotic prescribing patterns within the medical intensive care unit (MICU) at an academic medical center.

Methods: This retrospective review included adult patients who received antibiotics during MICU admission from February to March 2023. Data points collected included admission diagnosis, antibiotic selection and duration, culture data, mortality, and number of antibiotic changes. Descriptive statistics were used to identify patterns in ICU antibiotic usage. Chi-squared was used for nominal data, and student’s t-test for continuous data.

Results: A total of 250 patients were included with a median age of 60. Median (IQR) number of ICU antibiotic days was 4 (3-7), and the median number of antibiotics received was 3 (2-4). The most common empiric regimen was cefepime plus vancomycin (32.4%). A total of 108 patients (43.2%) had positive cultures . Therapy escalations occurred in 115 patients (46%) but only 65 (56.5%) of these patients had positive cultures. At least one de-escalation occurred in 221 patients (88.4%). Resistant bacteria were cultured in 26 patients (10.4%) and were associated with increased changes in antibiotic therapy (4.3 vs. 2.5, p=0.02), as were infectious disease (ID) consults (3.9 vs. 3.1, p=0.02). In hospital mortality was 26.4%. Patients who died had a higher likelihood of antibiotic escalation (OR 2.7, p=0.001), ID consult (OR 7.3, p< 0.001), and immunocompromised status (OR 2.4, p=0.003).

Conclusions: Antibiotic prescribing trends overall favored de-escalation more than escalation; however, some regimens were escalated despite lack of culture data. ID consults, resistant organisms, and positive cultures were associated with more changes in antibiotics. Further investigation into the factors that affected decisions to escalate therapy would be useful to assess appropriateness of therapy changes.