Antibiotic Stewardship in the Medical Intensive Care Unit of an Academic Medical Center: Impact of a Pneumonia Diagnostic Bundle with Pharmacist Intervention

BACKGROUND: Acute bacterial pneumonia is a common empiric diagnosis in medical intensive care unit (MICU) patients. Clinically, however, it may be difficult to distinguish from nonbacterial causes of inflammation and infection of the lung(s). Incomplete diagnostic workup at the time of empiric antibiotic initiation or misinterpretation of available data may impede antibiotic de-escalation and discontinuation. We aimed to reduce unnecessary antibiotic use in the MICU by (1.) bundling pneumonia diagnostic orders into a single comprehensive order set and (2.) by implementing a daily pharmacist-driven antibiotic time-out. METHODS: This before-and-after quality improvement pilot project was conducted in the MICU of Baystate Medical Center, a closed 16-bed unit, from December 2016 through March 2017. Outcomes were compared with a baseline period from December 2015 through March 2016. At baseline, all diagnostic orders were entered individually via computer physician order entry (CPOE) and daily antibiotic stewardship was not provided. For the pilot, a pneumonia order set was built which includes all diagnostic tests and recommended empiric antibiotics based on the local antibiogram. Of note, serial procalcitonin levels first became available at our institution through this order set. An interpretation algorithm was adapted from the literature to aid in their interpretation. A new MICU clinical pharmacist position was created which allowed antibiotic time-outs to be conducted 7 days per week. Antibiotic discontinuation was assessed by comparing days of antibiotic therapy per 1000 patient-days. RESULTS: For all antibiotics used to treat bacterial pneumonia, total days of therapy per 1000 patient-days in the MICU decreased from 905.7 in the baseline period to 688.4 in the pilot period (rate difference -217.3, 95% CI -270.8 to -163.9). The usage of narrow spectrum antibiotics increased during the pilot period. CONCLUSION: Bundling pneumonia diagnostic orders together into a single order set inclusive of serial procalcitonin measurement as well as providing daily pharmacist-led antibiotic time-outs were associated with decreased antibiotic usage in the MICU. DISCLOSURES: All authors: No reported disclosures.