Biweekly Case Audit with Concurrent Intensivist Feedback to Improve Antimicrobial Use in a Medical Intensive Care Unit

Background: How best to optimize antimicrobial (AM) use in intensive care units is unclear.  As part of a multimodal intervention to improve AM use in the medical intensive care unit (MICU) of our urban teaching hospital, we performed regular audits of AM recipients and concurrently discussed recommendations for improvement with attending intensivists.

Methods: We implemented infection management guidelines for common ICU infections, promoted daily formal review of AM indications and infection status during team rounds, and gave monthly teaching sessions for rotating MICU trainees.   An attending and fellow ID physician reviewed the records of current MICU AM recipients, with guideline-based recommendations for improvement discussed with the two rotating attending intensivists in person on the first Friday of the intensivists’ two-week rotations.  Demographic, clinical, microbiologic and AM-related data were collected. 

Results: We audited 99 ICU AM recipients (61 men, median age 59, median 2 major comorbidities each) during 10 audits conducted over 6 months.   A median of 2 AMs were given per patient for a median duration of 4 days; piperacillin-tazobactam (36 patients), IV vancomycin (35) and metronidazole (16) were used most often.  Pneumonia (50 patients) and C difficile(16 empiric, 6 confirmed) were the commonest indications; infections were community-acquired in 72; microbial etiology was confirmed in 34; 36 patients were followed separately by ID consultants.  Clarifications for AM indications or duration were requested 26 times.   Changes in AM regimens were recommended 42 times: de-escalating coverage, AM discontinuation (d/c) and broadening coverage were recommended 16, 15 and 11 times, respectively; recommendations were accepted 23 (54%) times.  Audit discussions were uniformly collegial but were often hurried. 

Conclusion: Our audits led to AM de-escalation or d/c recommendations for 30% of patients but were only accepted half the time.  Ambiguity regarding AM indications and duration were common audit barriers.  The impact of this and associated interventions on antimicrobial use and resistance require further study.