Methods: A quasi-experimental study was conducted in adult patients (age >18 years) admitted to Nebraska Medicine for suspected AECOPD before and after clinical decision support was introduced. Patients in the pre-implementation period (10 weeks, 4/15-6/30/16, N=44) and a similar post-implementation period (10 weeks, 4/12-6/29/17, N=51) were included if COPD was the primary diagnosis code or the COPD exacerbation order set was used at admission. Exclusion criteria included AECOPD admission within the previous 30 days and transfer from an outside hospital. Outcome measures included: percentage of patients receiving antibiotics, median length of therapy, order set usage, antibiotic choices, length of stay (LOS) and oral steroid use.
Results: Post-implementation, the percentage of patients prescribed antibiotics decreased (86.4% vs 60.8%, p=0.006) as did antibiotics ordered from the order set (29.5% vs 13.7%). Median length of therapy decreased from 5 days to 1 day pre- versus post-implementation, respectively . Fluoroquinolone use decreased from 43.2% to 25.5% while azithromycin use remained consistent (18.2% vs 17.6%). Oral steroid use increased post-implementation (27.3% vs 41.2%) and average duration of steroid use decreased (11.1 vs 8.7 days). Average LOS was 3.7 days in both groups and in-hospital mortality was low (2% vs 0%).
Conclusion: Implementation of an AECOPD guidance document, order set with CDS, and education resulted in significant decreases in antibiotic usage, particularly for fluoroquinolones. Other areas of care also improved using a syndromic stewardship strategy. Our data supports the utilization of this strategy to promote evidence-based antibiotic management in AECOPD.
S. Bergman, None
T. Van Schooneveld, None