Methods: A single center, quasi-experimental retrospective cohort analysis comparing oral antibiotics prescribed at hospital discharge before (July 2012 - June 2013) and after (October 2014 - February 2015) an intervention consisting of provider education and real-time pharmacy audit and feedback of discharge prescriptions. The main outcome measures were total duration of therapy, use of antibiotics with broad gram-negative activity (fluoroquinolones or amoxicillin-clavulanate), and appropriateness of discharge prescriptions, as adjudicated by a blinded panel of infectious disease (ID) experts on a subset of randomly selected cases.
Results: 300 cases were included in the baseline cohort and 200 cases in the intervention cohort. Compared with the baseline period, use of broad gram-negative antibiotics decreased during the intervention period (51% vs 40%, p<0.05) and use of azithromycin increased (12% vs 18%, p<0.05). There was a significant decrease in median duration of therapy prescribed at discharge from 6 days (interquartile range [IQR] 4-10) to 5 days (IQR 3-7) (p<0.05); however, reduction in total duration did not reach significance (10 days [IQR 7-13] to 9 days [IQR 6-13], p>0.05). In subgroup analyses, community-acquired pneumonia (CAP) and skin infections each had a reduction in use of broad gram-negative agents (60% to 31%, p=0.01 and 34% to 15%, p=0.04, respectively) and total duration of therapy (8 days [IQR 6-9] to 6 days [IQR 5-7] and 12 days [IQR 8-15] to 9 days [7-12], respectively; p<0.05). In the blinded review by ID experts, antibiotic selection was more frequently classified as appropriate during the intervention period (72% vs 90%, p=0.02).
Conclusion: A multifaceted intervention to optimize antibiotic prescriptions at hospital discharge led to a decrease in use of antibiotics with broad gram-negative activity and shorter post-hospital treatment durations. The intervention was particularly effective in cases of CAP and skin infections.
K. Shihadeh, None
S. Calcaterra, None
B. Knepper, None
W. Burman, None
P. Mehler, None
T. Jenkins, None
See more of: Oral Abstract Session