232 - Impact of Weekend Initiation of Vancomycin or Piperacillin/Tazobactam on Days of Therapy Received upon Hospital Admission

Background: Antibiotic therapy for inpatients with suspected infections is typically empirically initiated and therapy narrowed or altered when additional diagnostic evidence becomes available.  For patients whose therapy is initiated on a weekend, differences in hospital staffing may impact the timing of therapy changes.  We aimed to compare the duration of therapy of vancomycin and piperacillin-tazobactam between those who had therapy initiated on a weekday versus a weekend day.

Methods: We performed a cross-sectional study among U.S. hospitals that contributed pharmacy data for inpatients to the Vizient clinical database in 2016.  We identified vancomycin and piperacillin-tazobactam courses initiated within the first 48 hours of admission; courses were categorized as weekend initiation (Friday, Saturday, Sunday) versus weekday initiation. The median days of therapy were compared between weekend and weekday initiation using the Wilcoxon rank-sum test.

Results: Among the 145 hospitals representing approximately 3.7 million patient encounters there were 401,101 encounters with vancomycin and 221,751 with piperacillin/tazobactam initiated within the first 48 hours of admission. Of these courses, 33% of vancomycin and 40% of piperacillin/tazobactam were initiated on a weekend day. The median (IQR) days of therapy for vancomycin initiated on a weekend was 2 days (1-4 days) compared to 2 days (1-3 days) when initiated on a weekday (p< .01). The median (IQR) days of therapy for piperacillin/tazobactam was 3 days (2-5 days) for courses initiated on either a weekend or weekday (p< .01).

Conclusion: We observed a statistically significant difference in the days of therapy received by patient encounters with vancomycin or piperacillin/tazobactam initiated on weekdays versus weekends. However, because of the large sample size in this study, we had power to identify small differences as statistically significant. Still, for vancomycin the 75^th percentile received at least one additional day of therapy when initiated on a weekend versus a weekday. Further exploration is needed to identify if weekend initiation is associated with extended durations of therapy in specific sub-populations of patients.