O2185: Enhancing Antibiotic Stewardship: Increasing Discharge Smart Set Utilization in Pediatric Urgent Care for Skin Soft Tissue Infections, a Quality Improvement Initiative

Purpose/Objectives: In our large urban urgent care setting we see many patients with skin soft tissue infections (SSTI). We updated our SSTI clinical practice guideline and created a discharge smart set to reflect these changes in EMR, thus prescribing a shorter course of cephalexin to treat cellulitis per national guidelines. We tracked provider data to monitor the change in prescribing habits. However, we noticed that as our baseline practice group ordered the correct antibiotic with correct duration 60% of the time with some seasonal variability. A priori education alone was not enough to improve antibiotic stewardship. Our QI initiative set out to improve a process outcome measure to facilitate antibiotic stewardship with easier workflow and less guesswork for providers.
Our goal was to increase discharge smart set utilization from baseline 47% to 58% for all the providers within six months. Our global aim was to increase appropriate choice of antibiotic and duration to 80%. Our balancing measure was a return visit within 72 hours.

Design/Methods: Initially, a survey was sent to all the providers to gauge utilization of discharge smart sets and identify areas for improvement. We created a Process Map and Key Driver Diagram. First intervention was provider education and guideline review. Next, we created laminated cards and placed them at all provider workstations as a reminder. We provided individual feedback to providers with the lowest utilization of smart sets. We utilized PDSA cycles to determine which interventions to adapt, adopt and abandon.
We used QlikView Accesspoint to obtain initial baseline weekly data for all the providers who ordered the correct antibiotic and length of therapy and used the discharge smart set. We used a p-chart in QI Macros to plot both the data for number of smart sets used with correct antibiotic prescription and correct antibiotic use and duration over time.

Results: Per graph below, we saw variable improvement with Intervention # 1(education), Intervention #2 (laminated card) and Intervention #3 (individual feedback). The central line did not shift but we did have some special cause variation.

Conclusion/Discussion: There is seasonal variability in the diagnosis of cellulitis as reflected by central line shifts in both graphs over the summer months. Very small sample sizes in the winter months create variation which makes it challenging to demonstrate improvement with our current interventions. We plan to continue the project for 12 months to account for seasonal variability and opportunity to include additional interventions including expanding our trigger list for discharge smart sets based on chief complaint and creating a Best Practice Advisory (BPA) alert as an additional reminder.