Improving surgical antibiotic prophylaxis timing at a small community hospital

Abstract

INTRODUCTION: In November 2003, a six sigma quality improvement team was initiated to improve surgical prophylactic antibiotic timing from the yearly rate of 71% at this 138-bed community hospital located in southwestern Illinois and a member of a broader healthcare system. METHODS: From January 2003 to August 2004, data were prospectively collected for hip and knee arthroplasty, colon resection, and hysterectomy (n=393). Centers for Medicare and Medicaid Services (CMS) definitions were used for antibiotic on-time delivery (<60 minutes before the incision). The other CMS populations, heart and vascular surgeries, are not performed at the hospital. From December 2003 through February 2004, the team collected and analyzed data on antibiotic delivery, flowcharted current processes, developed cause-and-effect diagrams of untimely antibiotic delivery, and trialed new processes. In March 2004, a new practice of antibiotic intravenous push (IVP) on anesthesia induction was initiated. In June 2004, standing orders for automatic substitution to appropriate prophylactic antibiotic choice and IVP were approved. Analysis was performed using Pearson's chi square tests and Mantel-Haenszel's common odds ratio estimates. RESULTS: Prior to the IVP process, 199/277 (72%) cases received antibiotics on time, 50 (19%) were early (>60 minutes before incision), 19 (7%) were after the incision, and 9 (3%) were not given. After IVP, 98/116 (84%) received on-time antibiotics, 10 (9%) were early, 5 (4%) were late, and 3 (3%) were not given, results which were significantly better (p=0.009). Before the IVP process, the mean time of antibiotic delivery was 33 minutes preoperatively (SD=120 minutes); after IVP the mean was 22 minutes (SD=67 minutes). Knee and hip arthoplasty patients were 2.7 times more likely (CI=1.4–5.1) than other surgical patients to receive antibiotics on time prior to IVP. After IVP, there was no difference in timeliness between the surgical procedures. The IVP process was identified when evaluating the data of the surgeons with timely delivery. Since automatic substitution began in June 2004, on-time delivery has been 48/53 (90%) (p=0.006). CONCLUSION: The six sigma team improved the timeliness of prophylactic antibiotics at the hospital. IVP on induction with automatic substitution were successful interventions. Current evaluation of each untimely case helps identify process failures. Continued efforts are needed to achieve 100% compliance.