Effectiveness of Antimicrobial Use and Diagnostic Testing in the Postoperative Setting for Colectomy, Craniectomy and Knee Arthroplasty

Abstract Background Patients with postoperative fevers tend to receive empiric antibiotics and a diagnostic workup even though postoperative fevers are frequently physiologic. The yield of postoperative workups ranges from 2–24%. Studying the effectiveness of antibiotic use and diagnostic testing across surgeries may provide insight to improve value of care for the postoperative patient. Methods We evaluated all patients who had elective colectomy, craniectomy and knee arthroplasty (KA) at NorthShore University HealthSystem between October 2009 and December 2013. Inpatient clinical data were extracted from the electronic data warehouse. All complications were confirmed by chart review. Antimicrobial Effectiveness was defined as the use of antibiotics only when patient had a confirmed infectious complication (pneumonia, urinary tract infection, etc) during hospital course. Diagnostic Effectiveness was defined as the use of diagnostic test only when patient had a confirmed complication. Length-of-stay (LOS) and cost were compared using Student’s t-test. Results 233 patients had a colectomy, 506 had a craniectomy and 4,693 had a knee arthroplasty. Antibiotic and Workup Effectiveness rates were 51% and 45.33% for colectomy, 18% and 22.53% for craniectomy, and 27%and 9.89% for KA, respectively. Effectiveness of antimicrobial use and diagnostic workups are shown in Figure 1. Patients without complications in all 3 surgeries who had a workup, as compared with those patients without a workup, were more likely to have a longer LOS (colectomy: 8.95 days vs. 5.08 days, P < 0.0001; craniectomy: 4.88 days vs. 2.62 days, P < 0.001; KA: 3.22 days vs. 2.95 days, P < 0.0001) and a higher cost of hospitalization (colectomy: $76,034 vs. $42,865, P < 0.001; craniectomy: $83,779 vs. $58,998, P < 0.001; KA: $53,451 vs. $52,159, P < 0.001). Conclusion A significant fraction of empiric antibiotics in the postoperative setting are unnecessary across multiple surgeries. Diagnostic testing effectiveness rates are low and have consequences in terms of healthcare utilization. Disclosures All authors: No reported disclosures.