The Utility of Discharge Antibiotics in Pediatric Perforated Appendicitis Without Leukocytosis.

Abstract

Optimal management of pediatric perforated appendicitis remains a topic of active investigation. Our institutional clinical practice guidelines (CPGs) were modified to discontinue antibiotics on discharge for patients with normal white blood cell count (WBC) without left shift. We hypothesized that patients would receive fewer antibiotics without increased complications. Patients <18y old with perforated appendicitis who underwent laparoscopic appendectomy between November 1, 2016 and May 31, 2021 at a tertiary care children's hospital were included. Primary outcome was adverse events: postdischarge surgical site infection (SSI), 30-day emergency department (ED) visits, or readmissions. Outcomes were compared before and after CPG modification. Multivariable regression was performed to identify factors associated with SSI. There were 113 patients pre- and 97 patients post-CPG modification. 23.1% of patients in the pre-cohort had an elevated discharge WBC or left shift compared to 18.9% of patients in the post-cohort (P=0.48). Significantly fewer patients were prescribed antibiotics on discharge in the post-cohort (70.8% pre versus 14.4% post, P<0.0001) and for fewer days (2 pre versus 0 post, P<0.0001). Total antibiotic days decreased significantly (6.1 pre versus 4.6 post, P<0.0001). There was an increase in postdischarge SSIs on univariate analysis (1.8% pre versus 9.3% post, P=0.03), ED visits (9.7% pre versus 19.6% post, P=0.04), and readmissions (5.3% pre versus 11.3% post, P=0.13). On multivariable analysis, being in the post-cohort was not significantly associated with post-discharge SSIs after adjusting for sex, symptom duration, initial WBC, and discharge antibiotic duration (OR 0.25, 95% CI 0.04-1.4, P=0.11). Modification of a pediatric perforated appendicitis clinical practice guideline to discontinue antibiotics on discharge with a normal WBC without left shift was effective in decreasing antibiotic duration. This was associated with an increase in SSIs on univariate analysis, which did not persist on multivariable analysis and requires further investigation.