Factors associated with failure of nonoperative treatment of complicated appendicitis in children

Abstract

Appendicitis remains the most common cause for emergency abdominal surgery in children. Immediate appendectomy in complicated, perforated appendicitis can be hazardous and nonoperative therapy has been gaining use as an initial therapy in children. Previous studies have reported failure rates in nonoperative therapy in such cases ranging from 10% to 41%. Factors leading to treatment failures have been studied with various and disparate results. We reviewed our institutional experience in treated complicated appendicitis, with focus on those initially managed nonoperatively. MethodsRecords of all children admitted with the diagnosis of perforated appendicitis to NYU Langone Medical Center and Bellevue Hospital Center from January 1, 2003 to December 31, 2013 were reviewed. The diagnosis was made with ultrasound and/or computed tomography scan. Those with abscesses amenable to drainage underwent aspiration and drain placement by an interventional radiologist. Broad spectrum intravenous (IV) antibiotics were given until the patient became afebrile, pain free and tolerating a regular diet. Oral antibiotics were continued for an additional week and interval appendectomy was done eight weeks later. The primary outcome measure was treatment response with failure defined as those who did not improve or required readmission for additional IV antibiotics and/or early appendectomy. Multiple patient and treatment related variables, including those previously reported as predicting failure in nonoperative therapy, were studied. Continuous variables were reported as means ± standard error and compared using 2-tailed unpaired t tests; nonparametric variables were analyzed by Mann–Whitney U tests. Categorical variables were reported as medians ± interquartile ranges and compared using Chi-square testing. Statistical significance was accepted for p<.05. ResultsSixty-four patients were identified as undergoing initial nonoperative therapy. Fifty-two (81%) were categorized as treatment successes being treated nonoperatively and 12 (19%) were failures. Variables showing no significance in predicting treatment failures included duration of symptoms, presence of appendicolith, presence of phlegmon, presence of abscess, initial white blood cell count, and SIRS (Systemic Inflammatory Response Syndrome) positive. The variables that predicted failure of nonoperative therapy vs. successes were presence of bandemia (75% vs. 40%, p=0.052) and small bowel obstruction on imaging (42% vs. 15%, p=0.052) and presence of bandemia ≥15% which was highly predictive of failure (67% vs. 4%, p<0.01). ConclusionsPredicting which patients with complicated perforated appendicitis will respond well to nonoperative therapy may allow us to more effectively treat patients with complicated perforated appendicitis. In our study the presence of small bowel obstruction and bandemia, especially ≥15% correlated with treatment failure; this suggests that these select patients may need a modified treatment strategy.