Impact of a randomized clinical trial on children with perforated appendicitis

Abstract

We previously conducted a randomized, clinical trial comparing early appendectomy with interval appendectomy for perforated appendicitis. The purpose of the present study was to evaluate the effect this clinical trial had on subsequent practice patterns and outcomes for patients with perforated appendicitis at the free-standing children's hospital conducting the trial. A retrospective study was conducted comparing children with perforated appendicitis treated before the trial (2005-2006) and after the trial (2009-2011). Early appendectomy was performed within 24 hours of diagnosis; interval appendectomy occurred 4-6 weeks after initial treatment with antibiotics. Patient characteristics, treatment variables, and outcomes were collected and compared. The pretrial group consisted of 92 patients-62 (67%) underwent early appendectomy, and 30 (33%) patients had interval appendectomy. The posttrial group was composed of 103 patients, with 87 (84%) undergoing early appendectomy and 16 (16%) interval appendectomy (P = .005). The groups were similar in patient and admission characteristics, although the posttrial group had a lower percentage of self-pay patients and fewer computed tomography scans; health care use was similar between groups. Overall, the posttrial group had fewer adverse events (18% vs 34%, P = .02), specifically fewer wound infections (2% vs 14%, P = .001) and fewer unplanned readmissions (7% vs 16%, P = .04) than the pretrial group. In the posttrial group, those patients selected for interval appendectomy were more likely to complete the planned course of therapy than in the pretrial group. A clinical trial conducted at our institution to evaluate currently available treatment options for perforated appendicitis did change practice patterns at our hospital. After the trial, there was an increase in the use of early appendectomy, a decrease in the number of computed tomography scans performed per patient, and a reduction in the overall adverse event rate.