Antibiotic treatment versus appendicectomy for acute appendicitis in adults: an individual patient data meta-analysis

Abstract

Background: Randomised controlled trials (RCTs) have found antibiotics to be a feasible and safe alternative to appendicectomy in adults with imaging-confirmed acute appendicitis. However, patient inclusion criteria and outcome definitions vary greatly between RCTs. We aimed to compare antibiotics with appendicectomy for the treatment of acute appendicitis using individual patient data and uniform outcome definitions.

Methods: In this individual patient data meta-analysis, we searched PubMed, Embase, and the Cochrane Central Register of Controlled Trials without language restrictions between database inception and June 6, 2023, for RCTs comparing appendicectomy with antibiotics for the treatment of adults (≥18 years) with imaging-confirmed acute appendicitis. Studies without 1-year follow-up data on complications were excluded, as were patients. Corresponding authors of eligible studies were contacted and invited to share data; individual patient data were merged after validation. One-stage meta-analyses were conducted using a generalised, mixed-effects linear regression model, accounting for clustering of patients within studies. The primary outcome was the complication rate at 1-year follow-up, uniformly harmonised across trials using the Clavien-Dindo classification. Complications were further divided into minor (grade 1-2 or equivalent) and major (grade 3-5 or equivalent) complications. Appendicectomy rate during 1 year was a key secondary outcome but not considered a complication for the antibiotics group. Outcomes were described separately for patients with and without an appendicolith. This study is registered with PROSPERO, CRD42023391676.

Findings: Of 887 potentially relevant articles, eight were eligible for inclusion, of which six RCTs could provide data for 2101 eligible patients (1050 assigned to antibiotics and 1051 assigned to appendicectomy; 830 [39·5%] women and 1271 [60·5%] men). All studies raised some bias concerns due to absence of blinding. One study was judged to have a high risk of bias due to the exclusion of eligible patients after randomisation, but these patients were eligible for inclusion in our meta-analysis. At 1 year, 57 (5·4%) of 1050 patients randomly assigned to antibiotics had a complication compared with 87 (8·3%) of 1051 patients randomly assigned to appendicectomy (odds ratio [OR] 0·49 [95% CI 0·20 to 1·20]; risk difference -4·5 percentage points [95% CI -11·6 to 2·6]). At 1 year, 1025 (97·5%) patients in the appendicectomy group had undergone appendicectomy compared with 356 (33·9%) patients in the antibiotics group. In patients with an appendicolith at pre-interventional imaging, there were more complications at 1 year among patients who received antibiotic treatment than among those who underwent appendicectomy (29 [15·0%] of 193 patients vs 12 [6·3%] of 190 patients; OR 2·82 [95% CI 1·11 to 7·18]; risk difference 13·2 percentage points [95% CI 2·3 to 24·2]). In the antibiotics group, 94 (48·7%) of 193 patients with an appendicolith underwent appendicectomy within 1 year versus 262 (30·6%) of 857 patients without an appendicolith.

Interpretation: This meta-analysis showed that antibiotic treatment in adults with imaging-confirmed acute appendicitis was a safe alternative to surgery and resulted in around two-thirds of patients avoiding appendicectomy during the first year. In patients with an appendicolith, initial antibiotic treatment increased the risk of complications compared with appendicectomy, and around half of these patients assigned to antibiotics underwent step-up appendicectomy within 1 year. These data should be key components in shared decision making.

Funding: None.