Outcomes of initial medical compared with surgical strategies in the management of intra-abdominal abscesses in patients with Crohn's disease: a meta-analysis

Background Several studies have also evaluated the efficacy of initial medical management compared with initial surgical management strategies with regard to abscess resolution with variable results. Objective The aim of this study is to evaluate the efficacy of initial medical management compared with surgical management of Crohn's disease (CD)-related intra-abdominal abscesses. Data sources A comprehensive search of multiple databases (MEDLINE/PubMed, Cochrane databases, CINAHL, Scopus, and Google Scholar) was performed in August 2014. Study selection All studies on adults comparing initial surgical versus medical approaches to treat CD-related abscesses were included. Main outcomes measured The durability of abscess resolution and rate of stoma creation between the groups undergoing initial surgical versus medical approaches were compared. Results The pooled analysis of the nine studies including a total 603 patients showed an overall rate of abscess resolution were 56.6% in the medical group compared with 80.7% in the surgical group. There was over three-fold higher chance of achieving abscess resolution when an initial surgical strategy was used at the time of abscess diagnosis compared with the medical strategy (odds ratio 3.44, 95% confidence interval: 1.80, 6.58, P<0.001). The number needed to treat using the initial surgical approach to prevent a recurrent abscess was four patients. Limitations All included studies were retrospective case series with potential clinical confounders not fully accounted in the analysis. Conclusion Initial surgical management appears to be superior to medical management in patients with CD-related intra-abdominal abscesses. Though all the included studies in this meta-analysis were retrospective, this meta-analysis is likely the strongest level of evidence with regard to the management of CD-related abscesses, given that a randomized-control trial may not be feasible given the low rate of abscess development in CD. Copyright (C) 2015 Wolters Kluwer Health, Inc. All rights reserved.