Abstract
OBJECTIVE: A subgroup of patients with intractable constipation has persistent dilatation of the bowel, which in the absence of an organic cause is termed idiopathic megabowel (IMB). The aim of this systematic review was to evaluate the published outcome data of surgical procedures for IMB in adults. METHODS: Electronic searches of the MEDLINE (PubMed) database, Cochrane Library, EMBase, and Science Citation Index were performed. Only peer-reviewed articles of surgery for IMB published in the English language were evaluated. Studies of all surgical procedures were included, providing they were performed on 3 or more patients, and overall success rates were documented. Studies were critically appraised in terms of design and methodology, inclusion criteria, success, mortality and morbidity rates, and functional outcomes. RESULTS: A total of 27 suitable studies were identified, all evidence was low quality obtained from case series, and there were no comparative studies. The studies involved small numbers of patients (median 12, range 3-50), without long-term follow-up (median 3 years, range 0.5-7). Inclusion of subjects, methods of data acquisition, and reporting of outcomes were extremely variable. Subtotal colectomy was successful in 71.1% (0%-100%) but was associated with significant morbidity related to bowel obstruction (14.5%, range 0%-29%). Segmental resection was successful in 48.4% (12.5%-100%), and recurrent symptoms were common (23.8%). Rectal procedures achieved a successful outcome in 71% to 87% of patients. Proctectomy, the Duhamel, and pull-through procedures were associated with significant mortality (3%-25%) and morbidity (6%-29%). Vertical reduction rectoplasty (VRR) offered promising short-term success (83%). Pelvic-floor procedures were associated with poor outcomes. A stoma provided a safe alternative but was only effective in 65% of cases. CONCLUSIONS: Outcome data of surgery for IMB must be interpreted with extreme caution due to limitations of included studies. Recommendations based on firm evidence cannot be given, although colectomy appears to be the optimum procedure in patients with a nondilated rectum, restorative proctocolectomy the most suitable in those with dilatation of the colon and rectum, and VRR in those patients with dilatation confined to the rectum. Appropriately designed studies are required to make valid comparisons of the different procedures available.