Analysis of colonoscopic perforations at a local clinic and a tertiary hospital.

AIM: To define the clinical characteristics, and to assess the management of colonoscopic complications at a local clinic.
METHODS: A retrospective review of the medical records was performed for the patients with iatrogenic colon perforations after endoscopy at a local clinic between April 2006 and December 2010. Data obtained from a tertiary hospital in the same region were also analyzed. The underlying conditions, clinical presentations, perforation locations, treatment types (operative or conservative) and outcome data for patients at the local clinic and the tertiary hospital were compared.
RESULTS: A total of 10  826 colonoscopies, and 2625 therapeutic procedures were performed at a local clinic and 32  148 colonoscopies, and 7787 therapeutic procedures were performed at the tertiary hospital. The clinic had no perforations during diagnostic colonoscopy and 8 (0.3%) perforations were determined to be related to therapeutic procedures. The perforation rates in each therapeutic procedure were 0.06% (1/1609) in polypectomy, 0.2% (2/885) in endoscopic mucosal resection (EMR), and 3.8% (5/131) in endoscopic submucosal dissection (ESD). Perforation rates for ESD were significantly higher than those for polypectomy or EMR (P < 0.01). All of these patients were treated conservatively. On the other hand, three (0.01%) perforation cases were observed among the 24  361 diagnostic procedures performed, and these cases were treated with surgery in a tertiary hospital. Six perforations occurred with therapeutic endoscopy (perforation rate, 0.08%; 1 per 1298 procedures). Perforation rates for specific procedure types were 0.02% (1 per 5500) for polypectomy, 0.17% (1 per 561) for EMR, 2.3% (1 per 43) for ESD in the tertiary hospital. There were no differences in the perforation rates for each therapeutic procedure between the clinic and the tertiary hospital. The incidence of iatrogenic perforation requiring surgical treatment was quite low in both the clinic and the tertiary hospital. No procedure-related mortalities occurred. Performing closure with endoscopic clipping reduced the C-reactive protein (CRP) titers. The mean maximum CRP titer was 2.9 ± 1.6 mg/dL with clipping and 9.7 ± 6.2 mg/dL without clipping, respectively (P < 0.05). An operation is indicated in the presence of a large perforation, and in the setting of generalized peritonitis or ongoing sepsis. Although we did not experience such case in the clinic, patients with large perforations should be immediately transferred to a tertiary hospital. Good relationships between local clinics and nearby tertiary hospitals should therefore be maintained.
CONCLUSION: It was therefore found to be possible to perform endoscopic treatment at a local clinic when sufficient back up was available at a nearby tertiary hospital.