Current treatment and outcomes of esophageal perforations in adults: meta-analysis and meta-regression of 75 studies.

BACKGROUND: The current prognosis of esophageal perforation and the efficacy of available treatment methods are not well defined. METHODS: We performed a systematic review of esophageal perforations published from January 2000 to April 2012 and subjected a proportion of the retrieved data to a meta-analysis. Meta-regression was performed to determine predictors of mortality immediately after esophageal perforation. RESULTS: Analysis of 75 studies resulted in a pooled mortality of 11.9 % [95 % confidence interval (CI) 9.7-14.3: 75 studies with 2,971 patients] with a mean hospital stay of 32.9 days (95 % CI 16.9-48.9: 28 studies with 1,233 patients). Cervical perforations had a pooled mortality of 5.9 %, thoracic perforations 10.9 %, and intraabdominal perforations 13.2 %. Mortality after esophageal perforation secondary to foreign bodies was 2.1 %, iatrogenic perforation 13.2 %, and spontaneous perforation 14.8 %. Treatment started within 24 h after the event resulted in a mortality rate of 7.4 % compared with 20.3 % in patients treated later (risk ratio 2.279, 95 % CI 1.632-3.182). Primary repair was associated with a pooled mortality of 9.5 %, esophagectomy 13.8 %, T-tube or any other tube repair 20.0 %, and stent-grafting 7.3 %. CONCLUSIONS: Results of recent studies indicate that mortality after esophageal perforation is high despite any definitive surgical or conservative strategy. Stent-grafting is associated with somewhat lower mortality rates, but studies may be biased by patient selection and limited experience.