Current results of therapy for esophageal perforation.

BACKGROUND: Prior reviews of esophageal perforation with delayed recognition have reported mortality rates as high as 66%. We performed a retrospective review of patients with nonmalignant esophageal perforation to assess the outcome of current management techniques.

PATIENTS AND METHODS: Charts were reviewed of all patients who were treated for nonmalignant esophageal perforation between 1980 and 1993. They were 23 men and 10 women, mean age 49 +/- 3 years, 19 of whom were diagnosed early (< or = 24 hours) and 14 of whom were diagnosed late (> 24 hours).

RESULTS: Perforations were due to instrumentation (16), operative injury (7), spontaneous rupture (4), trauma (4), and other causes (2). Pre-existing esophageal disease was identified in 23 patients (70%), including achalasia (9), stricture (7), varices (5), and other (2). Treatment included closure and fundoplication or muscle wrap (10), closure with or without pleural flap (7), resection only (7), resection and reconstruction (3), drainage only (4), and observation (2). Nonfatal complications included empyema (4), arrhythmia (3), persistent leak following attempted closure (2), and other (5). They occurred in 50% of both the early and late diagnosis groups and were of comparable severity in both. The overall mortality was 9% (3/33). Causes of death were sepsis (1) and multisystem organ failure (2). Mortality was 5% (1/19) in patients diagnosed early and 14% (2/14) of those diagnosed late.

CONCLUSIONS: Current mortality rates in nonmalignant esophageal perforation are improved compared to previously published rates of 19% for all patients with the condition, 9% following early and 29% following late diagnosis (47 patients overall). We conclude that, despite a high incidence of associated complications, the survival rate following nonmalignant esophageal perforation is improving and the impact of delayed recognition is decreasing.