Reducing hospital morbidity and mortality following esophagectomy

Background

Esophagogastrectomy (EG) is a formidable operation with significant morbidity and mortality rates. Risk factor analyses have been performed, but few studies have produced strategies that have improved operative results. This study was performed in order to identify prognostic variables that might be used to develop a strategy for optimizing outcomes after EG.

Methods

The records of all patients (n = 379) who underwent EG patients at a tertiary medical center between 1996 and 2002 were retrospectively reviewed. Thirty-day morbidity and mortality were determined, and multivariable logistical regression analysis assessed the effect of preoperative and postoperative variables on early mortality.

Results

Operations included Ivor Lewis (n = 179), transhiatal (n = 130), and other approaches (n = 70). Operative mortality was 5.8%; 64% experienced complications, including respiratory complications (28.5%), anastamotic strictures (25%), and leak (14%). Increasing age, anastomotic leak, Charlson comorbidity index 3, worse swallowing scores, and pneumonia were associated with increased risk of mortality by univariate analysis. However, only age (p = 0.002) and pneumonia (p = 0.0008) were independently associated with mortality by multivariable analysis. Pneumonia was associated with a 20% incidence of death. Patients with pneumonia had significantly worse deglutition and anastomotic integrity on barium esophagogram compared with patients without pneumonia (p < 0.001, Mann-Whitney rank sum test).

Conclusions

Morbidity and mortality of EG are significant, but most complications, including anastomotic leak, are not independent predictors of mortality. The most important complication after EG is pneumonia. Strategies to decrease postoperative mortality should include careful assessment of swallowing abnormalities and predisposition to aspiration by cineradiography or fiberoptic endoscopy. After EG, acceptable pharyngeal function and airway protection should be verified before resuming oral intake.