Prediction of Anastomotic Leak and its Prognosis in Digestive Surgery

BACKGROUND: Anastomotic leak (AL) is a dangerous postoperative complication in gastrointestinal surgery. The present study focuses on whether our prediction scoring system, "Estimation of Physiologic Ability and Surgical Stress" (E-PASS), could predict occurrence of AL and its prognosis in various kinds of gastrointestinal surgical procedures. METHODS: We prospectively investigated parameters of E-PASS, absence or presence of AL, and in-hospital mortality in 6,005 patients who underwent elective digestive surgery with alimentary tract reconstruction in 45 acute care hospitals in Japan between 1 April 2002 and 31 March 2007. RESULTS: Incidences of AL were 19.6% for esophagectomy via right thoracotomy and laparotomy, 11.7% for pancreaticoduodenectomy, 7.4% for low anterior resection, 4.0% for total gastrectomy, 1.8% for open distal gastrectomy, 1.3% for open colectomy, for an overall incidence of 4.1%. The incidence in each procedure significantly correlated with median value of surgical stress score of the E-PASS (R = 0.78, n = 11, p = 0.0048). The incidences of AL increased when Total Risk Points (TRP) of the E-PASS increased; 1.1% at the TRP range of <500, 2.8% at 500 to <1,000, 4.8% at 1,000 to <1,500, and 13.6% at ≥ 1,500 (p < 0.0001). In patients who suffered from AL, an in-hospital mortality rate at TRP < 1,000 was significantly lower than that at TRP of ≥ 1,000 (1.1 vs. 15.9%; p = 0.00019). CONCLUSIONS: The E-PASS, requiring only nine variables, may be useful in predicting AL and its prognosis.