321. ANASTOMOTIC LEAK WITH ENHANCED RECOVERY PROTOCOL VERSUS CONVENTIONAL CARE AFTER ESOPHAGECTOMY FOR CANCER: CLINICAL CHARACTERISTICS AND OUTCOMES

Abstract Objective The anastomotic leak (AL) is greatly concerned in the clinical pathway of enhanced recovery after surgery (ERAS) with esophagectomy. This retrospective study was to compare the clinical characteristics and outcomes of AL between the ERAS protocol and conventional care in patients undergoing esophagectomy for cancer. The strategy for improving the safety of ERAS was explored. Methods The clinical data and outcomes were retrospectively analyzed in patients with thoracic esophageal carcinoma who received esophagectomy with esophagogastric anastomosis and experienced a postoperative AL from January 2018 to August 2023. The “Non Tube Non Fasting” edition was implemented in the ERAS group. Results 33 ERAS and 141 conventional patients experienced an AL in 489 ERAS and 1424 conventional patients with esophagectomy. The incidence of AL was significantly lower in the ERAS group (6.7%, 33/489) than in the conventional group (9.9%, 141/1424) (P0.05). There were no significant differences in the time-related AL between the two groups: delayed AL (39.4%, 13/33 vs. 54.6%, 77/141) and AL after planned discharge (24.2%, 8/33 vs. 23.4%, 33/141) (P>0.05). 164 patients received endoscopy and presented no significant differences in the average length of leakage between the two groups (5.7±3.8mm vs. 6.1±4.6mm, P>0.05). The comparison of the incidence of the most common early clinical manifestations showed no significant differences between the two groups: cervical wound infection (24.2%, 8/33 vs. 23.4%, 33/141), tachyarrhythmia (93.9%, 31/33 vs. 87.2%, 123/141), fever beyond 38°C (87.9%, 29/33 vs. 91.7%, 135/141), leukocytosis (81.8%, 27/33 vs. 89.4%, 126/141), cervical subcutaneous emphysema (42.4%, 14/33 vs. 41.1%, 58/141), purulent drainage (6.1%, 2/33 vs. 19.1%, 27/141), dyspnea (21.2%, 7/33 vs. 31.9%, 45/141), and right pneumothorax (54.5%, 18/33 vs. 50.4%, 71/141) (P>0.05). The incidences of the time-related early alarm signals were not significantly different between the two groups: within 3 days after surgery (100%, 33/33 vs. 95.7%, 135/141), during postoperative 4 days to 7 days (0%, 0/33 vs. 3.5%, 5/141), and after postoperative 7 days (0%, 0/33 vs. 0.7%, 1/141) (P>0.05). The 30-day mortality was significantly higher in the ERAS group than in the