Background: Many studies have reported that prophylactic abdominal drainage is unnecessary after liver resection, but it is required in some cases. This study aimed to elucidate risk factors for abdominal drainage after liver resection. Methods: This study included 182 patients who underwent liver resection for hepatocellular carcinoma. We compared clinical outcomes between patients with abdominal drainage (drainage group, n=110) and those without (non-drainage group, n=72). A total of 20 patients (drainage group: 12 in whom it was difficult to remove the drain and six in whom a drain was reinserted; non-drainage group: two in whom a drain was inserted) were included in the drain necessary group versus 162 in the drain unnecessary group. We analyzed the risk factors for abdominal drainage. Results:In the drainage group, anatomical hepatic resection was performed significantly more frequently (61% vs 44%, P=0.034), operative time was significantly longer (344 vs 264.5min, P=0.004), and operative blood loss was significantly greater (879.5 vs 328mL, P<0.001). Complications occurred in 22 patients (20%) of the drainage group (P=0.008). In the drain necessary group, preoperative aspartate transaminase (AST) and total bilirubin levels were significantly higher (44.0 vs 31.5IU/L, P=0.031; 1.0 vs 0.8mg/dL, P=0.050), as was the American Society of Anesthesiologists Physical Status (ASA-PS) (P=0.043). Furthermore, in the drain necessary group, hemi-hepatectomy was performed significantly more frequently (55% vs 23%, P=0.005), operative time was significantly longer (404 vs 288min, P<0.001), and operative blood loss was significantly greater (1293 vs 532mL, P<0.001). Multivariate analysis indicated that ASA-PS (Class 3), elevated AST, and massive blood loss were risk factors for abdominal drainage (P=0.045, 0.040, and 0.029, respectively). Conclusions: outine use of abdominal drainage is unnecessary, but it is likely necessary in cases of high ASA-PS, liver dysfunction, longer operation time, or massive blood loss.
View full abstract