RESUMEN
PURPOSE: Preoperative hypoalbuminemia has traditionally been used as a marker of nutritional status and is considered a significant risk factor for anastomotic leak (AL). METHODS: The Westmead Enhanced Recovery After Surgery (WERAS) prospectively collected database, consisting of 361 patients who underwent colorectal surgery with primary anastomosis, was interrogated. Preoperative serum albumin and protein levels (measured within 1 week of surgery) were plotted on receiver operating characteristic curves (ROC curves) and statistically analyzed for cutoff values, sensitivity, specificity, positive predictive values (PPV), and negative predictive values (NPV). RESULTS: The incidence of AL was 4.4% (16/361). Overall mortality was 1.4% (5/361), 6.3% (1/16) in the AL group, and 1.2% (4/345) in the no AL group. The median preoperative albumin and protein level in the AL group were 39 g/L and 75 g/L, respectively. The median preoperative albumin and protein level in the no AL group were 38 g/L and 74 g/L, respectively. The Mann-Whitney U test showed no statistically significant difference in albumin levels (p = 0.4457) nor protein levels (p = 0.6245) in the AL and no AL groups. ROC curves demonstrated that preoperative albumin and protein levels were not good predictors of anastomotic leak. Cutoff values for albumin (38 g/L) and protein (75 g/L) both had poor PPV for AL (4.8% and 3.8% respectively). CONCLUSION: In patients undergoing elective colorectal surgery as part of an ERAS program, preoperative serum albumin and protein levels are not reliable in predicting AL. This may be because of nutritional supplementation provided as part of an ERAS program may correct nutritional deficits to protect against AL or that low albumin/protein is not as robust a marker of AL as previously reported.