Abstract
Development of an enteric fistula after surgery is a major therapeutic complication. In this study, we retrospectively examined the potential relationship between preoperative laboratory data and patient mortality by collecting patient data from a tertiary medical center. We included patients who developed enteric fistulas after surgery for gastrointestinal (GI) cancer between January 2005 and December 2010. Patient demographics and data on preoperative and pre-parenteral nutritional statuses were compared between surviving and deceased patients. Logistic regression analysis and receiver operating characteristic (ROC) curves were used to determine the predictors and cut-off values, respectively. Patients with incomplete data and preoperative heart, lung, kidney, and liver diseases were excluded from the study; thus, out of 65 patients, 43 were enrolled. Logistic regression analysis showed that blood urea nitrogen-to-creatinine (BUN/Cr) ratio [p = 0.007; OR = 0.443, 95% confidence interval (CI), 0.245-0.802] was an independent predictor of mortality in patients who developed enteric fistulas after surgery for GI cancer. In conclusion, the results of our study showed that a high preoperative BUN/Cr ratio increases the risk of mortality in patients who develop enteric fistulas after surgery for GI cancer.