Mortality Prediction Model Before Surgery for Acute Mesen-teric Infarction: A Population-based Study

Surgery for acute mesenteric infarction (AMI) is associated with high mortality. This study aimed to generate mortality prediction model to predict the 30-day-mortality of surgery for AMI. We included patients ≥18 years received bowel resection in treating AMI and randomly divided into the derivation and validation groups. After multivariable analysis, the ‘Surgery for acute mesenteric infarction mortality score’ (SAMIMS) system was generated and was including age>62-year-old (3 points), hemodialysis (2 points), congestive heart failure (1 point), peptic ulcer disease (1 point), diabetes (1 point), cerebrovascular disease (1 point), and severe liver disease (4 points). The 30-day mortality rates in the derivation group were 4.4%, 13.4%, 24.5%, and 32.5% among very low- (0 point), low- (1–3 point(s)), intermediate- (4–6 points), and high- (7–13 points) risk patients. Compared to the very-low-risk group, the low-risk (OR=3.332), intermediate-risk (OR=7.004), and high-risk groups (OR=10.410, p<0.001) exhibited higher odds of 30-day mortality. We identified similar results in the validation group. The areas under the ROC curve were 0.677 and 0.696 in the derivation and validation groups. Our prediction model, SAMIMS, allowed for the stratification of the patients’ 30-days-mortality risk of surgery for acute mesenteric infarction.