[Establishment of a nomogram prediction model using common preoperative indicators for early weight loss after laparoscopic sleeve gastrectomy]

Objectives: To construct a nomogram prediction model using common preoperative indicators for early weight loss (EWL) 1 year after laparoscopic sleeve gastrectomy (LSG). Methods: Relevant data of obese patients who had undergone LSG from January 2015 to May 2022 in Fujian Medical University Union Hospital and Quanzhou First Hospital Affiliated Fujian Medical University were analyzed. Patients with a history of major abdominal surgery, severe gastroesophageal reflux disease, pregnancy within 1 year after surgery, or who were lost to follow-up were excluded, resulting in a total of 200 patients in the study (190 from Fujian Medical University Union Hospital and 10 from Quanzhou First Hospital Affiliated Fujian Medical University). The participants were 51 men and 149 women of a mean age 29.9±8.2 years and a body mass index (BMI) 38.7±6.5 kg/m2. All patients in this group underwent standardized LSG procedure. Achieving ideal weight (BMI≤25 kg/m2) 1 year after LSG was defined as goal of EWL. Logistic regression analyses were performed to identify factors that independently influenced EWL. These factors were incorporated into the nomogram model. Receiver operating characteristic (ROC) curves (the larger the area under the curve [AUC], the better the predictive ability and accuracy of the model), likelihood ratio test (higher likelihood ratio indicates greater model homogeneity), decision curve analysis (higher net benefit indicates a better model), Akaike information criterion (AIC; smaller AIC indicates a better model), and Bayesian information criterion (BIC; smaller BIC indicates a better model) were used to validate the predictive ability of the column line diagram model. Results: In this study of 200 obese patients who underwent LSG surgery, 136 achieved EWL goal, whereas the remaining 64 did not. The rate of EWL goal achievement of the entire group was 68.0%. Compared with patients who did not achieve EWL goal, those who did had lower BMI, alanine transaminase, aspartate transaminase, triglycerides, and higher cholesterol. Additionally, the proportion of female was higher and the proportions of patients with fatty liver and hypertension lower in those who achieved EWL goal (all P<0.05). Univariate and multivariate logistic regression analysis revealed that preoperative BMI (OR=0.852, 95%CI: 0.796-0.912, P<0.001), alanine transaminase (OR=0.992, 95%CI: 0.985-0.999, P=0.024), presence of fatty liver (OR=0.185, 95%CI: 0.038-0.887, P=0.035) and hypertension (OR=0.374, 95%CI: 0.144-0.969, P=0.043) were independently associated with failure to achieve EWL goal. Cholesterol (OR=1.428, 95%CI: 1.052-1.939, P=0.022) was independently associated with achieving EWL goal. We used the above variables to establish an EWL nomogram model. ROC analysis, the likelihood ratio test, decision curve analysis, and AIC all revealed that the predictive value of the model was better than that of BMI alone (nomogram model vs. BMI: area under the curve 0.840 vs. 0.798, P=0.047; likelihood ratio: 58.785 vs. 36.565, AIC: 193.066 vs. 207.063, BIC: 212.856 vs. 213.660). Conclusion: Our predictive model is more accurate in predicting EWL after LSG compared with using BMI.