PROGNOSTIC VALUE OF THE ABCD AND IMS FOR EVALUATION OF TYPE 2 DIABETES MELLITUS REMISSION AFTER BARIATRIC SURGERY

Abstract

Background. Significant proportion of patients with obesity and type 2 diabetes mellitus (DM) have significant weight loss and improved metabolic outcomes as a result of bariatric surgery. To predict the effect of surgical treatment of DM, several scales have been proposed, including ABCD and IMS.Objective. To estimate accuracy of the ABCD and IMS scales in predicting DM remission in patients undergoing bariatric surgery.Design and methods. 38 patients with type 2 diabetes were identified after bariatric surgery with a follow-up period of at least 1 year. The KNIME Analytics Platform 4.3.6 (KNIME AG, Switzerland) was used for data processing.Results. 12.8 % of patients achieved partial remission, 52.6 % complete remission, and 31.6 % did not achieve remission. According to the IMS, no significant results were detected in the remission groups. IMS scale have a low predictive value. A significant result was obtained for the ABCD after CAIM binning. For the binary classification (“Remission”/”Haven’t remission”): AUC = 0.98 and Cohen’s kappa k = 0.86 for probability treshold 0.55399, that maximized F-measure (0.96) were obtained. So, ABCD predictive value is high.Conclusion. The ABCD have a better predictive value. Ease of use, good prognostic effect allows us to recommend ABCD before bariatric treatment.