Clinical Data in Short Bowel Syndrome (SBS): A Data Driven Approach

Introduction: Short Bowel Syndrome (SBS) is characterized by a peculiar pathophysiological frameworks: affected patients, often undergoing parenteral nutrition (NPT), present alterations in the laboratory data for which there are no specific reference values. Materials and Methods: A software platform has been developed with functions for the management, analysis, data visualization (Data Warehouse) and for the definition and training of predictive models and algorithms (Machine Learning), aiming at a future use on large amounts of information (Big Data) and creating a prototypical environment dedicated to SBS but potentially adaptable to other biomedical contexts. The information used for this study was handled anonymously and, where required, as part of the consent to the processing of personal data provided during hospitalization of patients, in compliance with the requirements of the General Data Protection Regulation (officially regulated by the European Union No. 2016/679 and signed GDPR19). Results: The database allows to obtain statistical data regarding the anamnestic and laboratory parameters of patients with SBS. Among the results data obtained we found that the prevalence of SBS is equal in both sexes and an incidence of Gastroschisis, Volvulus and NEC was found as the primary cause of intestinal resection. About 2/3 of SBS patients were treated with intestinal lengthening, after which 76% of them were weaned from parenteral nutrition or decreased the amount. The glycemia in these patients is higher than in the standard pediatric population, with an increase in the average value as the age of the subjects examined increases. Conclusions: The data-driven approach has proven to be effective in highlighting correlations between different aspects of the patient’s clinical history and, with a greater number of data, will be able to provide predictions on patient outcome allowing for therapeutic modulation.