Team 10: Visual dashboard and predictive analytics for pediatric endocrinology
Type 1 diabetes is a condition estimated to affect about 1 million people in the United States, and close to 2500 children with the condition are followed at Children’s Health System of Texas. Children with type 1 diabetes have lost their ability to make insulin and as a result must take insulin either via subcutaneous injection or continuous subcutaneous infusion (insulin pump therapy) to try to maintain normal glucose levels. Despite the large numbers of children with the condition nationwide, multiple barriers remain which limit our understanding of risk factors for worse outcomes in children with type 1 diabetes.
To begin to address this problem, we have developed SAP based descriptive reports (spreadsheets) of our patient population which allow manual queries of the population to be done which can provide users with tediously derived summary statistics of the population of interest.
Two goals remain: 1) to turn these data sets into visual dashboards which can more efficiently display the desired summary descriptors of the population and 2) to use predictive analytics to begin to stratify individual patients based on risk factors for undesired outcomes (emergency room visits, admissions, or average glycemic control above specified thresholds) so that patients can be placed into care pathways most appropriate for their risk stratification. Deidentified data sets in spreadsheet format will be made available.