Youth in the United States collectively exhibit deterioration in glycemic control between the ages of 8 and 18, with small improvements beginning to show from ages 18 to 30. Evidence suggests that suboptimal control early in the course of disease has an irrevocable impact on the risk of developing future health complications. To combat this problem and improve diabetes self-management in youth, we partnered with Cyft to develop a prediction model that identifies patients who are at an elevated risk of T1D-related complications (rise in HbA1c or hospital admission for Diabetic Ketoacidosis).
Using Cerner electronic health records, we constructed a predictive model that draws on a patient’s health history (up to 1 year prior to visit date) to evaluate their glycemic control and predict HbA1c reading at the next visit. Patients who are predicted to have a significant rise in HbA1c or a DKA admission between their current visit and their subsequent visit (approximately 70-110 days) are deemed eligible for the Remote Patient Monitoring (RPM) program and given the choice to participate.
During recruitment, patients and their families are approached for inclusion in RPM by phone outreach or patient portal message. If a family opts into RPM, they will schedule a time to meet with an interventionist through a virtual face-to-face platform or phone call. During these RPM calls, interventionists assess the needs of the patient/family and address any questions or concerns a patient or family member may have regarding diabetes care, such as insulin regimen, prescription refills, urgent management, concerns at school, barriers to care, or any other unique issues. Upon initial enrollment, weekly or biweekly scheduled check-in appointments will be created for the family with the interventionist with the aim to schedule patients for 3 or more visits total; however, exceptions can be made for families whose needs have been met in fewer visits, and those who would like to continue the additional assistance, provided that the interventionists have the capacity to do so.
As of June, RPM has served over 100 families at Children’s Mercy. We have expanded the program to families whose first language is not English, as well as patients who have had two HbA1c readings over 9.0% in the last year. By increasing accessibility of RPM, we can provide additional aid to families that do not appear on the A1c rise/DKA admission predictions list but can still benefit from having this additional resource. Analysis of previous participant data shows that the proportion of individuals experiencing a 90-day rise in A1c is trending down (55% expected versus 38% actual) with the RPM intervention (N=25).