Pediatric Endocrinology – Improve Care Delivery for T1D Patients Through Automation and Capacity Optimization

LPCH Pediatric Endocrinology Clinic’s CGM Program uses TIDE – An interactive tool to facilitate the patient review process:

  1. Number of flags and CGM metrics highlight high-risk patients
  2. Daily glucose levels reveal important behavioral trends
  3. Flags and CGM data in one dashboard for quick access

Patient prioritization algorithm determines the patients clinicians should review each week. The prioritization rule is chosen based on its performance on historical patient data and our goal to reduce the number of patients reviewed per week while still reviewing the patients that clinicians had prioritized to contact based on safety and opportunity to improve mean glucose and time in range (70-180mg/dl) in the past after reviewing their CGM data.

The goal of our project is to test if the Pediatric Endocrinology clinic can increase its capacity without increasing resources through technological and operational innovations. We are looking at options that facilitate population-level glucose management by enabling the clinic to provide care for more patients while maintaining or improving glucose metrics. We are working on a 3-pronged strategy to achieve our goal:

  1. A model that will allow the clinic management to have an estimate of CDE capacity and the number of patients that can be reviewed and contacted.
    • Cadence Analysis: Determine an optimal cadence of remote monitoring based on reasonable CDE capacity and patient needs.
    • Current cadence: Once a week for the patient’s first year in the study
      • Once a month month after the first year
  2. Technology improvements to optimize provider capacity & process efficacy:
    • Limitations (Current State):
      • Data Preparation – Manual setup and execution
      • Scalability – limited scope to support large number of patients & CDEs
      • Historical Data – Only point-in-time data available
    • Goals:
      • Business Process Optimization
        • Optimize CDE staff time
      • Data Protection & Security
        • PHI Safe Data Storage
        • Server based deployment
        • Historical data available for analysis
      • Technical Process Optimization
        • Automate Data Pipeline
        • Automate Prioritization Algorithm
  3. Financial Model that allows for proper reimbursement for the services provided
    • Challenges:
      • Stanford Children’s commitment to experimenting with risk-based payment (in parallel with FFS model)
      • Cultural alignment for clinical staff
    • Payment Model Innovation – Feasibility Framework
      • Understand the unit economics of managing peds diabetes
      • Define clear outcomes goals that are clinically reasonable
      • Develop payment model that rewards outcomes delivered

Annie Chang

B.S. Human Biology


Ryan Leonard Pei

Sloan Fellow – Graduate School of Business


Michael Gao

MS – Management Science & Engineering

David Scheinker

Founder & Director, SURF Stanford Medicine

Prashant Yadav

Sloan Fellow – Graduate School of Business