AID Systems in Clinical Practice: Evidence and Case Studies Workshop | Recording-AID Systems in Clinical Practice: Evidence and Case Studies Workshop

Video Transcription

Video Summary

This ADA workshop, “AID Systems in Clinical Practice,” reviews FDA-approved automated insulin delivery (AID) systems and how to interpret pump/CGM reports to guide clinical decisions. Presenters highlight updates from the 2026 ADA Standards of Care, emphasizing AID as the preferred insulin delivery method for people with type 1 diabetes and increasingly for insulin-requiring type 2 diabetes, with device choice based on individual needs and preferences.

An overview compares major systems and the settings most responsible for outcomes—especially glucose targets and how each algorithm uses adjustable parameters. Five real-world cases illustrate practical use:

1) <strong>iLet (Beta Bionics) bionic pancreas</strong>: A 76-year-old man with long-standing type 1 diabetes and poor engagement with carb counting improved from 35% to 73% time-in-range. Key concept: iLet initiation primarily requires <strong>body weight</strong>, and the <strong>glucose target</strong> is the main lever.

2) <strong>MiniMed 780G</strong>: A 40-year-old man with pancreatic diabetes, severe comorbidities, and missed doses improved dramatically (A1c to 7.6%, ~72% time-in-range) with fewer hospitalizations. Most impactful settings: <strong>target (ideally 100)</strong> and <strong>active insulin time (2 hours)</strong>; education focused on avoiding late bolus “stacking” with auto-corrections.

3) <strong>Omnipod 5 (type 2 diabetes)</strong>: A 76-year-old woman with CKD and CVD reduced A1c to 7.4% and improved time-in-range using simplified “small/medium/large” meal entries. Key lever: <strong>lowering algorithm target</strong>; basal rate and correction factor don’t drive automated delivery.

4) <strong>Tandem Control-IQ(+)/Mobi/X2</strong>: A teen with type 1 had hypoglycemia from frequent overrides and overtreatment of lows; guidance included fewer carbs for lows (5–10g) and appropriate use of sleep mode.

5) <strong>TwiST (Loop)</strong>: Highly customizable targets and presets (pre-meal/workout); insulin sensitivity strongly affects algorithm behavior.

The session ends with resources and an AID comparison handout for clinical implementation.

Keywords

automated insulin delivery (AID) systems

FDA-approved insulin pumps

CGM pump report interpretation

2026 ADA Standards of Care

type 1 diabetes AID preferred therapy

insulin-requiring type 2 diabetes AID

iLet Beta Bionics bionic pancreas

MiniMed 780G target 100 active insulin time 2 hours

Omnipod 5 algorithm target settings

Tandem Control-IQ and TwiST Loop customization