Interview Summary

This interview highlights IQVIA’s AI-enabled clinical decision support tool, developed with Breakthrough T1D, and its real-world impact on improving identification of adult-onset type 1 diabetes (T1D). Bringing together data science, clinical and patient perspectives, the discussion focuses on how IQVIA Healthcare-grade AI® addresses the persistent challenge of misdiagnosing and misclassifying adult-onset T1D as type 2 diabetes (T2D).

The tool was recently recognized with the 2026 AI Breakthrough Award for “Predictive Modelling Solution of the Year”, reflecting its innovation and measurable impact in enabling earlier diagnosis and improved patient outcomes.

Speakers

• Raquel López Díez, PhD, Senior Scientist, Breakthrough T1D
• Nadea Leavitt, Senior Director, IQVIA Applied AI Science
• Joe McFadden, Consultant, AI Solutions Delivery, IQVIA Applied AI Science (Person living with T1D)

Take home points

• Critical unmet need: Misdiagnosis and misclassification of adult-onset T1D as T2D is common and can delay appropriate treatment and increase the risk of life-threatening complications.
• AI enables targeted action: The model analyzes EMR data to identify high-risk patients and support earlier clinical evaluation.
• Demonstrated real-world impact: 99.5% reduction in screening burden – from approximately 60,000 to 300 patients – thereby requiring 200 times fewer assessments. Of the patients flagged, 28% were confirmed to have a clinical presentation of T1D versus a 0.22% baseline.
• Designed for clinical use: Transparent outputs support clinician trust, usability and decision making.
• Better outcomes through earlier diagnosis: Timely identification enables appropriate treatment sooner, improving long-term health and quality of life.
• Power of collaboration: Close partnership across AI, clinical and patient communities is key to scaling real-world impact.

Key minute marks

• 00:00–00:30 – Introduction and overview of the IQVIA–Breakthrough T1D collaboration
• 00:30–03:30 – The clinical challenge: misclassification and its impact
• 04:30–06:30 – Model development, validation and performance
• 06:45–09:15 – How the algorithm works in practice
• 09:15–11:45 – Clinical impact and care pathway improvements
• 11:45–14:50 – Patient perspective and importance of early diagnosis
• 14:50–15:00 – Closing remarks and future outlook

For more detail on the research behind this work, read the following studies:

• Prospective validation of an AI algorithm to identify adult-onset type 1 diabetes misclassification: protocol for a non-interventional multicentre study
• Predicting misdiagnosed adult-onset type 1 diabetes using machine learning
• Prospective Validation of AI for Detecting Misclassified Adult Type 1 Diabetes: Insights on Precision, Clinical Workflows, and Adoption