New England Journal of Medicine Catalyst - Using AI to avert missed opportunities for diagnosis in radiological settings
02/18/2026
Members of the PCCI and Parkland Health published a paper on using AI to reduce missed opportunities for diagnosis in the March 2026 edition of New England Journal of Medicine Catalyst, one of the most respected peer-reviewed medical journals in the world.
You can view the article here: https://catalyst.nejm.org/doi/full/10.1056/CAT.25.0401
Authors include PCCI's, Alex Treacher, PhD, George Oliver, MD, PhD, Steve Miff, PhD, Olayide Adejumobi, RDN, LD, MHA, and Albert Karam, MS, MBA. Authors from Parkland Health were Brett Moran, MD, and Molly Case, RN, MHA.
A preview of the paper's abstract outlines the problems PCCI's innovative AI solution is attempting to resolve:
Abstract
Missed opportunities for diagnosis are a critical subset of diagnostic errors that can lead to adverse patient outcomes. These errors frequently arise from failures in the diagnostic process, particularly in ensuring that recommended follow-ups are scheduled and completed. In large health systems, such as Parkland Health in Dallas, Texas, which conducts over 500,000 radiologist studies annually, the challenge of reliably identifying and managing follow-up recommendations is amplified by the reliance on structured note templates (macros) within electronic health records.
The paper revealed the effectiveness of the AI system, finding accuracy and efficiency that saves time and improves outcomes for patients:
This study demonstrates that a custom AI agent built using a pretrained LLM can achieve high predictive performance of follow-up prediction from radiologist notes and tabularize the corresponding details, and significantly improve the detection of radiologist-recommended follow-up imaging compared with a macro-based flagging system. On a held-out test set of 10,000 radiology reports, the AI agent achieved a balanced accuracy of 98.7%, correctly identifying over six times more follow-up recommendations than the existing DIS flag. Furthermore, the agent extracted actionable details — modality, body part, time frame, and rationale — with accuracies exceeding 94%, enabling more precise and scalable integration into care coordination workflows.
For information about this innovative AI program, go to PCCI's Annual Impact Report: https://online.flippingbook.com/view/183669062/8-9/