Preliminary budget impact assessment of AI assisted radiograph review for suspected fractures in NHS emergency departments

A United Kingdom retrospective health economic modeling study evaluated BoneView (Gleamer), an AI tool for fracture detection to estimate its national budget impact. The effect of AI on assisting ED clinicians when detecting fractures was sourced from published literature (Guermazi et al., 2022; Bent et al., 2013). Data on resource use were extracted from St George’s University Hospital NHS Foundation Trust in London and extrapolated to a 1-year national model.

The model estimated 658,564 radiographs annually for suspected wrist, ankle, or hip fractures in English NHS EDs. AI-assisted review was projected to improve sensitivity (hip: 77% vs 67.2%; wrist/ankle: 73.1% vs 61.8%) and specificity (hip: 96% vs 93%; wrist/ankle: 96.1% vs 89.8%), reducing 21,674 return visits and 20,916 unnecessary fracture clinic referrals. The budget impact analysis showed annual care costs of £66.65M under standard care vs £63.01M with AI, producing an estimated £3.63M net savings. Scenario analyses highlighted ED clinician accuracy and AI sensitivity as the primary drivers of economic benefit.

AI-assisted fracture detection has the potential to reduce missed fractures, avoid unnecessary patient visits, and save the NHS millions annually. However, these findings are model-based, combining public AI performance data with local NHS resource-use patterns, and prospective multicenter NHS studies are required to validate real-world clinical and economic impact.

Read full study

Early Budget Impact Analysis of Artificial Intelligence to Support the Review of Radiographic Examinations for Suspected Fractures in National Health Service Emergency Departments

Value in Health, 2025

Abstract

Objectives

To develop an early budget impact analysis of and inform future research on the national adoption of a commercially available artificial intelligence (AI) application to support clinicians reviewing radiographs for suspected fractures across National Health Service (NHS) emergency departments (ED) in England.

Methods

A decision tree framework was coded to assess a change in outcomes for suspected fractures in adults when AI fracture detection was integrated into the clinical workflow over a 1-year time horizon. Standard of care was the comparator scenario, and the ground truth reference cases were characterized by radiology report findings. The effect of AI on assisting ED clinicians when detecting fractures was sourced from US literature. Data on resource use conditioned on the correct identification of a fracture in the ED were extracted from a London NHS trust. Sensitivity analysis was conducted to account for the influence of parameter uncertainty on results.

Results

In 1 year, an estimated 658 564 radiographs were performed in the EDs across England for suspected wrist, ankle, or hip fractures. The number of patients returning to the ED with a missed fracture was reduced by 21 674 cases, and a reduction of 20 916 unnecessary referrals to fracture clinics was also noted. The cost of the current practice was estimated at £66 646 542 and £63 012 150 with the integration of AI. Overall, a return on investment of £3 634 392 to the NHS was generated.

Conclusions

The adoption of AI in EDs across England has the potential to generate cost savings. Nevertheless, additional evidence on radiograph review accuracy and subsequent resource use is required to further signify this.