Potential change in healthcare costs of implementing artificial intelligence for detecting cervical spine fractures on CT: an early health technology assessment
European Radiology, Springer Nature, 2026
Summary
This early health technology assessment evaluated the in-hospital economic impact of introducing an artificial intelligence (AI) algorithm as a concurrent reader to radiologists interpreting cervical spine (C-spine) CT scans for fracture detection. The study used an early health technology assessment model comparing standard radiologist interpretation of cervical spine CT scans with an AI-assisted concurrent reading scenario, integrating published diagnostic accuracy data and in-hospital cost parameters. A decision-amodelling approach was applied to estimate differences in diagnostic outcomes and downstream healthcare costs between the two strategies. The analysis firstly showed that AI had the potential to improve diagnostic accuracy, by increasing sensitivity to 98.6% and specificity to 100%, showing 10.5% and 0.8% relative increases respectively. Regarding the total in-hospital healthcare costs, in the AI-assisted scenario there was only an increase by 0.3% compared to radiologists working alone. Overall, these findings suggest that integrating AI in this context can improve diagnostic accuracy with a very small impact on in-hospital cost burdens, supporting more accurate and efficient fracture detection in clinical practice.
