AI to be used by NHS to help prevent bone fractures being missed

The NHS in England is set to utilise artificial intelligence (AI) to help prevent medical professionals from overlooking bone fractures on X-rays. According to experts, these technologies can expedite the diagnostic process and decrease the necessity for follow-up appointments. Urgent care centres in England have been given the green light to use four AI platforms following the release of draft guidelines by the National Institute for Health and Care Excellence (Nice).

Nice has highlighted that clinical evidence indicates AI could enhance fracture detection on X-rays in urgent care settings without increasing the risk of incorrect diagnoses. By reducing the number of missed fractures, the technology can lower the likelihood of further injury or harm to patients between the initial assessment and treatment decision. Additionally, improving diagnostic accuracy may lead to a decrease in hospital recalls and unnecessary referrals to fracture clinics.

Typically, individuals with suspected fractures are assessed by urgent care staff who then request X-rays to be conducted by radiographers. The current protocol recommends these X-rays be reviewed by a radiologist or another trained professional before the patient is discharged. However, reporting delays lasting days or weeks are common due to workforce shortages in radiology departments nationwide.

Mark Chapman, Director of Healthtech at Nice, emphasised the necessity for additional support to alleviate the high workload faced by radiologists and radiographers. He stated that AI technologies are safe and have the potential to identify fractures that might be missed under the current work pressures. The AI platforms recommended for NHS use are TechCare Alert for all age groups, Rayvolve for adults only, and BoneView and RBfracture for both adults and children aged two and above.

The introduction of AI to aid professionals in urgent care centres could lead to quicker diagnoses and reduce the need for follow-up appointments resulting from missed fractures during initial assessments. While the true cost of implementing these AI technologies remains uncertain, it is crucial for centres to ensure the cost per scan aligns with the estimate in the draft guidance. A consultation on the draft recommendations is scheduled to run until November 5, 2024.

Charlotte Beardmore, Executive Director of Professional Policy at the Society and College of Radiographers, supports Nice’s draft guidance on AI usage for fracture detection in urgent care. She stressed that AI tools should complement the expertise of radiographers and radiologists while enhancing patient care. It is essential that all results are reviewed by certified professionals to ensure compliance with regulations on ionizing radiation exposure.

In conclusion, the integration of AI technologies in healthcare could significantly improve the accuracy and efficiency of fracture detection, benefiting both patients and healthcare professionals. This development underscores the ongoing effort within the NHS to leverage innovative solutions for enhanced patient outcomes and streamlined healthcare services.