Search the hub

Medical device makers have been rushing to add AI to their products. While proponents say the new technology will revolutionize medicine, regulators are receiving a rising number of claims of patient injuries. This Reuters Special Report investigates some of the hazards associated with AI-enabled medical devices, including errors in a navigation system integrated into a medical device used in ENT surgery, AI software used for prenatal ultrasound scans that misidentified fetal body parts and AI assisted heart monitors that failed to recognise abnormal rhythms.  Issues with the capacity of the U.S. Food and Drug Administration (FDA) to review the flood of new AI-enabled medical devices are also raised, as well as concerns that the FDA's traditional approach to regulating medical devices may no longer be fit for purpose.

Recent advancements in artificial intelligence (AI) and the vast data generated by modern clinical systems have driven the development of AI solutions in medical imaging, encompassing image reconstruction, segmentation, diagnosis, and treatment planning. Despite these successes and potential, many stakeholders worry about the risks and ethical implications of imaging AI, viewing it as complex, opaque, and challenging to understand, use, and trust in critical clinical applications. The FUTURE-AI guideline for trustworthy AI in healthcare was established based on six guiding principles: Fairness. Universality. Traceability. Usability. Robustness. Explainability. Through international consensus, a set of recommendations was defined, covering the entire lifecycle of medical AI tools, from design, development, and validation to regulation, deployment, and monitoring. In this paper, the authors describe how these specific recommendations can be instantiated in the domain of medical imaging, providing an overview of current best practices along with guidelines and concrete metrics on how those recommendations could be met, offering a valuable resource to the international medical imaging community.

This report shares findings from complaints made to Parliamentary and Health Service Ombudsman (PHSO) about failings in imaging in the NHS. The majority of these complaints involve people who had cancer at the time they used imaging services. Through highlighting these complaints, the PHSO’s objective is to support NHS services to improve. It suggests that failings in imaging services can only be addressed and learned from through collaboration across clinical specialties, looking at the whole imaging journey and its intersections as part of the patient’s care pathway. Findings The PHSO analysed 25 complaints relating to failings in the imaging journey since 2013. Most concluded between 2018 and 2020. This report focuses on four key findings they have seen in their casework Failure to follow national guidelines on reporting unexpected imaging findings. Failure to act on important unexpected findings. Delays in reporting imaging findings. Failure to learn from past mistakes. Recommendations For the system as a whole: Recommendations from previous work related to imaging must be implemented as a priority (referring to previous reports by the Healthcare Safety Investigation Branch, Care Quality Commission and the Independent Review of Diagnostic Services). Digital infrastructure must now be treated as a patient safety issue. Department of Health and Social Care and NHS England and NHS Improvement should ensure there is national guidance on the roles and responsibilities of clinicians, and expected timeframes, at each stage of the imaging journey. Department of Health and Social Care and NHS England and NHS Improvement should write to the Health and Social Care Select Committee and the Public Administration and Constitutional Affairs Committee by the end of March 2022. For imaging services specifically: All NHS-funded providers that have a radiology service should ensure staff working in those services have sufficient allocated time in their job plans for meaningful learning and reflection. Clinical directors and senior managers of NHS-funded radiology services should triangulate the learning from across their departments on a regular basis. The Royal College of Radiologists should review existing guidance on reporting unexpected findings and peer review of radiological reports to learn from the findings of PHSO’s casework.

Diagnostic errors have a negative impact on patient treatment and cost healthcare systems a large amount in wasted resources. This paper published by the Deeble Institute for Health Policy Research looks at diagnostic errors related to medical imaging in Australian public healthcare. It also looks at health policies that have been used internationally to improve the use of diagnostic imaging and reduce the consequences of diagnostic errors. The authors recommend: implementing a national strategy in Australia to identify and prevent diagnostic errors analysing medical indemnity claims to help measure the incidence and consequences of diagnostic errors.

Half of COVID-19 patients who received a heart scan in hospital showed abnormalities in heart function, according to new research funded by the British Heart Foundation. In this study, Dweck et al. describe the cardiac abnormalities in patients with COVID-19 and identify the characteristics of patients who would benefit most from echocardiography. The study, published in the European Heart Journal - Cardiovascular Imaging, found that around one in seven showed severe abnormalities likely to have a major effect on their survival and recovery. It also showed that one in three patients who received an echocardiography scan had their treatment changed as a result. The findings suggest that heart scans could prove crucial for identifying patients who may benefit from additional treatments to improve their COVID-19 recovery and prevent potential long-term damage to their heart. Professor Marc Dweck, British Heart Foundation Senior Lecturer and Consultant Cardiologist at the University of Edinburgh, said: “Covid-19 is a complex, multisystem disease which can have profound effects on many parts of the body, including the heart. Many doctors have been hesitant to order echocardiograms for patients with Covid-19 because it’s an added procedure which involves close contact with patients. Our work shows that these scans are important – they improved the treatment for a third of patients who received them.”

There has been an increase in the number of units providing anaesthesia for magnetic resonance imaging and the strength of magnetic resonance scanners, as well as the number of interventions and operations performed within the magnetic resonance environment. More devices and implants are now magnetic resonance imaging conditional, allowing scans to be undertaken in patients for whom this was previously not possible. There has also been a revision in terminology relating to magnetic resonance safety of devices.  These guidelines, by the Association of Anaesthetists, have been put together by organisations who are involved in the pathways for patients needing magnetic resonance, reinforce the safety aspects of providing anaesthesia in the magnetic resonance environment and suggest that hospitals should develop and audit governance procedures to ensure that anaesthetists of all grades are competent to deliver anaesthesia in the magnetic resonance environment.

The Care Quality Commission (CQC)’s annual report on Ionising Radiation (Medical Exposure) Regulations in England has been published. The report gives a breakdown of the number and type of statutory notifications of errors received from healthcare providers in 2018/19 where patients were exposed to ionising radiation. These notifications are where there have been significant accidental or unintended exposures, for example where a patient received a higher dose than intended or where the wrong patient was exposed. Professor Ted Baker, CQC’s Chief Inspector of Hospitals, said:“It is important that organisations learn from incidents and take action to mitigate any risks when patients are exposed to ionising radiation from x-rays, radiotherapy or radiopharmaceuticals as part of their diagnosis or treatment." “The number of errors involving patients is small in the context of the many millions of procedures undertaken each year involving radiation. That said, in too many cases errors happen as a result of inadequate checks, poor communication, or because of a simple failure follow procedures around radiation protection." The report includes recommended actions that providers can take to improve compliance with the regulations and the quality and safety of care for patients. It also shares examples of good practice to help leaders and healthcare professionals identify where they can make improvements in their own services.

This is a story of a patient in whom the emergency department missed the same diagnosis twice, four years apart. The first occasion (prior to his diagnosis of ankylosing spondylitis) was understandable. The second was not. As a result of this case, the hospital have changed their x-ray policy for non-traumatic back pain. They also want to share key learning points (the majority of which were due to lack of awareness about a relatively rare condition and its complications) as widely as possible, to help others avoid the same errors.  This reflective learning features guest educator, Mr Gareth Dwyer (the patient).

The objective of this investigation was to understand the context of magnetic resonance imaging (MRI) scanning under general anaesthetic and how care may be reasonably adjusted for patients with autism or learning disabilities. The ‘reference event’ was Alice, a teenage girl who had autism. Sadly, Alice died following her MRI scan under general anaesthetic. The findings and conclusions of this investigation may be applicable to other non-invasive procedures carried out on patients who are under general anaesthetic. The investigation identified: There is an opportunity to clarify the consent requirements for diagnostic imaging facilitated by a general anaesthetic. There is variation in the information given to patients regarding anaesthesia at the point of referral for an MRI scan under general anaesthetic. The observations and examinations to be routinely performed in pre-anaesthetic assessment are not defined nationally. The investigation found variation in the hospitals it visited. Children coming into hospital for an MRI scan who had been assessed as fit for anaesthetic were perceived as “well” by ward staff. Children with autism, learning disabilities and/or learning difficulties often find clinical environments distressing, which may be reflected in their physiological observations. This may result in diagnostic overshadowing, where problems such as autism (or a medical condition) are attributed as the cause of other new problems, rather than considering other underlying causes, thereby leaving other co-existing conditions potentially undiagnosed. Children with autism, learning disabilities or learning difficulties may benefit from reasonable adjustments being made when attending hospital. Electronic flagging systems can help staff identify patients who may benefit from reasonable adjustments. Hospital passports provide valuable information to assist with implementation of these adjustments. The model of care for learning disability nursing teams is not standardised nationally. There is an opportunity to enhance the existing published guidance available to assist clinicians involved in general anaesthetics to prepare for adverse events in the MRI scanning environment. Professional networks for anaesthetists provide the opportunity for shared learning and consensus regarding best practice. It is challenging to comply fully with the existing published standards for anaesthetic equipment used in MRI environments.

Incidental imaging findings (IIFs) are things that show up on a diagnostic image that are not related to the reason a healthcare professional ordered the test. Inadequate follow-up of IIFs can result in poor patient outcomes, patient dissatisfaction and provider malpractice. In an effort to improve awareness of IIFs, this study aimed to investigate communication of IIFs on inpatient discharge summaries after implementation of a new electronic health record (EHR) notification system. The results showed that IIFs were included in 51% of discharge summaries. The authors identified that lack of inclusion of IIFs on discharge summaries could be related to transitions of care within hospitalisation, provider alert fatigue and many diagnostic testing results to distil. The findings demonstrate the need to improve communication of IIFs and the need for care coordinators to follow up on IIFs. This year’s World Patient Safety Day on 17 September 2024 (WPSD 2024) is focused on the theme “Improving diagnosis for patient safety”. Find out more.

Peripheral nerve blocks (PNB) are safe and effective alternatives or supplements to general anaesthesia. They may improve pain control both during and after surgery, thus avoiding many of the side effects of systemic opioids. PNBs may also lead to improved patient satisfaction, decreased resource utilization, and may be better for the environment by decreasing usage of aesthetic gases and other medications. With the growing use of peripheral nerve blocks in the United States, this paper examines safety issues surrounding the procedures. It examines the safety of nerve blocks as it relates to: nerve injury recognition and treatment of local anaesthetic systemic toxicity (LAST) appropriate health care professional performance of timeouts to avoid wrong-site blocks.

This investigation by the Healthcare Safety Investigation Branch (HSIB) explores the timely recognition and treatment of suspected pulmonary embolism in emergency departments. Pulmonary embolisms can form when clots from the deep veins of the body, usually originating in the legs, travel through the venous system and become lodged in the lungs. A person suffering from a pulmonary embolism requires urgent treatment to reduce the chance of significant harm or death. Investigation scope This national investigation: examined clinical decision making in the diagnosis and treatment of pulmonary embolism (PE) and the role of expertise (significant knowledge and skill that supports effective and practical decision making) using an Applied Cognitive Task Analysis (ACTA). identified factors in the wider healthcare system that support or inhibit diagnostic decision making when staff are seeing patients with non-specific symptoms and signs that may suggest PE. Findings Recognising that a person may have a PE is challenging, particularly for less experienced staff and when the person’s signs and symptoms are non-specific or atypical. Deciding whether to initiate treatment for a suspected PE requires a decision that balances risks, and this decision can benefit from expert knowledge and skill. Despite expertise and the available tools to help identify patients who may have a PE, a small number of PEs may always be missed. Experts use different thought processes and show different behaviours when making decisions compared to more novice staff. Decision-making skills in healthcare are commonly developed through experience, without formal training or opportunities to practise making decisions. Simulation-based learning has the potential to help staff acquire decision making skills more quickly. Other industry sectors, such as aviation and the fire service, aim to accelerate the development of decision-making skills through structured training and the use of ‘generic decision tools’ for analytical decisions. Emergency departments (EDs) do not always provide the conditions which support the development of decision-making skills. Decision making in EDs is affected by workload, workforce availability, and performance targets. ED staff asked for further guidance to be provided on the use of decision aids to support the diagnosis of PE. The design of ED processes influences the decisions staff make. There is no standard model of initial patient assessment in EDs; this contributes to variation in the requesting of tests which can affect later decisions. Pathways for the diagnosis and treatment of PE in outpatient settings may create a safety risk where patients are discharged on anticoagulation medicines without a confirmed diagnosis; the capacity of imaging services is a significant contributor to this. Loss of clinical information when a patient’s care is handed over was identified as a further safety risk. This can contribute to harm if tests, such as D-dimer (a blood test used as part of the assessment of likelihood of PE), are not followed up. Work procedures for the diagnosis and treatment of PE are not routinely designed in line with human factors principles to support their access and use. The physical design of environments may also affect decision making. Safety observations It may be beneficial for healthcare to learn from other industries and develop its own evidence base on strategies to accelerate the development of expert decision-making skills. These strategies may include: development of a generic decision tool for implementation in healthcare, training and clinical practice to support analytical decision making; incorporation into education programmes of theory around how people make decisions and influences on decision making; the use of simulation as a regular intervention to support practice and development of decision-making skills across scenarios with different levels of complexity; and consideration of the role of simulation in competency assessments for key skills. It may be beneficial if the findings of this investigation are used to support the development of staff expertise in decision making through: building understanding of how experts think and make decisions; supporting reflection on the outcomes of simple and complex decisions; development of clinical supervision skills of senior staff; and regular multidisciplinary case review. It may be beneficial for individual organisations to understand the extent to which national guidance on the diagnosis and management of pulmonary embolism is implemented across their organisations. This would help to identify local barriers to implementation to address. In particular it may be helpful to consider, in line with the findings of this investigation, local engagement with the scoring systems available to help predict the likelihood of a pulmonary embolism. It may be beneficial for emergency departments and same-day emergency care units to have rapid access to recommended imaging for patients who require it for the diagnosis of pulmonary embolism. It may be beneficial for the positivity standard for computerised tomography pulmonary angiography (CTPA) (that at least 15% of CTPAs should show a pulmonary embolism) to be evaluated to understand its effects on emergency department decision making. It may be beneficial for healthcare work procedures to be written in line with the principles for effectiveness and usability provided by the Chartered Institute of Ergonomics and Human Factors.

A broken hip or ‘hip fracture’ is a serious injury, which each year in the UK leads to around 75,000 people needing hospital admission, surgery and anaesthesia, followed by weeks of rehabilitation in hospital and the community. The National Hip Fracture Database (NHFD) is an online platform that uses real-time data to drive Quality Improvement (QI) across all 163 hospitals that look after patients with hip fractures in England and Wales. This report highlights key research carried out using data from the NHFD in 2021, and makes a number of recommendations to improve treatment and outcomes for patients with hip fractures. Recommendations Hip fracture teams should use quarterly governance meetings to review the quality and outcome of the care they provide. Where performance is significantly below average, units should formally discuss possible reasons for this within their regular MDT meeting, and plan a QI project to address it. Quarterly governance meetings should be taken as an opportunity for team members and trainees from all disciplines to make use of the NHFD website as a driver for QI; the new Quarterly Governance Tool is designed to help them do this. The NHFD recommends that governance meetings of surgical, orthogeriatric, anaesthetic, nursing, therapy and management leads should take place on at least a monthly basis. Monthly governance meetings should be used to plan appropriate QI interventions, and to monitor the impact of these using the real-time data reported in the NHFD run charts. Hip fracture teams should use their KPI caterpillar plots to identify better-performing neighbouring units, so they can share best practice and network with them in designing QI work. Hip fracture teams should use KPI 0 as a marker of initial care and a driver to improve the provision of local anaesthetic nerve blocks and fast-tracking of patients to an appropriate ward. Performance should be considered alongside the figures for their unit in the Anaesthesia run chart and Assessment benchmarking table. To help patients avoid further fragility fractures, hip fracture team governance meetings should review KPI 7 alongside their Bone Medication Table and arrangements for 120-day follow-up. Hip fracture teams should signpost patients, their families and carers to the NHFD website resources designed to help them understand their care and recovery following a hip fracture. Hip fracture teams should use monthly governance meetings to review their policies and protocols, and to compare these with those in other units as described in the Facilities Survey. Hip fracture teams should minimise inequalities in health care; specifically by reviewing whether support and information are provided in formats and languages appropriate to their patients.

Closed-loop communication—when every test result is sent, received, acknowledged and acted upon without failure—is essential to reduce diagnostic error. This requires multiple parties within the healthcare system working together to refer, carry out tests, interpret the results and communicate them in language the patient can understand. If abnormal test results are not communicated in a timely manner, it can lead to patient harm. This Quick Safety case study looks at the case of a 47-year-old school teacher who had a screening mammogram. The radiologist identified a suspicious area of calcifications, which required follow up. The patient’s GP was not on the same electronic medical record (EMR) as the imaging centre and, because of front office changes, missed the notification to follow up. The patient was told that the radiologist would contact her if the results were abnormal and therefore assumed she was okay. A year later when seeing her GP, the patient was told that she needed follow-up testing and that she had stage 3 cancer. Her lesion had grown significantly, and she now required surgery, chemotherapy and radiation for advanced breast cancer. The case study suggests safety actions that should be considered to prevent this error from happening again.

This study in the journal Current Problems in Diagnostic Radiology aimed to explore the perspectives of radiology and internal medicine residents on the desire for personal contact between radiologists and referring doctors, and the effect of improved contact on clinical practice. A radiology round was implemented, in which radiology residents travel to the internal medicine teaching service teams to discuss their inpatients and review ordered imaging. Surveys were given to both groups following nine months of implementation. The vast majority of both diagnostic radiology residents and internal medicine residents reported benefits in patient management from direct contact with the other group, leading the authors to conclude that this generation of doctors is already aware of the value of radiologists who play an active, in-person role in making clinical decisions.

In July 2022, HSIB launched a national investigation into the safety risk of clinical investigation booking systems failures. Specifically, the investigation explores the use of paper or hybrid booking systems and the production of appointment letters. This interim bulletin highlights a safety risk identified by the investigation and presents a safety observation for the attention of NHS care providers. Some vital NHS services still use paper-based or hybrid systems, which may have been developed over time and could leave unintended gaps where patients can be lost in the system. The reference case for this investigation is the experience of a patient whose magnetic resonance imaging (MRI) scan was not rescheduled following a cancellation, leading to a delay in the diagnosis of cancer. Hybrid systems were in use, which did not assist staff to keep track of patients. Additionally, the hybrid systems in use did not allow appointment letters in non-English languages to be produced. Safety observation It may be beneficial for NHS care providers to explore options for the translation of written appointment communications, including pre-attendance guidance, for patients whose preferred written language is not English.

This letter from Dr Robert Farley, President of the Institute of Physics and Engineering in Medicine (IPEM) to Karen Reid, the Chief Executive Officer of NHS Education for Scotland (NES) highlights that lack of funding for Clinical Scientist training places is putting patient safety in Scotland at risk. Dr Farley says, "We understand NHS Education for Scotland are proposing funding that equates to less than a single training post in medical physics and clinical engineering in 2023. ‘This is despite the Scottish Government's Chief Healthcare Science Officer’s public acknowledgement of the importance of training. "Scotland currently has a 10 per cent Clinical Scientist vacancy rate across the medical physics specialisms. This equates to seven vacancies in radiotherapy, three in nuclear medicine, four in diagnostic radiology and radiation protection. These posts are critical to supporting diagnostics and cancer treatments."

This article describes how a radiology group in Arizona allegedly missed dozens of breast malignancies, some of which were obviously cancer. Breast surgeon Dr Beth Dupree and a team of expert radiologists identified 25 missed cancer diagnoses that required either surgery, chemotherapy, radiation or a mastectomy at Northern Arizona Healthcare between 2016 and 2018. The team felt that there was a high chance of the number of women with missed cancers being higher than those uncovered by the review, but their request to expand the investigation did not go ahead.

The Sentinel Stroke National Audit Programme (SSNAP), which assesses the care provided for patients during and after they receive inpatient care following a stroke, has published its ninth annual report. Based on data from April 2021 to March 2022, the report aims to identify which aspects of stroke care need to be improved with a particular focus on changes in stroke care over the last two years and the ‘roads’ that need to be followed in order to restore the quality of care. SSNAP measures the process of care against evidence-based quality standards referring to the interventions that any patient may be expected to receive. These standards are laid out in the latest clinical guidelines and include: whether patients receive clot busting drugs (thrombolysis). interventions for clot retrieval (thrombectomy). how quickly they receive a brain scan. how much therapy is delivered in hospital and at home.

In this opinion piece in The BMJ, consultant radiologist Giles Maskell examines changes to the ways in which medical imaging is used in the health service. He states that imaging used to be ordered, when necessary, at the end of a diagnostic process, whereas now many doctors are asking for scans before they will see a patient for the first time. The article highlights some of the implications of this shift in practice, including on screening service capacity and on the interpretation of test results.

This article outlines the results of a recent investigation by the Parliamentary and Health Service Ombudsman (PHSO) which found that a 65-year-old man died after doctors failed to notice serious abnormalities on his X-ray. The patient, known as Mr B, was admitted to University Hospitals Birmingham NHS Foundation Trust in May 2019 after being unwell for several days with abdominal pain and vomiting. An X-ray of his abdomen was taken, which two doctors said did not show any apparent abnormalities. The following day Mr B's condition deteriorated and he suffered a heart attack and died. The PHSO investigation found that the Trust failed to notice a blockage in his intestine on the X-ray. Because of this failure, Mr B did not receive treatment that could have saved his life.