Mechanical thrombectomy (MT) is the most effective evidence-based treatment for acute ischaemic stroke caused by large vessel occlusion and could transform outcome for the 10% acute stroke admissions who could benefit. However only 1 in 5 eligible patients currently receive this time-sensitive treatment. Increasing this figure is a priority highlighted in the NHS Long Term Plan. One of the barriers is lack of neuroradiology imaging expertise at smaller stroke units where most patients first arrive leading to delays in diagnosis and referral. Fast identification of large vessel occlusion and reliable quantification of the extent of ischaemic injury is crucial to the robust delivery of a mechanical thrombectomy service, so that the patients can be referred to a specialist centre within the critical time window to receive the treatment. COVID-19 posed additional challenges, with Emergency Departments busy with suspected virus patients, and physicians redeployed across the hospital, creating the need for more remote decision making. The Covid-19 pandemic has highlighted the need for more support for rapid, efficient, clinically effective remote working.
What we did
An innovative AI-enabled decision support tool (e-Stroke Suite – Brainomix) helps clinicians quickly and accurately decide the type and severity of stroke, and the most appropriate treatment. Initially introduced before the pandemic, it has enabled new ways of working in a world disrupted by Covid-19 eg if one hospital is overburdened another can help out by interpreting brain scans without delay. Previously, these scans had to be reviewed by a specialist in limited locations. Now they can be interpreted by a non-specialist, shared securely and seen within a few minutes of being processed – anywhere, any time – and advice given immediately.
Brainomix worked with the Health Innovation Oxford and Thames Valley, one of the five OAHP Partners, establishing a Thrombectomy Innovation & Transformation (TITAN) quality improvement team to support the introduction and real world evaluation of the AI tool into the hyperacute stroke pathway in the Thames Valley to improve the quality of stroke care by enabling a robust thrombectomy referral pathway. Initially introduced at the Royal Berkshire Hospital, Reading, it was extended to all five primary stroke centres (Aylesbury, High Wycombe, Milton Keynes, Northampton and Swindon) and the regional tertiary neuroscience centre in Oxford in summer 2020, making the Thames Valley the country’s first AI-enabled regional stroke network.
Impacts and delivery
Implementation has led to substantial reductions in Door in Door Out times – a key metric of stroke system performance for delivering MT, DIDO reduced from 140 to 79 minutes at RBH (Dr K Nagaratnam) following adoption of Brainomix. The regional deployment of the software, enabling an integrated care model for thrombectomy referrals, demonstrates its scalability. By integrating seamlessly into the existing stroke treatment pathway, and by operating automatically, the technology can provide powerful results for stroke clinicians without interrupting the normal flow. Impact on the system is speedier decision-making and more rapid treatment of stroke patients leading to reductions in death and disability. Secondary benefits include enhanced productivity and establishing a virtual network supporting remote working and social distancing. The software can be deployed virtually or installed at a hospital with a server. The main potential constraints relate to IT and data governance. We have worked with national and regional authorities to address these concerns and issue standardised reports for each hospital. https://vimeo.com/561831060 which demonstrates the impact and benefits for patients.
Value and potential implications for OAHP
Health Innovation Oxford and Thames Valley is leading wider expansion across the NHS England South East region and evaluating the impact of the Brainomix AI-enabled software on clinical outcomes, stroke care pathways, workforce productivity and healthcare professionals’ views of the system.
Brainomix was founded more than decade ago by leading University of Oxford (OU) stroke academics, including the current CEO Dr. Michalis Papadakis, and Prof Alastair Buchan, who developed the ASPECTS scoring method for stroke patients. Since spinning out from OU in 2010, Brainomix have developed award-winning, AI-powered imaging biomarkers and software solutions, assisting physicians across the world to make better life-saving decisions. Its flagship technology, e-Stroke, is now used in more than 250 hospitals across 30 countries, having processed more than 500,000 patient scans.