Every minute counts in stroke care. A patient experiencing an ischaemic stroke loses approximately 1.9 million neurons per minute without treatment. The difference between a good outcome and permanent disability can come down to how quickly a clinician can interpret a scan and make a treatment decision. Across the UK, 100,000 people are affected by strokes every year. The pressure to get diagnosis right, and fast, is immense.
That clinical reality is precisely what the NHS SBS AI Software in Neuroscience for Stroke Decision Making Support framework will address. It is also why the evidence bar for suppliers is exceptionally high. Right now, only four companies hold a place on this framework.
If you are developing or supplying stroke AI software and want to be among them, or among the suppliers who secure a place on its successor, the question is not simply whether your technology works. It is whether you can prove it, convincingly and compliantly, in writing.
This framework is not a standard procurement vehicle. It was developed by NHS SBS in close collaboration with NHS England, NHS Improvement, all 20 Integrated Stroke Delivery Networks, the Academic Health Science Network, NHSX, and the Care Quality Commission. That development process reflects the clinical seriousness of what it covers.
The framework provides AI software that supports the analysis of CT and MRI imaging for ischaemic and haemorrhagic stroke. It is designed to help clinicians make faster, more accurate decisions about time-critical interventions, such as thrombolysis and thrombectomy. This is because it can reduce interpretation time from as long as 30 minutes for remote manual review down to seconds. It is the only framework of its kind in the UK, open to all NHS organisations across the country. Due to these factors, it won the 2024 Public Finance Award for digital solutions partnership.
For suppliers, that pedigree matters. It means evaluators are not generalist procurement professionals assessing a broad IT product. They are specialists who understand stroke pathways, imaging protocols, and clinical workflow in depth. Your application will be read by people who know the difference between a well-evidenced solution and a well-marketed one.
The current framework is due to expire on 12 November 2026. But the renewal is not simply a like-for-like replacement. NHS SBS has confirmed it is developing a new Healthcare Artificial Intelligence Solutions framework. This will consolidate this framework and the separate AI Imaging and Radiotherapy Equipment framework into a single, significantly broader offering.
Provisionally structured across six lots, the new framework will cover:
Stroke AI suppliers will most naturally sit under Lot 1. However, the wider scope of that lot (spanning neurology, oncology, cardiology, urology and ophthalmology) means the competitive field will broaden considerably.
No tender notice is available yet. The framework is still at an early stage of development, and NHS SBS has indicated that the structure may evolve following further market analysis. What is clear is that the window between now and publication is a preparation window. Suppliers who are already building and documenting their evidence base will be in a meaningfully better position when the tender drops.
It is worth pausing on that number. Out of the entire global market for stroke AI software, four suppliers currently hold a place on this framework. That is not an accident of low demand or limited awareness. It reflects a genuinely exacting qualification threshold.
The framework was evaluated against clinical standards shaped by the people who designed the National Optimal Stroke Imaging Pathway (NOSIP). Evaluators understand the full clinical pathway, from first presentation through imaging triage, thrombolysis decision, thrombectomy referral, and post-event monitoring. A solution that addresses one element of that pathway but cannot demonstrate broader clinical integration will struggle to score well.
There are also significant technical and compliance requirements. UK Conformity Assessed (UKCA) or CE marking, compliance with NHS data standards, Data Security and Protection (DSP) Toolkit certification, Cyber Essentials Plus, and compatibility with DICOM imaging infrastructure are baseline expectations, not differentiators. Meeting them is the price of entry. What determines whether you win a place is the quality of evidence you can produce above and beyond those thresholds.
Many AI companies can describe what their software does. Far fewer can demonstrate, in a structured bid response, what it has measurably achieved in clinical settings. Fewer still can present that evidence in a way that maps cleanly to evaluation criteria.
This kind of evidence is non-negotiable. Evaluators expect to see peer-reviewed literature, prospective trial data, or validated real-world outcome studies that establish the diagnostic accuracy of your solution. Marketing claims, pilot anecdotes, and generic statements about AI performance in stroke care will not score well. If your validation data comes from international studies, be prepared to address its applicability to the NHS clinical context explicitly.
Real-world NHS deployment data carries particular weight. How has your solution actually performed in NHS trust environments? Quantified improvements in time-to-treatment, demonstrated reductions in door-to-needle times, evidence of workflow integration with existing radiology systems: these are the kinds of outputs evaluators want to see. Generic deployment statistics are less persuasive than a clearly evidenced case study from a named NHS trust that shows before-and-after clinical outcomes.
Alignment with the National Optimal Stroke Imaging Pathway is a scoring opportunity that suppliers frequently underuse. NOSIP sets out the clinical standards against which NHS stroke services are measured. A bid response that references your solution’s contribution to specific NOSIP standards, and evidences that contribution, speaks directly to what NHS evaluators are trying to achieve. A response that does not mention it at all is a missed opportunity, and may signal to evaluators that the supplier has not engaged deeply enough with the clinical landscape.
You need to address interoperability and data governance with precision, not broadly. Vague assurances about data security or system compatibility are not sufficient. Evaluators want to understand how your solution integrates with NHS imaging infrastructure (PACS, RIS, DICOM workflows), what your data processing architecture looks like, how you handle patient data under UK GDPR, and what your approach to NHS login and identity management is. The more specific and technically grounded these responses are, the more credible they read.
Implementation and change management capability is an area where technically strong companies often underinvest in their bids. Deploying AI in a clinical environment is not a software installation; it is a change programme. Evaluators want evidence that you can support clinical teams through adoption: training, workflow redesign, ongoing clinical engagement, and escalation pathways when the system flags an uncertain result. If you have done this in NHS environments before, that experience should be front and centre.
There is a gap that appears in many technically excellent bids. It is the gap between what a company knows about its own solution and what actually lands on the page. It is surprisingly easy to write a response that is accurate and technically sound, but that fails to score well because it is structured around the product rather than the evaluation criteria.
A bid evaluator reads your response looking for specific things in a specific order. If the evidence they need to award marks is buried in the third paragraph, or implied rather than stated, or presented in the language of a product brochure rather than a clinical evidence framework, marks will be lost. Structuring each response around what the question is actually asking, and leading with the most mark-relevant content, is a discipline that makes a measurable difference to scores.
Case studies are particularly important on a framework like this, where clinical credibility is so central. Each case study should be selected because it mirrors the complexity, setting, and clinical context of the contracts being bid for. It should describe the challenge, the solution, and the measurable outcome. “Improved workflow” is not an outcome. “Reduced mean door-to-needle time by 23 minutes across six stroke centres over 18 months” is.
Before you submit, an independent bid review by someone outside the team that wrote the responses is one of the most valuable steps you can take. It is genuinely difficult to evaluate your own bid objectively. A reviewer who understands NHS procurement and bid evaluation can identify where the evidence is thin, where the response has drifted from the question, and where scoring opportunities have been missed, while there is still time to address them.
The transition to the new Healthcare AI Solutions framework is not just an administrative change. It is a structural shift in how NHS SBS is thinking about AI procurement. The broadening of scope under Lot 1 means that stroke AI suppliers will be competing alongside solutions from oncology, cardiology, and ophthalmology, all presenting their own clinical outcome evidence.
That changes the competitive dynamic. In a framework assessing only stroke AI, evaluators could look at your clinical evidence on its own terms. In a broader radiology and diagnostics lot, the quality and rigour of your evidence presentation will be compared against suppliers from adjacent specialities who may have different (and in some cases more extensive) published evidence bases.
The implication is to start building now. Document deployment outcomes systematically. Gather case studies from NHS trust implementations with quantified impact data. Engage with the NHS Long Term Plan objectives for cardiovascular and stroke care and ensure your positioning aligns explicitly with them. And develop a clear bid strategy and positioning for how you will differentiate within a broader lot: not just from other stroke AI providers, but from the wider AI imaging market.
NHS SBS has also indicated it would welcome input from prospective suppliers in shaping the new framework. If you have not already engaged with that process, doing so is worthwhile. This will enable you both to influence the framework design and to build relationships with the category team ahead of the formal tender.
The framework application process for a framework like this demands more than good technology and good intentions. It demands responses that meet a high clinical evidence standard, presented with the clarity and structure that bid evaluation requires. That combination is exactly where specialist bid writing support makes the difference.
At Bidding, we work with suppliers in technically complex, high-stakes NHS procurement, helping them turn genuine capability into bids that score at the top of the quality criteria. We know what it takes to present evidence in the way that evaluators reward.
Whether the Healthcare AI Solutions tender has just been published or you are building your evidence base ahead of it, the earlier we work together the stronger your application will be. Get in touch with Bidding and let’s make sure your bid reflects everything your technology can do.
Bidding Ltd © 2026
