The Health Bill is currently progressing through Parliament, and, among the digital provisions travelling with it is the Single Patient Record (SPR), a unified record drawing a patient's data together across every care setting. I've spent much of my career on similar programmes and have found that ambitious legislation, like the Health Bill, is really only the first phase of change. The finer details will be decided in secondary legislation and implementation is where we really delve into how new technologies impact day-to-day life for clinicians and patients.

The NHS desperately needs an SPR, but for the SPR to achieve its ambitions, it needs structured data captured at the point of care. Investing in the right tools, that deliver the right data, therefore, becomes more important than simply choosing the cheapest option.

What goes into the record matters

Although the SPR has not yet been introduced into the NHS, now is the time for decision makers to begin setting the foundations for its success. DHSC has set ambitions for greater access to linked patient data by 2027, with full visibility for patients of their shared care record by 2030. This means redesigning systems with the SPR in mind and AI tools must be considered as part of this process. By procuring AI tools without considering how they would interact with the SPR, the NHS could be setting itself up for failure.

For instance, back in May, NHS Shared Business Services opened a framework worth up to £900m for healthcare AI. The AI tools that are procured now will provide the data that fuels the SPR later down the line, making it more important than ever to invest in the right tools for the job. Of course, no one sets out to buy badly, but some procurement frameworks are not designed with long-term value in mind, meaning that price wins by default.

Documentation tools chosen for their low price often produce free text rather than structured, coded data, with their output usually moved manually between systems. When a note only needed to stay where it was written, this approach wasn't a problem. But, when the SPR comes in, the bar will be set much higher.

The SPR will only work if the data fed into it is structured enough to be searched and relied upon by clinicians at every stage of a patient's journey. Built from a mess of notes and manual transfers, it will carry errors with the patient across the NHS, which could lead to misdiagnoses or delays in treatment, ultimately jeopardising patient safety. For NHS leaders, this also creates a new and expensive problem guaranteed to outweigh the appeal of any short-term cost saving.

From procurement to a change management programme

Ensuring the SPR achieves its ambitions also requires a shift in mindset. When considering any big technological change in the NHS, or even a large-scale procurement, we should look at it as a change management programme. Introducing new technology is really only the first step; everyone who is expected to use the new tool should be brought along on an adoption journey over a defined period of time. In my experience, the roll outs that worked best were those where the supplier shared responsibility for the implementation alongside NHS teams, rather than simply handing over a licence and a training module. Clinicians will ultimately be the ones that decide what they're willing to work with and that trust is earned.

This approach is particularly pertinent when it comes to introducing AI. For many healthcare organisations, AI remains relatively new, compared to the systems they have been using for decades. Yet, research has shown 90% of NHS healthcare professionals have already tested AI in their clinical work. It is high time, therefore, that healthcare organisations ensure their clinicians have access to high-quality AI tools and their employees are equipped with the skills and knowledge to use them effectively.

Ensuring suppliers are ready for the change

For technology suppliers, the questions will become increasingly practical as the NHS ramps up to begin introducing an SPR in 2027. What does your governance consist of? Does your output arrive structured or as free text somebody needs to sort later? Will your tool integrate through open standards, or a locally managed workaround? Who carries onboarding through the first week or two? How will performance be reported? And what will this platform be asked to do in three years, as documentation, guideline access and patient co-ordination converge over these various tools?

Fortunately, there are already examples of what good looks like. Across more than 15m NHS consultations, ambient voice technology has returned over 4m hours of clinical time to frontline teams and cut documentation time by around 85% in high-throughput hospital services. Procurement's job is to capture this kind of success and to keep checking it holds up. That doesn't mean buying the cheapest option but ensuring AI tools are delivering value for patients throughout their life cycle. The SPR, and the structured data it requires to work well, will start to show any cracks that remain.