HTN was joined for an interactive webinar on emerging pilot projects and innovations by Shanker Vijayadeva, GP lead – digital transformation in primary care, London Region, NHS England. With the expert help of Shanker, we explored examples of innovation, emerging technologies, opportunities, challenges, and the process of scaling successful innovations across the NHS.
Explaining his background as a GP, working at ICB level as well as in his role with NHSE for the London region primarily focusing on primary care, Shanker shared examples of current projects including early piloting of AI triage through the NHS App, a project exploring alcohol screening for teenagers using a digital questionnaire, and improving the workflow process of diabetic retinal screening.
Initial thoughts on innovation
“Innovation isn’t always about new technology,” Shanker reflected. “With the alcohol screening project, the digital stuff was more in the background.” On the other hand, the NHS App triage pilot has involved lots of digital challenges, making interface changes and deploying a digital online consultation platform, he noted.
Shanker talked about potential challenges around introducing innovation, including evaluation and measuring benefits, as well as funding and resources. “We know that these things are in a very difficult place at the moment as the NHS restructures,” he said. “That’s your first bottleneck, along with capacity, and then there can also be other contractual issues, transition periods, and so on to think about.”
Hearts and minds is another key factor, according to Shanker, with having the backing of stakeholders and those on the frontline being essential to success when introducing a new innovation. “It’s important to have their hearts first believe in it, then their minds on the workload, funding, interoperability, and all of those concepts,” he considered. “Scaling is something we think about at the beginning for the business case, but often we don’t think through the detail and it becomes an afterthought, which can quickly fall apart if the resources run out.”
A comment from our live webinar audience talked about framing innovation as meeting a need requiring a new approach, whether digital or not, rather than innovation for innovation’s sake. “I think that’s exactly right,” Shanker agreed, “I think there’s the potential for it to be overhyped.” The alcohol screening project arose from a broader look at public health by borough, he noted, with an initial identification of the issue and considerations for how it might affect things downstream, informing dependency and affecting admissions. “We had to think about what we could do with no resources or funding, and we used our existing EHRs and digital products to message patients, using quick digital questionnaires, dashboards, reports, so organisations have it all done for them.”
Shanker told us how his daughter helped with the testing of the innovation, recording an audio clip of her feedback to help win over hearts and minds. “She also did a poll of other teenagers at her school, and the results suggested we should be doing this, even if they weren’t drinking themselves,” he said. “I think that’s a great example of innovation, where there’s a definite need, and it doesn’t have to be a huge change or mean introducing new technology; it can be just changing an approach slightly to make an improvement.”
Moving on to evaluating innovation, Shanker gave an example from his experience in London working on AI and automation pilots. “We had a pot of money to give to each area, and we had to design the evaluation process,” he shared. “My first thought was that the scale of the cost to properly evaluate it would be bigger than the actual grant. You need to be pragmatic and think about scope in relation to what you’re doing, then the design, and also bias – if you’re close to something and passionate about it, it can be hard to step back. There’s quite often a business case attached which shows there’s going to be savings or efficiency improvements, and I think we’ve got to manage some of the conflicts of doing that.”
Opportunities and further examples
Sometimes, it can be feedback from elsewhere or word of mouth that prompts the exploration of a potential new technology, process, or way of working, Shanker explained. “As an organisation, you might be doing a brand new thing and taking on that risk; in primary care we definitely have first follower practices, second follower, and third follower practices. The problem from a leadership perspective is how to balance the resources, because I think there can be unintended consequences if you’re chasing one innovation and neglecting other things that might sit in parallel.”
A good approach is to focus first on needs, and then look at what innovations align with those needs, rather than risking being distracted by innovations that might look good but fail to scale or serve the right purpose for a particular region or organisation, outlined Shanker. Highlighting an example from the live audience about QFit samples being rejected by the lab and work to find out what the problem is and offering up simple solutions for practices, he noted: “That might also be called optimisation or problem solving, but it can have more innovative benefit than some really expensive tool. Innovation, to me, can start with just looking at your original problems and making some small changes to have huge outcomes for everyone.”
Innovative approaches don’t just end with actual patient care, Shanker considered, but can also involve connecting and communicating with patients in new ways. A 19-minute walkthrough video of the Universal Care Plan he made for TikTok ended up attracting three thousand views, he reported, “and I think that’s an example of how you can be innovative in how you engage with patients, by just taking the risk”.
Picking up on a comment from the audience about innovation in self-scheduling of MSK physio via a patient portal, Shanker noted similar efforts in London to build physio, talking therapies, and mental health flows into the NHS App with self-referral pathways. “A lot of these services, particularly MSK and mental health services, are struggling with demand,” he said, “so it’s about how you can get to an acceptable flow, using screening questions, etc. Often, as a GP, you can tick a box and refer a patient to MSK, but they don’t really want it, and they won’t engage, so nobody benefits from that. I think we need to start making referral pathways more oriented around the patient so they can self-refer, and that will offer better outcomes for everyone.”
Perhaps the biggest topic for innovation at the moment is AI, considered Shanker. “I think the consideration needs to be safety, and I know we worry about this a lot, but I’m going to compare this to when the computer started to beat the Grandmaster chess players – as humans, we can make mistakes, and if AI can reduce that and improve patient outcomes, I’m definitely all for it!” AI has the potential to scan EHRs and the single patient record to help highlight important information, he went on, “and I definitely am a great believer in putting patients at the centre and allowing them to navigate and control their own health”.
Shanker also shared with us some learnings from work on diabetic retinal screening in London – innovation that he says “should have happened years ago”. In general, findings from those screenings are not coded well by most practices due to time constraints, he told us, but all that is needed to improve that is some small tweaks to the formatting of the report, to allow EHRs and document management software to read it. “If you actually wrote in the report the SNOMED code, your document management software could just read the text and convert it into a code. Now, I’m trying to solve that across the country – that’s another example of how innovation doesn’t necessarily need a budget or new technology; it’s optimising something that should have been done many years ago.”
Learning from failure
“Over the course of my career, when I’ve seen certain digital things not work at a particular point in time, I’ve realised that a few years later, with a slightly different approach, a slightly different technology, or a slightly different patient population, it can work to different levels,” Shanker reflected. “We need to be prepared to revisit old ideas.” This is just one of the reasons that we need to be prepared to share learnings, which is an area that could be improved, he stated. “We don’t share enough, and often, by the time that we do, the world has moved on.”
A comment from the audience suggested that it is important to “reframe failure”, adding: “A failure is just data on the way to better understanding needs and refining; I’ve been trying a test case to establish whether automation and AI would improve a business process, or is the human process optimal and we need to avoid over-engineering. Thus, providing an insightful framework for go/no go decisions at any point in time. For me, success is the robustness of the framework and not necessarily having a ‘product’. But this message can be lost, and the drive to have ‘products’ can frame the success/failure perception.”
“There’s a culture, isn’t there, where you either hide your failures in the early stages of pilots or you brush over them,” Shanker agreed, “and it’s important that we fail fast – we can’t waste taxpayer’s money; but on the other hand, sometimes we need to give innovations enough time to prove themselves, so it’s a difficult balance.” Failure doesn’t necessarily mean the end of the road, however, he continued, “and some of the greatest innovations have been created for one purpose, completely failed, and then been repurposed successfully for another”.
Another challenge when piloting is governance, according to Shanker, as it can be impossible to predict all the problems or safety issues until something goes live. “You can’t fully evaluate something until you do it in a controlled release fashion, and that’s where you have to keep revisiting your governance and reassessing your position on safety. That’s preferable to a big bang approach where you never end up going live because you get stuck in a governance loophole.”
Shanker discussed progress with AVT, and how every organisation tends to have to do their own DCB0160. “If they have to do it individually, does it really work?” he asked. “There have been some commercial or third-party solutions that try to help do the governance at scale, where organisations can share their updated documents across lots of different stakeholders and work collaboratively on things like hazard logs, so it’s more cloud-based, and the provider of the digital solution can get more value out of the governance and safety.” In the NHS, however, there is more chance of things sitting in someone’s OneDrive or SharePoint, he acknowledged, “and that’s something we can definitely do better”.
We’d like to thank Shanker, and our live audience, for taking the time to share these insights with us.
