For a recent HTN Now panel discussion, we were joined by a panel of experts from across the health sector to discuss how general practice, PCNs, and ICBs can utilise data and leverage technology to support operational efficiencies and improvements across primary care.
Panellists included Kathryn Salt, assistant director of primary & community care, data and analytics for the Transformation Directorate, NHS England; Dr Shanker Vijayadeva, GP lead, digital transformation for the London region at NHS England; Dr Sheikh Mateen Ellahi, GP and practice partner at ELM Tree Surgery and South Stockton Primary Care Network; and Max Gattlin, digital consultant at X-on Health.
We first turned to our panellists to offer a brief introduction to their role, along with some initial thoughts on the topic area. Kathryn started us off by sharing that her work with NHS England, and prior to that NHS Digital, has spanned 18 years to date. “I’ve spent the last ten years or so focusing on primary care data and analytics,” she said, “and the landscape has changed a lot in that time.” Kathryn also highlighted current projects including the first-time collection of cloud-based telephony data, work in the community sector and linking to general practice, and wider work on modernising general practice.
Shanker told us about his roles with North West London ICB and with NHS England for the London region focusing on primary care, saying, “I feel like I do a bit of everything!”. Recently, he went on, “that’s been around things like pharmacy, registration, and automation”.
After starting out in primary care “five or six years ago”, Max told us, “I’ve worked on deployments covering things like online consultations, cloud-based telephony, and as we’ve developed more products and services to support digital transformation in primary care, a clear focus for us became the appropriate use of data.” That includes looking at “improving digital efficiencies, and how we can deliver that based on human efficiencies, utilising data on a broad spectrum to include all modes of access and find ways of putting in more efficient pathways to ultimately improve outcomes.”
Mateen highlighted his work leading his practice’s tech team, as well as his involvement in the private sector, and his upcoming role with the NHS Clinical Advisory Group for NHS England. He shared details of a couple of innovative projects currently underway at ELM Tree Surgery, saying: “I’m quite proud of the work we’ve done on improving patient outcomes and quality, and we want to keep improving, so we’re looking at things like using our new telephony system to make sure patients gain access to the reception team and then the GPs, within one minute of them calling.” Data analysis has helped the team offer 24/7 access for patients, and appointments “whenever they want”, he added.
Utilising data to optimise operational processes and drive change
When it comes to utilising data to optimise operational processes, Max echoed Mateen’s point about the importance of driving access to care. “It’s about offering assessment of need at first point of contact,” he said, “and one thing we did was to look at data to make sure patients were getting that access, and how they were getting that access.” In doing so, the team noticed that “whilst there had been incredible work on deploying the online consultation tool, lots of the demand was landing in our GP capacity’s lap”, he went on. By looking at data provided by patients, “we saw that there was confusion around which tool to use”, he said, “and 12 percent of patients were ending up with repeat prescriptions. By signposting them to the NHS app instead, the patient gets faster care, and it frees up capacity within the surgery”.
Mateen talked about similar work in his practice, highlighting a project which helped the practice to understand not only which patients were calling in, but what kind of appointments were needed, and which medical professional would be best suited to meet their needs. “For two weeks in the middle of winter we employed locums,” he continued, “and found out what demand actually was at the busiest time of year, analysing what times and days patients were calling, as well as what type of appointment they needed. We then recruited the amount of reception staff needed to ensure calls were answered in around one minute, as well as recruiting the types of medical staff most suited to the appointment need.” To actually overcome the problem around access, he summarised, “you have to understand what the demand actually is”.
“To mention something slightly different,” Shanker said, “we’ve done something looking at Healthwatch data at borough level, which tends to be more qualitative, comparing patient’s experiences and producing reports on differences – often we don’t compare feedback between practices, and when you see that side-by-side comparison you can see the story behind that raw data.”
From an NHS England perspective, “our goal is to ensure that we’ve got the data to tell the story that everyone on this call is probably familiar with about the demand”, Kathryn told us. “It sometimes feels like we collect a load of data and then publish it on our website and that’s it,” she went on, “but it’s just giving concrete evidence to that pressure that general practice is under”. In terms of efforts to optimise that, Kathryn cited the Modern General Practice initiative and the Capacity and Access Improvement plan, “which are there to find areas where practices need a bit more support to use their online consultation system or cloud-based telephony more effectively. There’s a lot more to do to make that data we’ve collected more usable, operationally, for everybody. Having a common set of metrics that everyone, including suppliers, can use and provide data around, would give us those national standardised stats we can start putting out.”
Standardisation of data collection and key metrics would be an integral part of moving forward in this space, Max agreed. “That would help us share those successes and avoid pitfalls in decision-making – we’re doing work at the minute looking at the Modern General Practice’s more recent pilot across 22 PCNs, looking at a standardised way to collect that data so it’s benchmarked across the whole cohort, so we can then share that in a standardised way,” he said.
How can data be used to overcome challenges around transforming primary care?
When data is being collected, it’s important to think about what the aim is, Shanker told us, “which is usually around change”. It’s then essential to consider whether it’s appropriate for the audience making that change, he went on, “as well as whether it’s accurate, timely, usable, and all those things”. It can be difficult to get the data we need, he considered, “and often there are errors or challenges within that data, or we collect data that is available to us, rather than the data we want but can’t get to”. Considering the audience is also integral, he continued, “because if you’re talking about primary care – they may not even have the capacity to listen to you giving them the data, and that should be a two-way conversation, because everyone will have a different context which affects how they interpret that.”
Usability is also a consideration, according to Shanker, “and you have to think about how user-friendly it is to actually put it into an action; because if the data is just on a strange dashboard on a population level, without clear guidance as to which patients it relates to, it’s not going to be useful.” It ultimately comes down to having the resources and time to understand what’s been presented, understand the change, and implement the change, he said, “and there can be a huge thing about relationships and trust when it comes to data, so we should also consider that emotional context”.
Max agreed with the fact that a lack of capacity makes it difficult for practices to consider the case for change, saying that his team often focuses on allocating that time, as well as “giving them a view of what the next 12 months could look like”, and going through some examples of where similar change has been successful. Like Mateen, he said, “we worked with a practice to tackle the 8am rush – they were at about three minutes at the time for patient calls to be answered, and we put processes in place like automated signposting within the call flows or digital assistant, as well as on the website, to help reduce that demand”. This process has resulted in getting that call waiting time down to “around 30 seconds”, according to Max, “and the transformative part of that is that now we have time to retrain and redirect staff into some of those proactive and preventative care models”.
When you’re talking about change, the cycle really starts with culture, Mateen told us, “and whilst that often gets forgotten about, getting that right is key”. His recommendations included involving key stakeholders within the practice, having conversations around why you’re collecting the data and how that is going to benefit the practice, as well as how it will ultimately benefit each individual member of that practice’s staff. “That really does make life so much easier,” Mateen said, “because otherwise you will face barriers and questions further down the line.”
Another area for consideration is around KPIs, Mateen shared. “I think it’s important for practices to always have a mission and a vision – for us that was trying to ensure patients can be seen whenever they want, and getting them the best care possible to meet their needs.” The practice’s new telephony system has “really helped”, he went on, “in actually finding out the pain points, the barriers, and which times are most difficult for our receptionists to be handling those calls”. Looking to clinical outcomes, Mateen talked about the prospect of measuring those by analysing whether patients were “coming back for the same problem”, with the practice also monitoring that data to understand more about what is motivating patients to keep coming back.
Ensuring consistency around data collection and coding is challenging, particularly when reception and administrative staff are overburdened, Mateen reflected, “and then it’s also important to audit and re-audit that data, to take a second look at areas where perhaps there hasn’t been an improvement, and why that might be.”
Responding to a question from our live audience, Kathryn also shared progress on telephony data, saying, “we’ve received data for around 50 percent of GP practices now via a couple of the telephony suppliers, and we’re expecting another 20 to 30 percent to land with us in January; it’s all looking good so far, we’ve collected about two months of data, and all looks as expected.” Once the quality of that data has been checked, she continued, “the first step is to share it with the ICBs, ideally through the Federated Data Platform if possible; and we’ll look to get feedback on that, because it’s important our data reflects what people actually see from within the system.” Kathryn also shared the link to the General Practice Access Data Hub, for those wanting to keep an eye on upcoming data.
Utilising data to move from a reactive to a proactive model in primary care
Looking at how data can help with the move from reactive to proactive care, Shanker said, “it’s often about health promotion, health awareness, and early detection”. In national contracts, there isn’t enough attention paid to this topic in primary care, he considered. “Examples could be in the core GP IT systems, around early detection of things like CKD, cardiovascular prevention; it’s about how you make it user-friendly, and what we’ve seen change recently is around digital tools, around recall and automation, booking systems, so you’re making it easier to do the workload of prevention,” he said.
On challenges, Shanker told us that “it can be a bit muddy when it comes to who’s responsible for extracting and delivering data”, and that getting communication right across the workforce can be difficult. “We talk about how difficult it is to change behaviour of our workforce,” he said, “and with patients that can be even harder. Digital tools and social media can help achieve that at-scale.”
Mateen thanked Shanker for sharing these points around communication and engagement, adding that at ELM Tree Surgery his team has begun to automate the filing of results. The number of appointments made where practices reach out to patients to make an appointment following an abnormal result is “quite high”, he continued, “and so we’ve put protocols in place whereby patients receive a message about their result and the next steps or tests required, and if they call to find out more, our receptionists can help answer their questions about why those investigations are happening, because we make sure they have an awareness of the whole process from the beginning.” Compared with before these protocols were put in place, “there’s a 90 percent decrease in the number of appointments being made due to filing away results”, Mateen shared, “and that’s significant when talking about capacity and demand”.
Travelling toward a risk stratification and segmentation level of patient care, where patients can be risk stratified in accordance with their clinical data, past medical history, demographics, and social determinants of health; Mateen said “ultimately I think what’s going to happen is we’ll have low-risk patients, moderate-risk patients, and high-risk patients, to help us in directing our preventative interventions”. Automation could help “keep those patients close”, he went on, “and you could have a RAG system to keep an eye on patient’s blood results and other metrics and get a better insight into when that intervention is needed.”
“I’m going to follow-up Shanker’s sentiment about improving the interface between primary, secondary, and community,” Max said, “because Kathryn’s right about data needing to be accurate, but when we see data trends it’s about knowing your population, your demographic; and preventative doesn’t always have to be outbound from the surgery, it could be utilising community services.” Max’s team is currently looking at data trends across ICBs, and with the right digital tools, “we could potentially get lots of emerging technologies to query large datasets on things like mode of access or intent, to find out more at that system level and strengthen those relationships for preventative care and patient engagement.”
In the short-term, data could be the key to tackling things like ordering prescriptions online, Max told us, “because if from the data we can see that only one percent of patients are accessing that service, we can look at how we can increase engagement around that and build that awareness, and digital tools can be pivotal in helping us do that.” Self-service also has the potential to be “one of the biggest solutions to managing demand and capacity, or at least finding a balance there”, he went on, “so it’s about checking you’re leveraging the tools available like the NHS app, which is completely free, and then building automated pathways around patients.”
Looking ahead
In terms of what he would like to see in this space in the next year or two, Mateen talked about his excitement at practice-level about the potential to introduce AI-powered decision-support systems to help personalise patient journeys. “It would be great to look at whether, based on historical data and ongoing assessments, we could get personalised reminders for patients with diabetes or hypertension, so we can provide more patient-centred care,” he asked, “and use automation to save time in processing data and automatically updating patient records, to reduce administrative workloads.”
“Absolutely,” Max responded here, “and we’re looking at implementing tools like this, but one issue we’re running into is the data capture side from a surgery level – if we’re going to leverage data, we need to get that data input, and it needs to be true and accurate at a surgery level.” Getting that data right will be key to capabilities when it comes to implementing things like AI and automation, he said.
“Often we don’t join the dots together,” observed Shanker, “because we have things like automation which can save time, but in the primary care space everything is about resources, and you need resources to fund all of this”. At a primary care level, “even just looking at AI in transcribing and capturing information would be transformative”, he went on, “and I think we still have to respect that because we are all one system, there’s a risk that primary care is ahead of secondary care and community; and we need to try and rollout things across the system.”
“As data people we want the source data to be really good,” Kathryn said, “and GP appointment data is a good example of that – we hear a lot that it isn’t accurate or doesn’t reflect what people see, and I think it’s this balance between us not wanting to dictate how practices use digital systems like their appointment books, but also recognising that that means everyone is using it very differently.” For some practices, “it’ll look like they only have one appointment in a day, because they’ve just booked the whole day out, probably to have a really effective triage”, she highlighted, “but when that’s reflected in the data it’s difficult to deal with.”
We’d like to thank all of our panellists for their time, and for sharing their insights with us from across the primary care data space.
