Advancements in machine learning have led to the rise of pre-trained Large-Language Models (LLMs) that can be used to interpret, understand, and process medical data. As a demonstration of this technology, we have built a proof-of-concept (POC) tool that transforms unstructured and doctors’ notes and chest x-rays into a structured database.

What is the opportunity for healthcare providers?

According to some studies, up to 97% of medical data sits unused within hospital premises. This is a huge untapped resource that could help providers optimise pathways, improve patient care and unlock financial gains through better coding and saving clinical time.

The main downside, however, is that much of this data sits in unstructured formats, such as images. Extracting meaningful data from these can be time consuming and expensive, but our proof of concept demonstrates how the latest AI developments could overcome these barriers, allowing greater utilisation of data for hospitals.

Proof of concept using chest X-rays

Our tool harnesses a state-of-the-art open-source pretrained language model (Zephyr-7B) and an image to text model that extracts information from DICOM images (llava-1.5-7b). We have tested the feasibility of our approach by applying it to the MIMIC-CXR public dataset of chest X-rays and doctors’ notes.

Our POC is able to reliably take this complex data source, which contains rich medical information, and distil the details into predetermined fields of a table, leading to a simplified version of the original dataset.

Processing pipeline for X-Ray images and doctor’s notes.

Why are structured tables, such as knowledge graphs, more useful than their unstructured counterparts?

By compressing rich clinical information into salient indicators and storing it in tables, we can create structured datasets that relay important details about patient care, such as their diagnosis, procedures and outcomes.

In turn, researchers, auditors, and clinicians can easily draw insights from large numbers of patients using simple data analysis approaches, allowing them to understand how interventions impact patient outcomes and determine methods to improve patient care.

Transformed knowledge graph output from the discharge notes and multiple X rays of a single patient.

Plenty of use cases

A tool like this could be employed to validate secondary-use datasets which are currently manually created by clinical coding experts, such as hospital episode statistics (HES), a nationally mandated dataset in English hospitals. This approach is error prone and is subject to variation in coding methodology at different hospitals and countries. By cross-referencing the data with the structured datasets generated by this tool, we can improve the accuracy and completeness of the information, unlocking deeper insights into patient care.

A more operational use case could involve the scheduling of further scans or treatment. The structured output produced by the tool could be fed into another model to automatically triage patients on hospital waitlists based on the severity of their conditions. This would ensure patients most at need of urgent care or at risk of deterioration are seen quickly.

There is a recurrent emphasis on the need for Integrated Care Systems (ICSs) to address health inequalities, as highlighted by media outlets, conferences, and reports.

Addressing these disparities is a monumental task, especially for the newly established ICSs, which have been tasked with not only establishing new governance and strategies but also tackling an elective backlog and long-standing health concerns like health inequalities.

One crucial area of focus, particularly in relation to the Core20Plus5 mandates, is cancer, as we are aware of inequalities in access impacting diagnosis and survival rates. The ambitious objective set out by the NHS Long Term Plan is to diagnose 75% of cancers at Stage 1 or 2 by 2028.

Lung cancer, where 64% of patients receive a diagnosis at stage 3 or 4, is an excellent example that underscores both challenges and opportunities for ICSs.

Why avoiding emergency diagnoses is key

We examined publicly available data on lung cancer care. Patients referred by a GP are more likely to be diagnosed at an early stage than those in emergency settings.

This relationship can be seen by comparing NHS clinical commissioning groups (CCGs), where a 10% increase in GP diagnoses is associated with a 3% increase in early diagnoses (stages 1 and 2), when adjusting for confounding factors, as shown in Figure 1.

Figure 1. Source PHE 2018

Since patients are over 3 times more likely to survive more than 5 years when diagnosed at stage 1 compared to stage 3[i], detecting cases via referrals from primary care has a direct impact on lowering lung cancer mortality rates associated with a late-stage diagnosis.

Table 1. Source NCRAS 2015-16

Reducing variation in primary care referral rates

The challenge, however, lies in the unequal volumes of lung cancer referrals made in primary care, which vary dramatically across NHS regions. We have used Public Health England data and machine learning approaches to uncover the relationship between the number of GP referrals and the percentage of all lung cancer cases diagnosed from these referrals.

As seen in Figure 2, there is a clear positive relationship between the volume of referrals from primary care and early diagnosis, which is particularly strong for areas with low referral volumes.

It follows that if GPs with low referring rates could be supported increase referral volumes, there will be a high impact in driving earlier diagnosis and improving survival rates. This means that increasing referral rates would be very material for the NHS and its patients. In fact, if all ICSs were able to bring all the lowest referring GP services in line with the bottom quartile, as shown in Figure 3, we would expect 700 extra early diagnoses and 100 lives saved per year across the country.

How these findings turn into practical implications for ICSs

Variation in urgent suspected cancer referrals and early diagnosis rates is likely a combination of both GP organisation/behaviour and broader patient behaviour. For the former, it is well known that there are pressures on GP numbers and overall workload, which will impact access locally.

Nonetheless, there will be opportunities for ICSs to surface data on variation in cancer referral rates and work with practices to understand variation and support where necessary.

ICSs can also lead improvements by understanding how their local population, demographic and health system factors are influencing access. Although this highlights the complexity of the challenge, it also offers multiple sources of opportunity for systems.

In our experience, some of the key actions for ICSs to address the above are:

• Involve primary care networks (PCNs) and cancer alliances early into conversations about improving cancer detection – we are currently working with a cancer alliance on data-driven research to better understand the drivers of variation in the detection rate and the most effective interventions for addressing them.
• Provide practices and PCNs with tools to better understand their local population and their health needs (see here for a population health management dashboard we developed for Surrey Heartlands).
• Plan adequately for workforce, particularly in primary care, to make sure there is enough capacity to boost referrals and avoid workforce overwhelm. Given the falling numbers of full-time equivalent GPs, this is a priority area for ICSs and nationally.
• Assess secondary care diagnostic capacity, including modelling demand and capacity and promote system-wide initiatives such as new community diagnostic centres, implementing rapid diagnostic services and supporting mutual aid between trust, as we have discussed previously.

As the landscape of healthcare continues to evolve, ICSs have a crucial role to play. The responsibility lies with them to implement innovative strategies, utilise data-driven research, and ensure a robust primary care workforce.
With a concerted effort towards these goals, ICSs have the potential to significantly influence early cancer detection rates and, ultimately, patient survival.

[i] Characteristics of patients with missing information on stage: a population-based study of patients diagnosed with a colon, lung or breast cancer in England in 2013. C Girolam and others BMC Cancer (2018). Volume 18, Page 492

Key messages:

Half of all paediatrics A&E attendances are from children and young people from the most deprived areas
Worrying parallels are seen in the lack of primary care provision, and the mistrust of people from deprived areas in their GPs, partially mitigated by higher numbers of doctors in A&E
To address inequalities in healthcare provision, we need to understand the context – this is where data can help
Four key areas to investigate are: causes for utilisation, complexity of patients, the wider context and the potential benefit of novel initiatives.

Unrelenting trends

Trends in persistent health inequalities remain a key policy issue. The COVID-19 pandemic laid bare the drastic health inequalities that exist among populations served by the NHS. Here we want to focus on inequalities among children, specifically how to identify them and take practical steps to address them.

A&E attendances among children for the most deprived population are higher compared to the least deprived in the UK. Although measures stratified by IMD quintiles show disparities across Trusts, insights are limited as the data is not available at the patient-level.[1] Even pre-pandemic, in 2015/2016, the most deprived Children and Young People (CYP) overall were 58% more likely to go to A&E than the least deprived (Nuffield Trust).
In our data, almost half of all paediatric A&E attendance are accounted for by the two most deprived quintiles across children aged 0 to 17 years.

What is behind the disparity?

The British Social Attitudes Survey from 2019 found that parents with children under the age of 5 living in the most deprived areas were the most frequent users of A&E in the preceding year, and they perceived it most difficult to obtain a GP appointment compared to families living in less deprived areas. Additionally, they expressed less trust in their GP but tend to utilise the internet more often to self-diagnose (BSA).

The inequality in primary care provision across IMD quintiles is evidenced by the above chart, where the most deprived areas have one whole less GP FTE per 10,000 patients compared to the least deprived. Sadly, this gap has been widening, rather than closing, as equitable workforce distribution remains a major challenge. Perhaps as a way to mitigate this and respond to the increased A&E utilisation, there are considerably higher numbers of A&E doctors across all grades in more deprived areas. Although this intervention responds to the observed differences in A&E utilisation by deprivation, workforce restructuring strategies are not enough without addressing the underlying issues causing greater inequalities (King’s Fund).

Why is this problematic?

A&E is often not the right place to provide care to children who may have more complex needs that span school settings, as well as community and secondary care (asthma, diabetes, epilepsy). The busier a paediatric A&E department, the less well suited to fully understand parents’ concerns and provide education to prevent re-admission.

The higher utilisation in A&E services reveals that the current efforts to move care in the community are failing children in most deprived areas, as A&Es respond to a lack of primary care and community investment. The increased demand for emergency services in more deprived areas is likely due to a combination of differences in need and issues with adequate primary care provision and utilisation.

Uncovering what drives higher A&E demand is paramount, especially if it is matched by lower primary care utilisation. This is because the long-term risk is fragmented children care, as A&E does not provide continuity and a holistic assessment of the wider context which is often needed in CYP.

How data can help uncover the root cause.

Data is a powerful tool that can support this quest. Here we propose four areas for investigation that are accessible to all providers, to start building solutions:

• Identify the causes for utilisation: Devise a clear picture of the main causes driving A&E utilisation to target and streamline services. Asthma is often cited as a primary cause of healthcare utilisation among children. This opens the opportunity to provide services in schools, such as the My Asthma in School (MAIS) intervention which conducted educational and self-management workshop to children. The intervention was successful: 91.4% of participants (n=1814) reported the workshop changed their perspective on asthma (PFS).

• Determine the complexity of patients: Are patients in more or less deprived areas presenting with more severe conditions requiring more treatment? Use admission rates from A&E, extent of treatment provided and attendance rates in minor/majors to find out. Parents of children living in areas of higher deprivation may wait longer to seek healthcare as other commitments take priority, such as working. On the other hand, a substantial proportion of ED attendances are non-urgent, especially in younger children (EMJ). Education in school, communities and GPs can have vast effects in both reducing unnecessary attendances to A&E and improving help-seeking behaviours.

• Understand the wider context: We have demonstrated that GP shortages in deprived areas may be a barrier for patient access. Other services such as pharmacies and other community services (e.g. health visitors, community and school nursing) may be equally affected and should be investigated. In Greater London, Child Health GP Hubs have been set up to address a shortage of GPs. These hubs are specific to paediatric children and provide more streamlined care by following a joined up care model (Imperial).

• Evaluate the benefit of novel initiatives: Could there be a benefit to integrating on-site primary care in children’s A&E? Combining community and trust data may reveal benefits in developing paediatric-specific ambulatory care centres sited in A&E. These could see children with less-acute conditions that present to A&E and provide more holistic care. Similarly, more specific paediatric services in primary care, and targeted education for primary GPs and nurses could support parents’ confidence in primary care providers, and combat the perception that paediatric A&E may provide a better service.

[1] The findings are also limited to latest available data resulting in temporal inconsistencies across analyses.

After what seemed to be a potential recovery of ambulance response times in November 2022, the latest data release from NHS Digital shows that response times have taken a significant downturn, hitting the highest on record.

Since 2020 there has been a large increase in mean response times across all incident categories. C2 incidents (serious conditions that are not immediately life threatening, such a strokes and chest pain) have suffered the most with mean response times reaching 93 minutes in December 2022, 5 times higher than 18 minutes pledge time. Waiting this long for transport to care will have drastic impacts on a patient’s outcome, not just for life threatening issues, but also for urgent conditions needing acute care, such as C3 incidents that have seen their pledge time exceeded by 455%.

When breaking down these figures into regions in England, significant differences emerge, with C1 calls in the South West waiting an average of 13.2 minutes, 32% higher than North West and the Midlands’ average response time of 10 minutes. Despite diverting 68% of calls out of 999 (compared to 19% in November 2022) and allocating an extra 9,000 ambulances (a 45% increase) to attend C1 calls, only 2/3 of them arrived on site, meaning the other third was stuck elsewhere.

Ambulances mean response times by region, December 2022

What is behind this huge disparity across regions?

Calls to ambulances have seen a significant increase since the start of 2021, growing by 20% nationally. And although the total number of calls resulting in an ambulance being dispatched (an “incident”) have decreased, the proportion of incidents attributed to C1 calls has grown to 18% (from 9% in 2019) the total number of C1 incidents has increased significantly, up by 23% compared to last year (Dec 2021), and the number of C1 incidents has seen a sharp rise. This is particularly significant in the South West, where C1 incidents have nearly doubled since 2021, suggesting that patients are becoming sicker, not just more willing to pick up the phone.

The situation in the South West should not be seen in isolation, but rather as a premonition of what might be coming for other regions if resources are not planned adequately. The remoteness of locations in the South West should not be the main culprit in the rising ambulance times – pre-pandemic they were performing in line with other regions -, but rather evidence of the strain that population health factors place on acute and community services, and the need to plan accordingly.

On the one hand, the population of the South West is amongst the oldest in England, which naturally leads to higher levels of demand across the entire health as well as the social care spectrum. Our recent work with NHSE/I on demand for secondary care shows that significant planning is required to deal with the demand associated with ageing.

Issues with capacity and bed utilisation are on the other side: 6 out of 8 ICBs in the South West have average G&A bed occupancy of above 92% (the recommended maximum), well above the national average of 88%. Last week, 20% of their entire G&A bed capacity was taken up by patients who are medically fit for discharge.

Being unable to shift patients out of hospital results in A&E departments too busy to take handovers from ambulances. In December, the average time lost to ambulances due to delays in handover more than doubled – in fact, time lost due to delays in handover was the equivalent of 40% of the total time spent dealing with incidents.

C2 ambulance response times and G&A bed occupancy, England

A new NHSE delivery plan for recovering urgent and emergency care sets specific funding to both increasing capacity of beds, ambulances and same-day emergency care services (£1bn), speed up discharge (£1.6bn), with further mentions for growing workforce, expanding community services and tackling unwarranted variation that did not receive specific funding mentions.

This is a step in the right direction, though it will now be up to ICBs, once the funding has been streamlined, to figure out how this can be used most effectively. Too narrow a focus risks creating bottlenecks downstream, rather than solving the issue, and solutions will need to both address patient flow while targeting the whole pathway, spanning from community care to addressing workforce.

At Edge Health we are experts in using forensic data analysis to target new capacity to solve the flow problem, not just move it. In our experience, full-spectrum capacity planning is what enables effective use and distribution of resources, and we have supported trust-wide planning and reconfigurations that have enabled trusts to recover the 4-hour A&E target. To find out how we can help you, get in touch.

Standards introduced in October 2021 mandate that at least 75% of patients urgently referred by a GP[1] for suspected cancer should receive a diagnosis (or be cleared) within 28 days. In September 2022, however, 33% of them – 78,000 a month – did not receive a timely diagnosis. This reduces their chances of survival considerably by preventing prompt treatment.

As shown in Figure 1, from April 2021 to September 2022, the number of patients meeting the 28-day target has dropped from 73% to 67%, and the 75% target has yet to be met. This is likely due to the large increase in the number of referrals registered during the past months which is constraining diagnostics capacity (as discussed in our previous post).

This varies greatly by suspected cancer. The 28-day rate for breast and children’s cancer in September 2022 was in fact close to 90%, while for tumour sites such as gynaecology, skin, lower gastrointestinal and urology (including prostate) it was as low as 50%.

For urological malignancies, the failure to meet the 28-day diagnostic target has a significant impact on the 62-day treatment target. As shown in Figure 2, the longer the delay in obtaining a diagnosis, the more likely patients are to miss the 62-day treatment benchmark. This highlights the crucial importance of timely diagnosis in ensuring prompt and effective cancer treatment.
Furthermore, as the number of diagnostic tests administered increases, so does the proportion of patients who meet the 28-day diagnostic target. This correlation suggests that delays and capacity limitations in diagnostic testing are playing a key role in the decline of cancer care outcomes nationwide[2].

This is not going unnoticed, with trusts racing to implement a number of solutions to reduce pressure on hospitals and provide quicker access for patients. There are three broad groups of approaches to streamlining cancer diagnostics:

• Creating additional diagnostic capacity using weekends, new diagnostics centres and independent sector diagnostics – For instance, the opening of new community diagnostic centres across England will provide elective diagnostics such as checks, scans and tests away from acute facilities and free up hospital capacity.
• New population screening programmes – An example of this is the new lung cancer screening programme which aims at improving early diagnosis by running a low-dose CT scan of the lungs on high-risk people and inviting them for further tests if abnormalities are shown.
• Implementing additional rapid diagnostic services for urgent patients – Like the national roll-out of fast-track testing, which from November 2022 allows every GP team to directly order CT scans, ultrasounds or brain MRIs for patients with concerning symptoms.

The examples provided above are just some of the latest initiatives aimed at cutting down waiting times for cancer patients across the NHS. In our experience, however, one approach has already shown huge potential for effectiveness: mutual aid between Trusts. In the next blog post, we’ll dive into the details of data-driven mutual aid and demonstrate the vast impact it can have on diagnostic and treatment efficiency and speeding up recovery.

[1] Along with patients referred by their GP with breast symptoms where cancer is not initially suspected; or referred by the National Screening Service with an abnormal screening result.
[2] The data used include both diagnosed and non-diagnosed patients. With more specific data this correlation is believed to be even stronger.

The pandemic has powered new ways of working across the NHS, and remote care has seen an unprecedented rise in both primary and secondary care. It allowed care to continue in the midst of a pandemic, and now, in the aftermath that has left the system heavily stretched, it provides a way to reach more patients despite a depleted workforce.

The data shows that the ratio of face-to-face appointments per GP across England has not changed. GPs are providing just as much face-to-face activity per person as they were pre-pandemic, though at first glance this may be easily missed.

Compared to the beginning of 2018, there are now 2,500 fewer fully qualified GP FTEs. With the current yearly ratio of appointments:GP, we would have expected them to carry out approximately 15m appointments. Despite this significant workforce loss, primary care is delivering more appointments than ever, and the number of monthly appointments per GP has increased by 30% since 2018 (first year for publicly available data on appointments in primary care).

GPs have undoubtedly been working harder, as they tell us, and one of the benefits of remote care shows in the chart below. GPs are now performing roughly the same number of face-to-face appointments per GP that they were at the end of 2018 (this is strictly GP-led appointments, rather than across primary care). Additional capacity is coming from the rise in telephone and video appointments and triage.

Moreover, despite the negative press that often surrounds primary care, the average wait per appointment to see a GP has decreased by 25% compared to 2018. Again, given the fall in the GP numbers, a more positive narrative is that this is a tremendous achievement despite having lost 10% of their workforce in four years. The majority of patients are also satisfied with the service provided, given that only 10% of patients would rather see a GP face-to-face.

How long primary care will be able to plug the holes and work harder is a key question. We have applied public health drivers to forecast demand in primary care, and it paints a dire picture: by 2032, we’d need 39,000 GPs to match demand. There is no quick fix, and a strategy is needed on several fronts, including addressing public health drivers, training and retention. In this scenario, we should expect and welcome the emergence of new ways of working, as innovation may help GPs find time for more patients when there are only 24 hours in a day.

Closing the GP Workforce gap – part 6

There has been an unprecedented focus in Primary Care to make better use of digital resources, providing virtual and remote consultations. Changing consultation modes will not provide the answers we need when looking at the growing GP workforce gap, which requires focussed planning on training, retention and addressing demand. We estimate that the benefit of innovation on workforce numbers will be relatively low compared to other strategies and that is by design – innovation should support, rather than replace, the workforce, though there is scope for potential productivity gains.

Closing the 2032 GP workforce gap, aiming for 8.5 % vacancy
Number of GPs expressed as Full Time Equivalent (FTE)

With that in mind, we should not put innovation on the back burner, for at least two considerations. One, that discovering new ways of working relies on innovation, which may yet provide us with strategies to tackle need; secondly, that innovative ways of work can have ripple benefits on other areas, such as workforce and patient satisfaction and increase the potential to attract and retain doctors in primary care.

What’s out there?

Innovative solutions are widespread in Primary Care, and some have demonstrated promising changes the way healthcare is provided, particularly through bridging primary and secondary care, and making care more accessible. Consultant Connect, a service developed for North Central London CCG that allowed GPs to directly message specialists for advice and guidance, reduced secondary care referrals in 64% of contacts as specialists could provide advice to be implemented in primary care rather than requiring a referral.

The virtual boost promoted by compelling need during the pandemic has allowed e-consultations to flourish and transform the way time is spent in primary care: e-consultations in North London have allowed non-clinical queries to be directed to staff other than GPs, and GPs were able to deal with e-consultations much faster than in person, reducing face to face consultations by 25%.

Our previous work has highlighted the vast impact of video consultations beyond healthcare savings – such as savings on inpatient travel resulting in lower greenhouse gas emissions as well as £108m savings in lost productivity through avoiding 3m lost work hours.

Productivity gains through harnessing innovation can be very material – virtual consultations have saved PCNs approximately 17,200 hours – scaled nationally, this is equivalent to freeing enough capital to employ 900 GPs –, and eConsult has shown to reduce missed GP appointments by 60%, with large time and cost savings. These strategies could give a substantial boost to our 2032 workforce concerns, though it is important to recognise the technology supports and enables the workforce, it does not (yet) replace it.

How can innovation become a reality?

Innovating does not always equate to revolutionising systems – sharing learning and making best use of resources in effective and innovative ways can be the best route to having a large impact. Working with Kaleidoscope and Cambridge and Peterborough STP, Edge Health supported the implementation of a digital first pathway for diabetic patients that enabled, through shared priorities and consistent approach to care, to improve care for diabetic patients and increase the uptake of the National Diabetes Prevention Programme without generating more work for GPs. Key to the success of the programme was the sharing of evidence that supported innovation within the STP, as well as the engagement with frontline staff that directly experienced the benefits of the innovative intervention.

We could draw four take home points from this successful implementation of innovation at scale:

• Ideas should be generated locally, where need is clearly understood
• Successful innovation relies on collaboration across practices – this promotes not only sharing of best practices, but also agreement on care delivery strategies that have the potential to level up working habits
• Taking on from the example of practices, collaboration across ICBs will also enable effective innovation to reach other areas through cross-communication; central channels need to be available for this to happen
• Innovation must involve frontline staff rather than be exclusively top-down, for effective implementation and longevity

The last consideration is particularly important in the context of workforce – attempting to implement innovative strategies will undoubtedly be a challenge amidst an overstretched, disenchanted workforce. Taking decisive steps towards tackling the workforce gap will be a necessary step to promote further innovation, and in return, innovation could hold the keys to better healthcare that makes best use of resources.

Closing the GP Workforce gap – part 5

30% of GP appointments are due to musculoskeletal (MSK) conditions – nearly 49million in 2021. The solutions that GPs can provide is often limited – self-management and analgesia, referrals to secondary care or physiotherapy.

By contrast, physiotherapists working in primary care can avoid unnecessary referrals, reducing admin and bureaucracy, providing cost-effective services that can reduce both primary and secondary care workloads, while empowering patients to take ownership of their health.

The role of direct patient care practitioners (DPCs, which includes physiotherapists) and nurses in primary care is expanding – and rightly so. 55.6% of primary care patients have more than one chronic condition; they need holistic, person-centred care. This can hardly be provided in 10-minutes GP appointments, where patients would like to have 2.5 issues resolved, but are told to only bring one. Multidisciplinary teams have already proven highly successful in secondary care and are key to empower patients to manage their conditions from multiple angles – covering education, nutrition, physical activity – without relying on over-medicalisation. Widening the healthcare team can achieve not only higher patient satisfaction, but also equal if not better care for patients than doctors alone. Expanding their role also offers solutions to addressing GP workforce demand which, as we have seen in our previous posts, poses significant challenges for the future.

Catch up with previous chapters of our blog series here.

Closing the 2032 GP workforce gap, aiming for 8.5 % vacancy
Number of GPs expressed as Full Time Equivalent (FTE)

So how many do we need?

Our projections suggest that by 2032, primary care demand will rise to 382million appointments a year. As outlined in our first post, providing today’s level of care requires closing a gap of nearly 30,000 GPs as well as employing 11,000 extra DPCs and nurses, who see just under half of all primary care appointments.

The valuable role played by DPCs is reflected in their growth of 8% a year since 2015 – though nurses numbers have remained largely stationary. Continuing the current trend will fulfil the demand requirement, though it might rely on continued funding through the Additional Roles Reimbursement Scheme (ARRS) which currently subsidises practices employing a list of professionals (including physiotherapists, dietitians, social prescribers and pharmacists). We are also faced with a choice: to continue “as is”, on a healthcare provision path that is facing significant challenges, or aim for a change. Could the role of DPCs be expanded to more consistently take over some of the functions that are traditionally the remit of GPs, mitigating to some extent the challenges in GP supply?

Collaborating with community pharmacists can save each GP practice £7,000, and spare patients from taking unnecessary, costly medications that may worsen rather than ameliorate their health. If physiotherapists could lead more MSK consultations, for instance taking over half of all MSK consultations carried out by GPs (24m), the saving in direct cost per appointment would be £375m a year, not accounting for potential costs saved from unnecessary tests, secondary care referrals and more. That would be enough to employ 9,000 full time equivalent physiotherapists in primary care.

Given the difficulties in matching GP supply to demand, DPCs could help reduce the overall requirement for GP time by 2032, while providing even greater benefits on population health through models of care more focused on prevention than our current.

There are two ways to achieve this:

• Increase the number of DPCs to slightly more than the required 11,000, to provide some buffer to the system in case GP training and recruitment efforts don’t bear fruit. Keeping the current ratio of professional to appointments, an extra 2,000 DPCs could free 1,800 GPs’ time by taking over the appointments that would be otherwise carried out by them (where appropriate). This time gain could be re-invested in providing training or clinical support to DPCs, keeping the system sustainable. The saving in direct employment costs would be £43m.
• Identify more appointments that could be appropriately carried out by DPCs and do not require a GP in the first instance, increasing the average appointments seen by DPCs, and slightly reducing GP consultations. Adding 2 extra appointments to each DPC’s working day could free 18million GP-led appointments. This requires appropriate triage, DPC training and collaboration amongst the practice team, so that patients are being seen by the most appropriate professional, with support from the wider team. For this option, the saving in direct employment costs would be £130m.

Primary care networks would be the optimum environment to lead on DPC training and expertise, as resources could be pooled amongst practices, allowing even the smaller ones to benefit from multidisciplinary teams.

Are there any drawbacks?

The body of evidence highlighting the clinical benefits – and to some extent, the economic benefit – of DPC staff and primary care nurses is growing. However, some sources have pointed out that roles may not be readily replaceable, and that patient education is needed to ensure satisfaction is maintained.

Innovating our models of care require two important considerations. First, that the increase in non-medical staff needs to be matched by adequate patient education on the diverse roles in primary care and that many concerns can be addressed by professionals other than a GP – for instance, that practice nurses are excellent educators for chronic conditions where many GPs may fall short and that pharmacists are experts in medication review and management. And secondly, that there will never be a 1:1 replacement ratio amongst any healthcare practitioner – what would be the point in a different job title and education, otherwise? It is the contribution provided by team members in a varied workforce that will reduce the burden placed on GPs and help with retention, while providing better, more holistic care for patients. However, for as long as primary care is seen as an exclusively doctor-led world, innovation and change in its structure will inevitably lead to unmatched public expectations, and discontent. Some practices have achieved this with success – much can be learned if these experiences are widely shared, and effective initiatives should be trialled across more centres under the direction of PCNs.

To support primary care through the future, one must think of the future. Outdated models of care are uncovering deep cracks in our system. It’s time to bring innovation into primary care, both to benefit patients and to improve primary care working conditions. More on this next week, for our last post of the series.