Undergraduate medical education: recommendations

Chapter Three: Funding and Infrastructure

Medical ACT

Scottish Government financially supports the delivery of undergraduate medical education in three ways: by support to universities via the Scottish Funding Council, to the NHS Boards via ACT (additional cost of teaching) funding which is currently administered and allocated by NHS Education for Scotland ("NES"), and via the Students Awards Agency Scotland (SAAS).

ACT is provided by the Scottish Government to cover the additional costs of teaching medical undergraduate students within the NHS in both primary and secondary care. ACT funds are distributed using an allocation model which is based on student numbers (stage one) and on the amount of teaching activity (stage two). The allocation of funds to primary care is based on a cost reimbursement model whereby GPs are paid for what they specifically do, which is ring-fenced within the ACT budget. For secondary care teaching, Boards receive a 'share' of the available funding relative to amount of teaching they deliver, the latter being calculated using a Measurement of teaching (MoT) tool. ACT funds are also used to support student travel, accommodation and subsistence costs in relation to undergraduate placements and this amount is also ring-fenced within the annual budget.

The ACT budget set by NES for 2018-19 was £77.2m and within this the contribution towards GP ACT was £7.1m (around 9% of total).

Brief history of ACT - "direct" and "indirect" ACT

ACT was originally introduced in 1977 and was based on an English formula (which was called the Service Increment for Teaching ‐ SIFT). ACT was based on the difference in costs between teaching hospitals and a sample of 45 District General Hospitals (in England and Wales). The total level of ACT funding in Scotland at that time was £86.6m pa.

Prior to the transfer of responsibility for ACT to NES in 2005, a review undertaken under the auspices of the Standing Committee on Resource Allocation (SCRA) identified 2 funding streams: (i) Direct Costs - equivalent to £57m (66%) and (ii) Indirect Costs - equivalent to £29m (34%).The direct cost element subsequently came to NES for management, and has risen over time (mainly as a consequence of passing on annual uplift) to the current level of circa £77m.

The indirect cost element was baselined to Health Boards around the same time. This section will focus only on direct cost element of ACT administered by NES. Similar support for the clinical placement costs of undergraduate medical students exists in other parts of the UK, where the direct and indirect cost elements have been kept in one 'pot' and managed by the relevant statutory body. As a consequence, identifiable 'SIFT' funding per student is significantly higher than direct ACT funding in Scotland.

Current GP ACT

Historically, primary care services were initially excluded from ACT funding until the early 1990s. The first allocation of ACT monies was made to primary care in 1992. GP ACT, like the hospital counterpart, is split into Category A and Category B costs. In general, Category A weeks are spent in placement in practice with a GP, usually for 1:1 teaching which currently attracts a sessional rate of £40 per student across Scotland. This payment amount was agreed by stakeholders, although the rate has not been subject to revision since 2010 and as such is due for review.

Category B costs relate to formal teaching or other teaching-related activities which take place outwith clinical placements, often on campus but also within a General Practice setting. There is variation between the type of activity and rates of remuneration provided by medical programmes.

NES Primary Care ACT Review and Work of Group

The key aims underpinning the management of Medical ACT are to deliver the highest possible quality of undergraduate medical education within the NHS to support the teaching of medical students in Scotland; to ensure a transparent and equitable approach to the distribution of ACT funding; to achieve best value through robust performance management of the use of ACT funding. NES commissioned a review of the costs of primary care ACT, which began in April 2018. The NES Review was necessary for a number of reasons namely:

• the increasing difficulties in providing teaching by GPs in primary care
• the emergence of new modes of delivery of undergraduate medical training in primary care, including the launch of ScotGEM in 2018 and new models with NHS Boards for which costs were not fully quantified
• the aspiration to increase the proportion of undergraduate medical teaching that is delivered in primary care.

The intention of the review was to quantify the current actual contribution and costs of Primary Care teaching and to describe the total contribution GPs make to the curriculum for each programme. A detailed exercise was conducted by the Medical ACT team to collate data from each medical programme relating to GP teaching, placement and support costs, along with the costs associated with travel and subsistence. In addition data was sought on activities delivered by GPs outwith these areas including those that were unfunded. The results confirmed that GPs contribute to medical undergraduate teaching in a variety of different ways and an important goal of this review was to reflect that diversity and to devise a means of capturing the actual impact on GPs' time.

There were close working relationships between the NES Primary Care ACT Review team and the undergraduate group (including joint group membership). The ACT Review is due to be published imminently and key findings from the review have helped shape the recommendations made in this report. The findings from the NES ACT review are attached in Appendix C.

Implications of current levels of ACT funding in general practice

Chapter 1, current context, sets out some of the difficulties experienced by Scottish medical schools in recruiting practices and delivering placements, resulting in declining numbers of practices taking part in teaching. We have detailed large falls in the number of practices teaching in Glasgow and St Andrews over the past three years, and a failure to recruit for expansion of GP teaching in Dundee in 2018. A specific example for a practice in Lanarkshire which gave up GP teaching due to multiple factors is given on page 28. We are aware that these difficulties in providing placements has triggered at least one medical school to ask the relevant Health Board to offer higher rates of remuneration than the historical rate of £40 per session.

Chapter 2 sets out in detail the evidence from the capacity survey. When asked what would make the most difference to capacity, 82 per cent of respondents said the backfill of GP time with locum provision, closely followed by increased financial compensation of time lost to teaching (with 79 per cent of respondents agreeing).There is no doubt that issues with recruitment and space within premises are all limiting factors. However, it seems clear from the context and the evidence of the capacity survey that the current level of reimbursement of £40 per session is inadequate to maintain current levels, and that undergraduate education in primary care will not increase at this level, and is likely to decline. Lack of resourcing for undergraduate education is also specifically mentioned as a barrier by Barber et al (2019).

What does it actually cost a GP practice to teach an undergraduate medical student?

A review of the costs of teaching in primary care was undertaken in 2017 by Professors Rosenthal, McKinley and Campbell for the Department of Health and Social Care (England), in association with Health Education England (HEE). The study methodology considered the various direct and indirect costs associated with teaching undergraduates in primary care and examined data collated from a sample of 50 practices, approximately 2 per medical programme. The study is currently in press but the authors have shared in personal communication their finding of an average cost per half-day student placement in general practice of £111, which is in stark contrast to the current payment rate in Scotland of £40. Publication of this study is due in 2019.

In late 2018, a similar exercise was undertaken in Scotland, using the algorithm employed by Rosenthal, McKinley and Campbell, which was shared by kind permission. The Scottish study, led by Dr Amjad Khan of NES, examined a sample of 8 GP Practices in association with 5 Medical Programmes (excluding ScotGEM) and was conducted using a combination of face to face visits to practices and by telephone conversation with the GP and/or Practice Manager. Given the small sample size, variations in the way that teaching was delivered and the size of the groups attending for education sessions, it was not possible to subject the data to the same detailed statistical analysis employed by the English study. However, a median cost of £85/student per 4 hour session, was calculated, which is broadly comparable to the English data.

Of note, the primary unit of analysis recognises the difference between single student placements and those placements where a group of students attends at the same time. This is important as the average number of students per placement varies between practices with values ranging from 1 student per placement to almost 10 (median = 2 students per placement), depending on the year of study and the type of activity being delivered. It is not assumed that groups with multiple students would attract a sum of £85 for an infinite number of students, but rather a ceiling payment per session would be applied, based on a value agreed by stakeholders.

Raising the ACT tariff for GP education

The strategic direction of travel as set out in the 2020 Vision and Health and Social Care Delivery Plan is to shift the balance of care from secondary to primary care. It is important to remember that 90% of healthcare contact happens in community settings and not hospital and it is important that we educate the doctors of the future accordingly. Currently 27% of doctors in Scotland are GPs. In addition, experience in general practice and primary care is important for all doctors, whether or not they choose to become GPs, as an understanding of the social determinants and cultural context of how people become ill is largely gained outside hospitals.

This substantiates the need to increase the amount of education that takes place in primary care and is taught by GPs. While it has been a longstanding principle that GP teaching should be treated on an equitable basis with hospital teaching, there is good evidence from the current decline in UG teaching practice numbers and the capacity survey that this principle will not support an increase undergraduate education in primary care. To increase from a baseline where 9% of teaching currently takes place in primary care, it will be necessary, as part of a package of measures, to urgently raise the ACT tariff for GP education.

It was agreed by the Group that there are three important principles that require to be considered in the context of potentially raising the tariff for GP ACT. These are that GP ACT should:

(i) adequately compensate those educating and training our future doctors and be attractive enough to allow new practices to come on board as demand for training in primary care settings increases;

(ii) be affordable to the public purse;

(iii) not destabilise secondary care ACT.

Modelling a tariff of £85 for Category A costs

• Based on the Khan data, NES have, at the request of the Group, modelled the financial implications of uplifting the £40 tariff for Category A costs to £85 on the basis of the status quo level of GP teaching for each of the programmes.
• Appendix D provides the financial implications of uplifting the rate to £85 for Category A costs based on current levels of activity. The calculations are based on curriculum data for Category A costs obtained from 5 Medical Programmes (Aberdeen, Dundee, Edinburgh, Glasgow and St Andrews).
• The financial impact of increasing the tariff from £40 to £85 for Category A teaching (no change to amount of GP teaching delivered) would be to increase costs for GP ACT by just over £2.5m.

What about rates for multiple students attending for group teaching?

The primary unit of analysis used in the survey by Khan was the cost per student session. This recognises the difference between single student placements and those placements where there is a group of multiple students attending at the same time. This is important as the average number of students per placement varied between practices in the sample with values ranging from 1 student per placement to almost 10 (median = 2 students per placement). In considering the scenario of multiple students attending for teaching, the cost of backfilling a GP's time may be the more appropriate figure to adopt where more than two students are attending, as the GP is unlikely to be able to run a surgery in that scenario. For example, the proposed payment for teaching 3 students would attract a rate of £255 per session (3x£85). This amount is comparable to a current reasonable market rate for a GP to attend an external meeting or to obtain a locum at standard costs. The group therefore agreed that a ceiling tariff value should be adopted at the level of 3 or more students.

Category B costs

The question of altering the rates for Category B teaching, as defined above, is considerably more complex. In contrast to the common rate for Category A teaching adopted by all medical programmes, there is wide variation between the type of activity and rates of remuneration provided by each of the medical programmes for Category B activities. Some examples for this (academic year 2017/18) are listed below and illustrate the heterogeneity that exists currently between programmes:

• Aberdeen - 15 GP practices deliver the Years 1 - 3 Foundation of Primary Care sessions. These are each remunerated via Service Level Agreement worth £13,149, which is the equivalent of 10% FTE Senior Clinical Lecturer contract for 1 session along with a £2016 practice fee. All practices deliver 2 out of 3 FPC years in any one academic year. In practice due to the remuneration method and timetable, the SLA cost for teaching is difficult to break down into a cost per teaching session.
• Dundee - A range of payments are in operation with rates of £40 per 2.5 hour session for a compulsory "experience" session in primary care to £40.19 per tutor per hour for small group sessions that generally last 2.5 hours. Each group has 40 students and involves two tutors, giving an approximate cost of £200 per session.
• Edinburgh - GP practices are paid a fixed rate per semester in year 1 (£1,257.60) or by quarter in year 2 (£3,795). The year 2 payment includes some funding for teaching facilities costs.
• Glasgow - Rates vary depending on type of teaching activity of which there are 4 different categories: Communication skills, Vocational studies, Clinical visit and GP visit. The sessional rates for these activities vary from £140 upwards. For some activities a fixed amount is paid by annual contract to a salaried tutor. Teaching facilities costs are also made if activities are based in a Practice.
• St Andrews (BSc Hons) - has the most straightforward rate of £161.81 which was paid across all sessions until late 2018 when it was increased to £180 per session

As each medical school uses a different method or methods of reimbursing practices or individual GPs for Category B activities, the group agreed that harmonisation of costing and reimbursement would be desirable to achieve a common rate or range of rates for category B teaching across all medical schools. This would also simplify the administrative procedures for Medical ACT within Universities and NES. However further work is required with input from relevant stakeholders in collaboration with NES to take this aspect further. Beyond that, it is likely that as the amount of teaching in primary care is increased towards 25%, according to aspirations of Scottish Government and Scottish Funding Council, ACT support costs could also be expected to increase further. Similarly, increasing teaching capacity to include more remote placements could also impact on travel and support costs for students who are placed outwith central areas. Currently it is difficult to predict the exact quantum of increase that would be required for the ACT budget to support both of these aspects. It is therefore essential that each of the medical programmes has a robust understanding of their current Primary Care support costs so that this amount may be captured accurately following any proposed change in policy.

Looking to the future

One criticism that may be made of the approach outlined in this report is that is looks at Primary Care ACT costs in isolation. Currently primary and secondary care work on the basis of different models - ACT is distributed through measurement of teaching in secondary care as opposed to cost reimbursement which applies in primary care. These differences reflect that the majority of GPs remain independent contractors, as opposed to consultants who are employees. GPs therefore do not have teaching in their job plans as can exist in secondary care.

While it is recommended that in order to encourage the changes we want to see that the category A tariff is raised to £85, it is arguable that more fundamental changes are required. While it is beyond the scope of this report, it is considered that, in the medium term, there would be merit in considering whether the current ACT model across both primary care and secondary care requires to be reviewed in a more fundamental manner. It could be further argued that such a review should not be limited to ACT but rather consider the overall approach to the funding of undergraduate medical education. Either way consideration should be given to alternative models of distributing monies to support undergraduate teaching, particularly against a background of changing models of delivering undergraduate education, different career pathways in general practice and constrained public finances.

Physical Infrastructure

Discussions with key stakeholders and analysis of the capacity survey show clearly that lack of physical space is seen as a major factor in limiting expansion of undergraduate education in primary care. This is particularly significant given not only the commitment to increase GP numbers but the growing need for training and clinical supervision of the wider MDT, including ANPs, pharmacists, paramedics and MSK physiotherapy practitioners.

Currently approximately 40% of GP premises are privately owned, 25% leased and the rest (35%) in health centres. It should be noted that it is difficult to provide completely accurate figures due to the number of branch sites and overlapping forms of ownership (a GP may own a building which the health board is operating another practice in without a lease for example).

As far as privately owned premises are concerned, the Scottish Government has already announced £50 million of GP Premises Sustainability Loans. These loans are available to an amount of up to 20% of the premises value as a GP surgery, except in exceptional circumstances where more may be provided. The Scheme aims to ease the financial risk associated with GPs owning their practice in turn helping to improve GP recruitment and retention.

As far as those premises owned by Health Boards are concerned, the Scottish Capital Investment Manual (SCIM) provides guidance to be applied in the development of all infrastructure and investment programmes and projects within NHS Scotland. Depending on the level of investment, investment decisions are either made at Health Board level or nationally by the Capital Investment Group (CIG). The CIG reviews all business cases for capital investment projects which are above a Board's delegated limit. In practice the vast majority of infrastructure investment comes through the CIG. All capital investment cases must include a strategic assessment of the need for service change. It is recommended that in future there should be a specific requirement for future training and clinical supervision requirements to be considered in all business cases that come before the CIG.

It should also be noted that the Scottish Government has asked NHS National Services Scotland to prepare a business case for digitising GP patient records. This will focus on the benefits of releasing space for clinical and teaching use. It will use information provided by a survey of GP premises funded by Scottish Government. The survey, to be completed by the end of May 2019, will provide information on the condition of properties as well as the use and size of rooms. Digitising GP patient records would incur significant costs but have the potential to increase physical capacity. It seems likely that the effectiveness, and cost-effectiveness of undertaking digitisation will often depend on the physical layout of each practice.

The biggest strategic issue however is that further capital investment in primary and community services is required. As care is increasingly shifted towards primary and community care settings, it is inevitable that space to educate and accommodate the workforce of the future, particularly the expanding multi-disciplinary team, will be required. A new NHS Scotland Capital Investment Strategy is due to be published shortly. It is recommended that this should make the case for investment in in primary and community care facilities recognising specifically the need for that to include purpose specific facilities in which to train the workforce of the future.

Biggar Medical Practice and medical students

Biggar Medical Practice has enjoyed a successful partnership with University of Glasgow Medical School for many years. Reports from students are always suggestive of an enjoyable rural placement offering a wide variety of experience. Sadly in the past year we have encountered numerous difficulties which has forced us to remove our support for students. We face challenges in areas such as IT, accommodation, physical working space and GP cover for student clinic.

Having recently lost a GP to early retirement we cannot justify reducing appointment numbers for student led clinics. We have also seen an increase in the MDT use of our practice which means we can no longer offer our students regular working space within our practice building. We have no library space or meeting rooms. Difficulties regarding accommodation are ongoing and we have offered subsidised accommodation to our students. Often the grant from the university is not enough to cover the cost of local accommodation. Being a rural practice students often note the broadband connectivity in poor and this in turn forces them to travel home at weekends to do course work. The sad impact of this being the students don't stay over the weekends to enjoy the richness of rural living.

We are hopeful of supporting medical students in the future, however there are numerous challenges to manage first, the most challenging being physical space.

Digital infrastructure

The Scottish Government Digital Health & Care Strategy 2018 in combination with ever more digital based education services, and indeed the more general move to digital services in our everyday life, is predicated upon connectivity. This issue is most pressing in remote and rural areas and significant focus and resource needs to be brought to bear. Families who move to remote and rural areas, whether they are health, social care or education professionals, expect high quality broadband connectivity. Absence of this is a barrier to recruitment. Access to university e-portfolios and educational material in general practice is very variable and tends to be worse in remote and rural areas and in independently owned GP premises.

The 'once for Scotland' approach to digital development will increasingly bring this into sharp relief if connectivity in these remote and independent locations is not significantly improved. The roll out of fibre-optic cable to cabinet through the Scotland Wide Area Network (SWAN) initiative is intended to offer high quality broadband to all general practices. Of the nearly 800 premises in the programme 60 remain unconnected, almost all of these in remote and rural areas with complex and challenging topography. This last single digital percentage will take a significant and disproportionate amount of resource to connect and timescales are unclear. On a positive note, the SWAN / Capita team have recently announced that the bandwidth restriction on fibre connections is being lifted and speeds for those practice premises will double over the course of the next few months. In addition, a recent bid to the European Space agency (in collaboration with NHS Cornwall) to develop a proof of concept for health services in remote and rural using satellite connectivity was approved. It is hoped this work will provide connectivity for the hard to reach practices and augmented connectivity (with much improved speeds) for those who have some connection.

This is a Scottish Government issue more generally; a lack of connectivity will increasingly exacerbate potential inequalities, social as well as health specific ones, if not fully addressed. Efforts to coordinate SWAN with the R100 project, which recently received an additional £600m, should be undertaken. This would allow for a better understanding of the timeframes for the remaining very hard to reach locations, and influence decisions on R100 priority areas, such as those places with GP surgeries without connectivity that urgently require access to be accelerated.