33.In this chapter, we examine the readiness of NHS England for its planned Genomic Medicine Service; in particular the infrastructure and training required.
34.NHS England plans to develop a network of seven Genomic Laboratory Hubs to deliver the genomic tests that appear in the Genomic Medicine Service’s National Test Directory. There were differing views, however, about the optimal number of such sites. Dame Sally Davies, the CMO, told us that consolidating the “cottage industry” of existing genomic laboratories into “factories” will provide “higher quality, faster throughput and turn-round, and cheaper prices”. She thought that the Service would ideally have “two or three” sites. The Association for Clinical Genomic Science supported NHS England’s planned move to seven hubs:
The range and complexity of tests in addition to whole genome sequencing is such that a smaller number of “factories” is unlikely to work […] Although the description of UK genomics services as being like a “cottage industry” has been construed as a negative comment, local services can be flexible and responsive to local need and driven by results not the rigidity of a central automated process.
They supported Professor Sue Hill’s proposed model in which, “these new hubs will be built around the extant [Regional Genetics Laboratory] network, with the aim of increasing quality, improving access and reducing waste through efficiency”. The Royal College of Pathologists suggested that the Genomics Laboratory Hubs should be co-ordinated with NHS Improvement’s plans to establish a consolidated network of 29 pathology hubs, emphasising the expertise required to prepare tumour samples for genome sequencing. The College took issue with the CMO’s vision of “a minimal number of DNA/RNA testing laboratories which are separate from other pathology services”.
35.Professor Sir John Bell told us that the digital, rather than the physical, infrastructure for the 100,000 Genomes Project had been the most difficult element to set up. Professor William Newman, of the British Society for Genetic Medicine, also believed that setting up the informatics infrastructure for the Genomic Medicine Service would be the most challenging aspect. Professor Sian Ellard, of the South West NHS Genomic Medicine Centre, highlighted the “huge amount of work to do” on the digital front before the Genomic Medicine Service is introduced:
Each of the genetics laboratories in England already has an IT system, and what we will need to do is link those systems to central databases and to hospitals’ pathology laboratories, and then have a way of delivering the results back to the clinicians through electronic patient records, while also having systems in place where those are not yet embedded across the country.
One person’s whole genome sequence comprises 3.2 billion ‘letters’ of DNA and takes 200GB of data to store, roughly equivalent to the capacity of an average laptop. Professor Sir Mike Stratton, of the Wellcome Sanger Institute, indicated to our predecessor Committee that analysing existing numbers of genome sequences already requires “the sort of IT that is available for the Large Hadron Collider”, which he said “gives you a sense of the challenges that there will be”. He calculated that “the rate at which we are generating DNA sequence, and could generate it from patients’ tumours, is much in advance of the rate of improvement through Moore’s law”.
36.Genomics England currently stores participants’ sample and clinical data in a dedicated data centre. Professor Sue Hill explained that “because we have learned about storing huge amounts of data”, the NHS Genomic Medicine Service will also store data in “one new data warehouse for the NHS that is not in individual laboratories around the country but in one place”. This data centre will be “built by Genomics England and based upon the existing storage for the 100,000 Genomes Project”.
37.Professor Hill told us that the NHS Genomic Medicine Service would need to be able to link and process more than just genomic data:
Routine genomic medicine requires collection and combination of information from [patient] clinical records about a patient’s medical history, their presenting condition and symptoms as well as a phenotypic characterisation, and for this to be considered and analysed alongside genomic test information.
This will require connectivity and interoperability between computer systems across the NHS, not just genomics departments. NHS England told us that this information transfer and processing “already happens as a routine in NHS clinical genetic services, although this may not always be digitally captured and recorded, with paper records still existing”. Professor Ellard, of the South West NHS Genomic Medicine Centre, described how current systems have genetic laboratories “chasing up small details about how names are spelt, for example, which wastes a huge amount of effort and time”:
When a referral to request a genetic test arrives at a laboratory, currently we have a request form that ideally is completed electronically in advance and emailed to us. I had hoped that our existing system would be set up so that we could automatically bring through those data, but I do not think any of the laboratories in the UK has that today.
38.Dr Hilary Burton, of the PHG Foundation, explained that the interpretation of genomic results can be an iterative process, which the current digital infrastructure does not facilitate:
A variant may be found and a question may go back to the clinician about whether such and such a clinical factor is visible in the patient, which sometimes requires the clinician to go back and look at the records […] To do that, they end up having to go through piles of records, piles of paper, to see whether that other clinical symptom or clinical finding is available. They do not have access to the sort of digital clinical records that would make it comparatively easy to add that extra bit of important information for the interpretation. There is still a lot to be developed in the digital infrastructure to make those things happen.
39.Professor Hill told us that “the plan is to provide all the genomic laboratory hubs, and then the NHS providers with whom they will work, with an informatics solution that enables everything to be ordered and all the data to be captured through the same system and to the same standards”:
The costs of this work fall into two distinct areas; the work needed to build the nationally provided solution and the work needed within the wider NHS to enable all local systems to fully interface with the national solution and to enable capture and uploading of all genomic data in every NHS organisation. The costs for the national solution are in the region of £6–9m in addition to the current run budget of Genomics England […] The costs for the wider NHS development are not yet fully determined as this is linked and integral to the whole of the digital maturity development of the NHS.
The Minister, Lord O’Shaughnessy, told us in November 2017 that the requirement for an additional £6–9m “depends on what is put in place”. Funds were not allocated in the 2017 Budget, he informed us, but:
Clearly, we are all committed to making this happen, so we will fund it from one route or another. Whether it is Paperless 2020 or whatever programme, we will make sure that funding is in place to put it in.
40.With regards to the funding for wider digital infrastructure needed to transfer data between separate parts of the NHS, Professor Hill told us:
There is not a budget for drawing the elements from the other clinical specialties. As to pushing the information from a laboratory hub out, there will be a mechanism to do it, but there still needs to be more budget to support how that would get out to every single provider, because it does not work like that at the moment.
The Minister told us that “plans for growing our genomics capacity” were included in the Government’s Industrial Strategy, and that “there is much more impetus and investment going into this as a whole”. The Life Sciences Sector Deal agreed by the Government does set out a range of measures to support genomics in the UK, including whole genome sequencing of the UK Biobank and an extension of the cancer branch of the 100,000 Genomes Project. The sector deal also includes commitments to support the development of the UK’s health data infrastructure. However, it is not clear whether additional funding has been made available specifically to deliver the infrastructure required for the Genomic Medicine Service.
41.Despite the budgetary uncertainty, Professor Hill told us that she expected the system for sharing genomic data to be in place by this Autumn:
There will be an operational system in place by October 2018. That is what we are planning. As we are going through live procurement processes, the intention is that between April and October those systems would be tested.
She made the point that although, at the moment, “genetic tests get ordered every day”, the genomic infrastructure being developed will be “more extensive and complex” than the system previously developed for the 100,000 Genomes Project. Professor Ellard told us that “things are moving forward”, but that it would be unrealistic to expect all of the planned infrastructure to be in place for the launch of the Genomic Medicine Service.
42.Progress on that front will depend on the broader digital information development of NHS organisations through the ‘Paperless 2020’ programme. The 2016 independent Wachter review of information technology in the NHS determined that “some trusts are currently too financially strapped, and/or lacking the staff, the training, and the culture to digitise effectively” and that “the target of ‘paperless by 2020’ should be discarded as unrealistic”. It recommended pushing back the target for digital maturity across all NHS trusts to 2023—five years later than the planned start of the Genomic Medicine Service. The Minister acknowledged that the infrastructure currently available for genetic testing varies across NHS England. Professor William Newman, of the British Society for Genetic Medicine, explained:
There are some centres and hospitals around the country that have solutions in place, but that is not uniform, and there is still a lot of work that needs to be done to provide that infrastructure.
43.Genomic medicine requires the collection and comparison of a wide range of data, and the digital infrastructure for whole genome sequence data cannot be developed in a silo separate from other data sources. Although whole genome sequencing and other genetic tests are already being delivered by Genomics England and NHS England respectively, it is clear that significant infrastructure remains to be put in place in order to enable an efficient Genomic Medicine Service. Unfortunately, the wider improvement of NHS data systems is running to a later timeframe than that needed for a Genomic Medicine Service that will begin operations later this year. Elements of the required infrastructure do not yet have clearly-allocated budgets.
44.Given the intention to have the Genomic Medicine Service in operation later this year, the budgets for the required digital infrastructure should be agreed and confirmed now. Decisions on when to provide whole genome sequencing in place of conventional alternative diagnostic tests should take into account the digital infrastructure available to support it, to avoid attempting to roll out a Genomic Medicine Service at a speed that cannot be delivered.
45.Genomic medicine will require new skills. The CMO’s report describes how genomic medicine is expanding the clinical team to include “diagnostic staff in laboratories and imaging; computer scientists; statisticians; (bio)informaticians; and data scientists who assemble, process and assess the data to advise on diagnosis and treatment”. The British Society for Genetic Medicine noted that existing genetic medicine specialists, including consultants in genomic medicine, diagnostic scientists, and genetic counsellors (experts in managing appropriate communication to patients and relatives of the long-term predictions inherent in genomic medicine), will be required. As Alison Hall, of the PHG Foundation, told our predecessor Committee, the wider workforce will also have to know enough about genomics “to identify the patients who will benefit most from having testing and to understand the results that are returned to them”. Dr Helen Firth, of the Wellcome Sanger Institute, explained how difficult the interpretation of genome sequences can be, and how the “ability to sequence genomes is running very fast ahead of our ability to interpret the data”. Professor William Newman emphasised the importance, and challenge, of getting the interpretation of genome data right, to avoid, for example, unnecessary preventative surgery.
46.Health Education England established the Genomics Education Programme in March 2014, with £20m of funding from the Department of Health for four years. The programme provides additional training and courses, and was described by Health Education England as the “NHS’s method of ensuring its staff have the knowledge, skills and experience to ensure that the health service remains a world leader in genomic and precision medicine”.
47.There remain, nevertheless, concerns about a lack of genomic expertise across several medical specialties. Dr Hilary Burton told us of concerns expressed to the Joint Committee of the Royal Colleges on Genomic Medicine:
We have 17 clinical champions. These are people in various specialties who are expert in genetics. All of them are concerned that there is not enough expertise within their specialty, at a general level.
The British Society for Genetic Medicine noted that there are unfilled posts in genomic diagnostic laboratories and across genomic services due to a lack of individuals with the necessary training and skills. The Royal College of Pathologists highlighted that not all NHS pathology laboratories have the trained personnel needed to process tissue for DNA testing, noting that it takes over 10 years to train a pathologist. The Association for Clinical Genomic Science similarly believed that there is “a serious risk of under-capacity in the workforce to deliver the full benefits [of] clinical genomics reorganisation”.
48.The Association of Genetic Nurses and Counsellors were “extremely concerned that there will not be enough genetic counsellors trained in the UK through the current training routes to meet the growing demands and developments in genomic medicine”. The standard training route for genetic counsellors in England had 15 enrolled in 2016–17 and was due to have just 10 for 2017–18. The Association found in 2017 that 16 out of 23 existing UK regional genomics centres would need more genetic counsellors in future, and 17 centres currently had difficulties filling such posts.
49.In response to these concerns, Health Education England told us:
We recognise that there is more work to be done by [the Genomics Education Programme] across the professions […] We recognise that the medical specialities and general practice are areas that required focused work by Health Education England over the next two years.
In the meantime, they highlighted that through the long-established Regional Genetics Centres there is “already a cadre of well-trained scientists, bioinformaticians and technical staff, clinical geneticists and genetic counsellors who are the backbone of the workforce requirement for the new Genomic Medicine Service”. They believed that, in combination with the Genomics Education Programme, these will be sufficient to introduce an operational Genomic Medicine Service as planned:
Through the work of NHS England, Genomics England and Health Education England in the implementation of, and the support given to, the 11 Genomic Medicine centres, this workforce has been responsive to the requirements of whole genome sequencing, its interpretation and clinical utility, and practise has changed accordingly as new knowledge has emerged […] Educationally, we feel the workforce is in a position to be operational from end [of Quarter 2] 2018. The Genomics Education Programme will be working with the NHS England Genomics Unit and others toward mainstreaming of Genomic Medicine by 2019. This will be challenging and the Genomics Education Programme will need to ensure that its work programme is co-ordinated with the work of the NHS England Genomics Implementation Unit and other key stakeholders.
Professor Ellard, of the South West NHS Genomic Medicine Centre, similarly told us that although the NHS is already delivering whole exome sequencing, which makes it “well-placed to manage the transition” to whole genome sequencing, “we need more trainee posts and more training for the existing workforce”.
50.Health Education England assured us that as the new Genomic Medicine Service takes shape, it will “undertake detailed workforce planning and modelling to inform future training numbers across the professions”. Asked what assessment they had made of the roles that would require genomics training, the level of training those roles would require and the proportion of people in those roles who had received such training, Health Education England told us in February 2018 only that a training needs analysis had been undertaken by each of the Genomic Medicine Centres, and that “this is currently undergoing detailed analysis but the common theme emerging is the complexity of determining the training needs when many of the workforce aren’t yet sure what and how much genomics they need to know”. In the meantime, Dr Hilary Burton explained that the Genomics Education Programme was helping to raise awareness, but not yet providing sufficient training for consultant-level clinical decisions. She was concerned that for the majority of specialty curricula, genomics was not being embedded quickly enough:
Outside specialties such as oncology and haematology, which are quite genomically oriented, I do not think there is a single specialty where [genomics] is properly embedded in the curriculum.
51.A 2013 review of NHS training curricula found that genomics had been embedded in some specialty courses, but not all. Health Education England will now undertake a new review “to identify the current omissions”. They told us that genomics had already been integrated into all of the scientific curricula, and that there was ongoing work with the Royal College of Physicians and the Royal College of General Practitioners to review the current genetic training provision and ensure that it was at the right level.
52.Professor Patrick Chinnery, of the University of Cambridge, believed that the challenge would lay in training the existing workforce, rather than the new staff coming into the NHS. He advocated including competency in genomic medicine in revalidation requirements. Health Education England similarly hoped that:
As genomic medicine practice becomes incorporated into the job plans of greater numbers of medical consultants, competence in this area will become part of annual appraisal, and therefore assessed during revalidation.
53.Lord O’Shaughnessy, Parliamentary Under-Secretary of State for Health, told us that the Genomics Education Programme would receive £1m a year to continue beyond March 2018, compared to the £20m it was awarded for the period 2014–2018. The Association of Genetic Nurses and Counsellors expressed concern at reports of “reducing budgets for continuing professional development for health professionals, including genetic counsellors”. The Minister told us that the decrease in the Programme’s annual funding was due to the ‘up-front’ training required at the beginning of the initiative, and the CMO explained that preparing coursework and online modules also had to be paid for up front. Health Education England told us that “the programme costs for 2018–19 and 2019–20 have yet to be agreed and are under discussion with the Department of Health and Social Care”.
54.There are widespread concerns about insufficient training and a lack of qualified NHS England staff ahead of the introduction of the Genomic Medicine Service later this year. Health Education England has still to undertake detailed workforce planning and modelling, and there is uncertainty within the existing workforce about the extent of genomics knowledge they will need. The Genomics Education Programme is playing an important role in raising awareness and expertise, however this was initially a 4 year programme. It is now to be continued, but with a substantially lower level of funding than previously. Genomics will need to be embedded in training curricula and revalidation criteria to ensure sufficient genomics understanding by all staff involved in clinical decisions.
55.With the Genomic Medicine Service due to be operational later this year, Health Education England should complete detailed workforce planning and modelling as soon as possible. They should also work with the Royal Colleges of Medicine and other stakeholders to embed genomics into relevant curricula and revalidation requirements as a priority. The Government must support them in this work, and ensure the necessary funding is available.
56.Asked about the readiness of the wider NHS workforce for the Genomic Medicine Service, the CMO told us that the entire workforce “will never be trained and ready to do [genomic medicine] themselves, in the sense of understanding all the genes”, but that this was not required if clinicians have sufficient support to be able to request the appropriate test and follow up on their results. She wanted to see a national network of multi-disciplinary teams established “to review and advise on complex and ultra-rare syndromes”. Such teams would include researchers, bioinformaticians and clinicians with expertise in genomics, as well as those treating the patient. The CMO explained that our understanding of the significance of different genomic features will continue to develop over the next five years or longer, and that “patients [will] need researchers to interpret them, because everyday clinicians will not be up to scratch”.
57.The British Society for Genetic Medicine pointed to multi-disciplinary teams of genetic medicine specialists already supporting clinical teams in the interpretation of genomic tests. However, a coalition of Cambridge-based institutions delivering genomic healthcare believed that some multi-disciplinary teams do not have sufficient commissioned support to be sustainable, but had instead been developed “using research resources and goodwill”. Fiona Murphy, of the Scottish Genomes Partnership, told us that multi-disciplinary teams for cancer patients had successfully evolved out of research teams, and that this could work well for genomics:
Sometimes, there can be a bit of a lag moving from a research-oriented team meeting to it being part of clinical practice, but I do not see any block towards that as these things become more organised. Generally, it tends to involve some small seed funding to have a co-ordinator who holds things together and has the infrastructure to bring people together.
Professor Sian Ellard explained that the need for multi-disciplinary teams would have to be “factored into the costing of the whole process”.
58.Multi-disciplinary teams will be critical to providing an effective genomic medicine service to patients and to keep abreast of a fast-evolving science. We agree with the CMO’s recommendation to establish a national network of multi-disciplinary teams. The Government should set out what funding and support it will provide to enable multi-disciplinary teams to develop from being research-oriented to supporting clinical practice, and factor their costs into the commissioning of the Genomic Medicine Service.
97 Association for Clinical Genomic Science ()
98 Association for Clinical Genomic Science ()
99 The Royal College of Pathologists ()
100 The Royal College of Pathologists ()
105 ‘’, Annual Report of the Chief Medical Officer 2016 (2017)
106 taken on 8 March 2017, HC (2016–17) 854, Q183
107 taken on 8 March 2017, HC (2016–17) 854, Q184
109 NHS England ()
110 NHS England ()
111 NHS England ()
116 NHS England ()
122 ‘’, HM Government (2017)
125 NHS England ()
128 NHS England ()
129 ‘’ (2016)
130 ‘’ (2016)
133 ‘’, Annual Report of the Chief Medical Officer 2016 (2017)
134 British Society for Genetic Medicine () and ()
135 taken on 8 March 2017, HC (2016–17) 854, Q133
136 taken on 8 February 2017, HC (2016–17) 854, Q69
138 Health Education England ()
139 Health Education England ()
141 British Society for Genetic Medicine ()
142 The Royal College of Pathologists ()
143 Association for Clinical Genomic Science ()
144 Association of Genetic Nurses and Counsellors ()
145 Association of Genetic Nurses and Counsellors ()
146 Association of Genetic Nurses and Counsellors ()
147 Health Education England ()
148 Health Education England ()
150 Health Education England ()
151 Health Education England ()
154 Health Education England ()
155 Health Education England ()
156 Health Education England ()
158 Health Education England ()
160 Association of Genetic Nurses and Counsellors (AGNC) in the UK and Republic of Ireland ()
162 Health Education England ()
165 ‘’, Annual Report of the Chief Medical Officer 2016 (2017)
168 British Society for Genetic Medicine ()
169 University of Cambridge and Cancer Research UK Cambridge Centre ()
Published: 20 April 2018