Specialised commissioning: lessons from history and looking to the future

Long read: Despite perhaps not receiving much attention, specialised commissioning is one of the fastest growing parts of the NHS budget, and from this month provider collaboratives will be responsible for commissioning specialised services for mental health. Nigel Edwards looks back at the historical problems of specialised commissioning, why finding solutions has often been difficult, and how policy-makers can this time avoid the mistakes of the past.

Blog post

Published: 01/10/2020

What is specialised commissioning?

Specialised services are generally provided in relatively few hospitals and accessed by small numbers of patients. They are usually for patients who have rare conditions or who need a specialised team working together at a centre.

There are currently 146 specialised services, covering a diverse range of disparate and complex services. These range from services for long-term conditions, such as renal (kidney) and specific mental health problems, to services for uncommon conditions such as rare cancers. Some specialised services, such as those for cystic fibrosis, cover the majority of care for patients with these conditions. However, most specialised services only form a part of a patient’s care and treatment pathway.

Some highly specialised services, including those for very rare diseases, are only provided at a very small number of centres across the country. Others, such as chemotherapy services, are provided by most acute hospitals.

The arrangements for planning and paying for specialised services receive comparatively little attention. This is surprising given that the budget hit £20 billion in 2019/20 (17% of the NHS budget), and is projected to grow to £25 billion by 2025. Pre-pandemic, this meant that specialised commissioning was one of the fastest growing elements of the NHS budget – increasing by an average of over 7% per year.

It has also seemed to occupy a permanent place in policy-makers’ ‘too difficult’ drawer, with attempts to improve budgetary control, transparency, data or to reform its scope never quite coming to fruition. There is a consensus there are oddities in what is classified as specialised, and more recently attention has returned to the idea of devolving elements of the specialised commissioning portfolio – and in some cases changing the way that money is allocated. 

Covid-19 meant that these plans were put on hold, but attention is now beginning to return to this area. While there are opportunities for improvements, there are also some challenging issues. Careful preparation is required before a major shift is made in the arrangements for these difficult and complex areas.

Currently NHS England is legally responsible for commissioning these services, and changing this requires legislation. Notwithstanding this constraint, changes in the size and number of clinical commissioning groups, and the development of sustainability and transformation partnerships (STPs) and integrated care systems (ICSs), appear to offer the opportunity to devolve the commissioning of all but the most specialised services.

The 2018/19 planning guidance made proposals for how STPs and ICSs should start to develop a specialised services planning board, while developing an advisory function relating to services that can be planned for populations of one million to 2.5 million, and collaborative arrangements for those with populations between 2.5 million and 10 million.  

In this long read, I look back at the historical problems of specialised commissioning over 35 years, and the issues that make finding solutions difficult. I also offer suggestions for policy-makers on how not to repeat the mistakes of the past.

Key lessons

We have reviewed the rather sparse library of reports in this area and consulted a number of people with a long association with this subject to draw some tentative conclusions. The key lessons seem to be:

  • As with many varieties of the question, “what is the right organisational level to do x?”, there is no one right answer. There are, however, a number of not very good or even wrong answers, and a current risk of rediscovering these again. 
  • This is an area where many of the disciplines of commissioning are still needed and, even where providers manage elements of specialist service, expert oversight is still required.
  • By their nature, specialised services are often at the forefront of innovation. We can expect that recent and imminent innovations in treatment technologies such as advanced imaging, CAR-T and other immunology, as well as new drugs and devices, will significantly increase costs. In common with other services, the increase in patients with co-morbidities is an additional pressure. 
  • National level commissioning provides a strong focus for lobbying by patient and pharmaceutical groups. This makes cost control more difficult, but there is a strong logic for national and regional level commissioning for a number of services.
  • Large populations are required to manage risks and random fluctuations in utilisation. For a number of these services, commissioning at scale is necessary because of the need to develop expertise. The development and use of nationally set standards are likely to be beneficial here.
  • It is extremely difficult to allocate some specialised services funding using a population needs-based formula, and this is not done in other countries. It is not clear how far this would produce beneficial changes for services that are not closely linked to local pathways.
  • Decentralisation of decision-making is generally the right way to go in many aspects of NHS service planning, but it needs to be to the level most able to deal with the challenges. 
  • ICSs are well placed to develop specialised commissioning, and focusing on a collaborative effort to deal with issues of discontinuity between elements of the pathway would be a useful place to start. For some services, given the difficulties NHS England have had in this area, caution might be wise and certainly they will need to pay attention to the skills, systems and processes required before taking over large parts of specialised commissioning.

The history of specialised commissioning

There has been some form of national level planning of specialised services since 1983, when the Supra Regional Services Advisory Group was established. The definition of what constitutes specialised services has changed over the years – becoming broader or narrower depending on the objectives of policy-makers, and the success of different patient or other lobby groups in getting their service included. 

Broadly, the definition covers treatments or specialties where there are small numbers of individuals who require the service, where costs are very high and specialist expertise scarce, and where the variability and high cost of treatments make them difficult for clinical commissioning groups to manage. 

Some specialties such as neuroscience are making good progress in redefining which conditions and treatments really do warrant being categorised as specialist or highly specialist, and this process needs to be taken further.

The Getting It Right First Time (GIRFT) programme and other long-standing analysis suggests there are still small pockets of specialist activity undertaken by some secondary providers that should probably be consolidated into specialised centres, and some specialised work that requires further centralisation. For example, according to an expert group that we worked with, not all 24 neurosurgical units should be doing very low-volume procedures (such as pineal tumour resection, epilepsy, pituitary, intra-ventricular tumour skull base or intramedullary spinal cord tumour surgery). 

In 2012, as part of the Lansley reforms, the definition of specialised services was greatly expanded. This seemed to be an attempt to improve budgetary control and to compensate for the fragmentation of commissioning that resulted from the creation of the large number of CCGs.

Instead, the budget overspent due to a combination of factors. The teams who had been managing specialised commissioning were disrupted, costs were not identified adequately and many new areas of care were added to the portfolio. This meant that a whole new set of skills and relationships had to be developed.

Progress on the aspiration to create more consistency and to develop a more strategic approach with more transparent decision-making was also slow, although better progress has been made more recently. The arrangements were established in a way that left patient groups feeling excluded from the process.

The current scope of specialised commissioning has included over 140 categories of care over the last five years, but 10 groupings account for 63% of the spend. The majority of these services are commissioned regionally, but a small number are identified as highly specialised services. These are commissioned nationally and usually have no more than 500 patients.

Difficulties defining specialised services and resource allocation

Generally, resources for commissioning are allocated to CCGs using a needs-based formula, but the specialist services are still funded on a largely historical basis and relate to providers rather than populations. The indicators of need used in the CCG formula do not capture a number of the predictors of using a specialist service. For example:

  • Patients with long-term specialised conditions may relocate to be closer to specialist centres.  
  • Some ethnic groups that tend to live in concentrated communities have a high prevalence of particular specialist conditions, such as Gaucher’s disease, cystic fibrosis, congenital abnormalities requiring surgery or neonatal intensive care.
  • Some areas have communities with a high prevalence of particular conditions unrelated to standard indicators of need, such as HIV or Hepatitis C.

A further complication is the tendency for some specialised services to be used more by populations in areas that are close to centres providing specialised care.

This may reflect how difficult it can be to draw boundaries between specialised care and other services. For example, a patient admitted as an emergency with epilepsy to a hospital with a neurosciences service might well be cared for by a neurologist (and counted as specialised), whereas a physician from another specialty might deliver very similar treatment to a patient admitted to a general hospital.

Similarly, distance can deter patients from accessing even life-sustaining services such as chemotherapy and radiotherapy. But we don’t know why this and other types of variation exist, and proximity to (or distance from) a tertiary centre is only a partial explanation. 

All of this means there is no easy way to allocate money to geographical areas for many of these services. While it is possible to pursue equity in outcomes, equity of access or funding is difficult to determine and even harder to achieve.

In fact, this is a rather uniquely British preoccupation. Other European countries do not take a geographical approach to resource allocation for highly specialised service in this way. In the Netherlands, for example, the approach is to pay subsidies to university clinics and tertiary providers and use the diagnostic related group (DRG) reimbursement system.

Although NHS England seem intent on moving to a new allocation method history, international experience suggests this will only work for some of the higher volume specialised services that should not have been on the list in the first place. 

Smaller-scale organisations have long struggled with specialised commissioning

Specialised commissioning was devolved to primary care trusts (PCTs) in 2002. The mismatch between needs-based allocations and historic patterns of spending soon became a problem for budgetary control, as commissioners whose populations had high rates of use of these services had no methods for controlling utilisation.

This was compounded by the rarity of the conditions that meant that, with the exception of some relatively high-volume areas (such as chemotherapy, radiotherapy and some cardiac services), there was little chance that local commissioners could develop sufficient expertise in specialised commissioning and limited incentives for them to do so.

Small numbers of cases meant there were large random fluctuations between years that could be smoothed over within big populations, but were a major issue for a small commissioner such as the average-sized PCT. 

Attempts to deal with these issues by pooling resources between PCTs led to disputes, with PCTs that had windfalls from lower levels of use reluctant to help out others.

All of these issues made it difficult for PCTs to manage specialised commissioning effectively, even though guidance issued in 2003 had “re-affirmed the need for PCTs to establish collaborative commissioning groups to coordinate the commissioning of specialised services, and for SHAs to oversee and performance manage such groups”.

The 2006 report by Professor Sir David Carter recommended that PCTs pool their risks and create consortia to do specialist commissioning at regional and national level. Carter also found that PCTs were very poor at cooperating, which in turn created difficulties, including for tertiary centres that had to manage a large number of different admission criteria, exclusions or other aspects of the pathway or commissioning policy.

The creation of national and regional specialist commissioning groups that followed the Carter review improved the situation. However, the health committee still found they were “hobbled” by the fact that PCTs still retained statutory responsibility for the budget, and by a poor level of participation and corporate working by PCTs.

They were unwilling to share sovereignty or resources or to pool risk, and even though the specialist commissioning groups were subservient to the PCTs, there were no mechanisms to join local and regional commissioning. The committee also found that specialist commissioning groups were not subject to rigorous oversight and performance management.

Moving forward: why decentralisation is being considered

There are some good reasons why decentralisation of some more specialist commissioning is being considered. When the national and regional specialist commissioning groups were established in 2006, they were responsible for a much smaller portfolio of services and proportion of the NHS budget than is the case now, and it is clear that NHS England has found the size of the portfolio and its diversity challenging

There may be some less good reasons, including the temptation to hand over an overspending budget or remove some of the opportunities for lobbying. There have always been incentives for patient groups, providers and pharma companies to get their service into the specialised services portfolio. The budget was seen as more generous, less well controlled, and more amenable to direct lobbying – including via parliament. Once a service was listed, it proved difficult to delist it without a tussle with well-organised patient groups and providers.  

Putting some specialised services into local commissioning or integrated care organisations can help to align incentives and lead to better service integration or investment decisions. For example, combining commissioning for transplantation with dialysis aligns the incentives for commissioners to increase transplants to control the costs of dialysis.   

These advantages are theoretically appealing, but in joining up one set of services (such as renal replacement therapy and transplant) there is also a risk of creating new fragmentation with others, such as prevention, diabetes management and hypertension reduction. A number of these services have opportunities for trade-offs, interfaces and handoffs, which need to be managed.

There is not an obvious structural solution that best resolves all these problems, and there is no substitute for those responsible talking to each other to coordinate decision-making. An approach is required that combines some of the advantages of scale with the ability to join up decision-making more locally. 

Potential models

Chemotherapy, radiotherapy, cardiac services and renal dialysis are areas of relatively stable activity that are sufficiently large for it to be possible to develop adequate commissioning expertise at a more local level. The STP footprint would generally be large enough, although some of these have a multi-way relationship with tertiary centres that might require some work. 

In mental health, work is already underway to develop provider collaboratives generally made up of NHS providers. As a first step, these will commission:

  • Child and adolescent mental health services (CAMHS)
  • Adult low and medium secure services
  • Adult eating disorder services.
  • Over time a range of other specialised mental health services, such as perinatal, forensic CAMHS.

Collaboratives should help local leaders improve continuity in patient pathways, ensure incentives are better aligned to outcomes, and decide whether it is better to procure these services externally or to provide them within the collaborative.

A lead provider model might work quite well for areas such as specialist neurology. For example, the Walton Centre provides neurology support across a range of sites in the north west by working with local hospitals. Other specialised areas that have a clear pathway, and components that can be bundled together, might also be suitable for this sort of lead provider model.

A key question is how to ensure that the consortium or lead provider gets proper oversight, and that its decisions are based on what is best for patients rather than its own priorities and needs. Failure to do so could see patients having to travel to tertiary centres for clinic appointments that could be delivered locally; or conversely, being cared for in mental health units within the provider collaborative when they would benefit from specialist care available outside. 

Gaps between local and national commissioning responsibilities can result in problems with flow to and from specialist services. Local providers often struggle to get their patients accepted by specialist centres, but they in turn often experience problems returning their patients to local care. For example, patients who have had neurosurgery followed by specialist neurorehabilitation will still need to transfer to local less specialist providers.

Discharge problems for specialist centres are often exacerbated by a shortage of rehabilitation capacity, and therefore their beds are occupied by patients who could be cared for elsewhere.   

What needs to happen next?

Reform of specialised commissioning has been back on the agenda since the Five Year Forward View, and it forms part of the thinking about the development of ICSs – with decentralisation and a focus on ICS-level population health management featuring strongly.  

In 2002, decentralisation led to the devolution of services that were unsuitable, financially volatile and high risk. If the push to decentralise commissioning continues, care should be taken to avoid a similar ‘overshooting’ resulting in the devolution of services, where the resource allocation mechanism is not clear, where there are large random fluctuations in use, or where it is not possible to develop commissioning expertise.    

NHS England will need to develop clear criteria for which services can be devolved, and it will remain responsible for the commissioning of these services until legislation changes.

It will also be useful to set out a development pathway that aligns information and outcome measures in the first instance. Incorporating needs assessments from ICSs into specialised commissioning, as well as joint work on the disjunction between local and specialised pathways, could form a valuable part of this. This could include tackling the question of the long tail of specialist work where there would be benefits from rationalisation.   

ICSs and CCGs are currently light on the sort of infrastructure required to commission specialised services, and local commissioning expertise will need to be developed. Regions could work with ICSs to pool commissioning expertise in areas such as mental health, so that ICS staff contribute to a wider pool of knowledge and develop individual areas of specialised commissioning interest on behalf of the region.

There is also further work to do on definitions, pathways and standards. Some areas such as neurosciences are working on better (and often tighter) definitions of what is truly specialised. There are clearly benefits to having national standards and pathways developed once rather than multiple times, and work has already been developed in this area that should not be lost.

The emergence of provider collaborative models is an interesting and important change, and builds on successful previous experience in this area in mental health. However, some formal evaluation would be of value to identify the strengths and issues that arise from this approach – in particular the impact on patient experience.

There is a dilemma. There are undoubtedly areas where a more joined-up approach to specialised commissioning would add value to population health management at an ICS level. It could also increase opportunities for prevention, improve step-down care, speed up discharge and create more holistic care. 

But there are challenges in developing the expertise, financial frameworks and other parts of the infrastructure, at a time when ICSs already have a very full agenda. As well as putative benefits, there are also risks from devolving too much or too quickly. Fortunately, there are some important lessons from past experience that can help to get this right this time.

Suggested citation

Edwards N (2020) “Specialised commissioning: lessons from history and looking to the future”, Nuffield Trust comment.