While much of the noise around the latest round of NHS reform is about abolishing NHS England and the promise of neighbourhood health teams, there has been little scrutiny about the central role that commissioning is set to play in the latest version of the health service.
Commissioning covers a wide range of activities within the health service, from detailed transactional work related to service planning and contracting, right through to strategic oversight and system leadership. It has been a feature of the NHS in England since the National Health Service and Community Care Act 1990, which created an ‘internal market’ that separated health service commissioning (then called purchasing) from provision. That commissioning layer of the NHS has since been the most reorganised part of the NHS, which is probably the most frequently reorganised health system in the world.
In the latest long-term plan for the NHS, commissioning is changing once again. ICBs are being merged and required to significantly reduce the size of their organisations. The new model integrated care board (ICB) guidance and the strategic commissioning framework very much positions ICBs as the driver of commissioning. This is in parallel with the development of neighbourhood health models and new models to create integrated providers.
As the NHS embarks on yet another attempt to develop effective commissioning, this time based on organisational units the size of medium-sized European countries, it is worth reflecting on some lessons from history and the opportunities that a new focus on neighbourhoods and population might provide.
Given past and international experience, what needs to happen for commissioning to work more effectively than it appears to have done in the past?
Lessons from the last 30 years
A review in 2018 of the implementation of “strategic purchasing” – what commissioning is called in other systems – in 10 countries in Europe found that none fully measured up in terms of their performance or capabilities in any of the definitions proposed by different authorities, which indicates how challenging the task is.
The follow-up to this review has the encouraging title, Power and purchasing: why strategic purchasing fails, and a number of the issues that this identifies are highly relevant to the NHS. The counterfactual provided by Wales and Scotland, which have largely abandoned commissioning (but also have different approaches to accountability and performance management) is difficult to interpret, but does not suggest that the lack of a commissioning-based approach has held them back.
The fact that policymakers and the NHS itself have so frequently reorganised the NHS’s commissioning layer suggests a long-standing problem with the model of commissioning, or at least a serious gap between the expectations of policymakers and reality. The abject failure to achieve the long-stated goal of shifting work from hospital to primary and community care noted in the Darzi Review also suggests a flaw with the approach over the last 30 years.
We can draw several lessons that would increase the chances that ICB-led strategic commissioning will perform better than the district health authorities, health authorities, primary care groups (PCGs), primary care trusts (PCTs) and clinical commissioning groups (CCGs) who came before them.
One important issue is that commissioners have often been given an overambitious scope. The NHS model of commissioning has been much more concerned with specifying the details of how care is to be provided than payers in other health systems. In the chronic disease management programmes and bundled payment models used by payers in Germany and the Netherlands respectively, the standards were set by external clinical experts and there was a tight focus on a small number of high-impact chronic diseases. Neither has been very successful.
Large numbers of objectives and the attempt to use commissioning to implement complex policy instruments have added to the problem. Other payers and integrated delivery systems tend to operate with far fewer key performance indicators (KPIs). Commissioners in the NHS appear to be more proactive and involved in detail than payers in other systems – intervening to change models of care and pathways or redesign whole systems of care.
However, they do not seem to have proportionately more resources to take on this role, so the scale of the task has often been too great for the generally small resources available. More seriously, commissioners have been on the wrong side of a major information asymmetry, which has made it difficult to understand and manage the level of detail they have been expected to encompass.
In any case, it is clear that commissioning skills and capacity have never been adequate to carry out the task demanded. The King's Fund reported in 2010 that 80% of GPs felt they lacked necessary commissioning skills, and similar capacity gaps have been mentioned in PCTs, CCGs and now ICBs. Frequent reorganisations have compounded this problem – often destroying institutional knowledge and relationships before they could mature.
The information asymmetry was often part of a wider provider capture of the commissioners, particularly the smaller versions of commissioning organisations. Shifting care from hospitals to other forms of care leaves stranded costs, and investment in the new services has not allowed hospitals to release cash, which creates a significant obstacle to change. In health systems with multiple payers, hospitals have the option to make up the shortfall in fixed cost coverage by finding replacement income from other payers.
It is also the case that payers in these systems often have more financial flexibility – they don’t spend 100% of their annual income and can manage their budgets over multiple financial years. They also tend to operate in systems that have many more hospitals and are larger than NHS commissioners (up to now), so they are less in thrall to an individual provider.
A financial problem in an NHS trust caused by strategic commissioning decisions very quickly has become a problem for the commissioner, either through intervention by the levels above (region or NHS England) or because the provider has found other ways to bill for work that the commissioners found difficult to control. It is arguable that all of this has meant that risk has been misallocated, with commissioners holding too much of the risk for overspending due to higher volumes.
The lesson from integrated models in Spain, New Zealand, Israel and the USA is that providers need to be doing much more of the heavy lifting in redesigning care, developing pathways and finding ways to deliver more cost-effective services. To some extent, NHS commissioners have often let hospitals off the hook.
These factors have come together in the expectation by policymakers and commissioners that they could effectively manage demand. Commissioners have been able to do this in some cases through changing the pathway for patients, developing new services, setting referral criteria and working with providers.
However, this has only been partially successful, and hospital activity has increased significantly above the levels expected given changes in the population. This is not just the case for non-elective work, as the potentially more controllable outpatient, diagnostic and elective categories of activity have also shown significant growth. This reflects the disconnect between commissioning and front-line clinical decision-making, and a failure to appreciate that new technologies, increased treatment intensity and other factors not related to demand are very important drivers of activity.
One of the reasons for this failure and a more general problem has been a lack of integration of the commissioning of primary care and the management of the GMS contract into the wider job of commissioning. The relative weakness of the GMS contract as an instrument for making change has not helped. This has also been an area where a lack of capacity has been an issue.
Commissioners appear to have been more successful when they approached their role using a mix of methods based on collaboration with providers and other stakeholders. But in many cases, they have attempted to micromanage provider behaviour using contracts, procurement and payment mechanisms. This has not generally been successful.
Writing complete contracts and using micro-incentives to change provider behaviour effectively are very difficult, but NHS commissioners appear to be much more willing to try this than payers in other countries. The information asymmetry and expertise problem applies here, but transaction cost economics suggest another reason for caution. In situations where there are high asset specificity (it is difficult to redeploy them), relatively high transaction costs and issues with the ability to measure outputs, then culture, management and hierarchy are more efficient and effective models of governance than contracts and externalised transactions.
The use of procurement is also a blunt instrument, and commissioners do not (and probably cannot) have the capacity to specify what they want in detail across the whole range of services they are buying. Even if they could, the complexity of interactions between different service areas would be a challenge, not least because it is not possible to optimise a complex system by individually optimising its component parts.
Highly detailed specifications also risk removing the scope for provider-led innovation and assume that the commissioners have the knowledge to predict which innovative approaches will work. It also often has the effect of excluding smaller providers and the voluntary sector, which is frequently a source of innovation. All of this has contributed to a situation where significant commissioner resources have been dedicated to bureaucratic and often low-value contract management and procurement, with an emphasis on compliance with the specification rather than population health.
A further issue is that commissioners’ attempt to change delivery is generally based on an assumption that there is a strong connection from contracts, payment models and specifications to clinical behaviour. This is often not the case.
In spite of the interest in a degree of micromanagement, commissioners have made limited use of utilisation management compared with payers in the USA and (to a lesser extent) in other parts of Europe. Some European countries use prior authorisation for expensive treatments. There is some monitoring of utilisation in Germany, where the sickness funds employ medical review services that do prospective, concurrent and retrospective reviews. Some countries have payers that monitor clinical pathway compliance with varying level of clinical challenge. With the exception of GP gatekeeping and the use of budget caps on providers, the NHS is relatively weak in this area when compared to the approach elsewhere.
The international review points out that, in addition to the information asymmetry, there are also often power imbalances. Hospitals have been seen as having better staff and superior information systems, and tend to be more politically influential. The barriers to change are often political rather than technical. This is also the case in other systems.
Research examining 123 National Clinical Advisory Team reconfiguration proposals between 2007 and 2012 found that "a significant number of proposals do not get implemented as planned, largely as a result of local public and/or clinical opposition". This and other factors may explain the lack of decommissioning that has made it more difficult to support innovation, experimentation or the double running often required for implementing new models or service redesign.
The frequent reorganisation and resizing of commissioning organisations may suggest some uncertainty about the right level at which decisions should be taken, and the search for a single configuration for functions that often have a different optimal scale.
The NHS seems unclear about decision rights – who is supposed to be taking which types of decisions and to what level these should be devolved, or what should be standardised. A lack of appropriate standardisation has led to unwarranted variation in access and outcomes and a wide variety of approaches to service delivery across the country. Even within small areas, it is not uncommon to find teams with overlapping mandates and slightly different referral criteria set up for similar purposes. Where there has been standardisation, this is often not adequately enforced – for example compliance with formularies and rules for prescribing, especially for dispensing GPs.
Problems with misaligned policies have been an issue in the past, and they remain an issue now. While one part of current policy stresses neighbourhoods and population health, which require a system approach, other parts stress more the use of market mechanisms and provider autonomy and a less system-oriented approach, such as the move away from system control totals. Moving provider performance management to the regional level is part of this, which creates the potential for conflicting priorities and a risk that this will reduce the incentive for providers to prioritise ICBs and neighbourhoods instead of the imperatives of regional performance management.
A lesson from Portugal is that developing a focus on population health and outcomes is difficult if the performance management system is concentrated on operational performance issues. There is a risk that the way the National Oversight Framework and its associated machinery are constructed will mean that commissioners will experience a similar tension.
Finally, patient and public involvement has not been as effective as it might. The reasons for this are complex, and may reflect the upward-facing nature of accountability in the NHS and a general lack of investment in this area.
There have also been positive lessons, and commissioners have made change happen very effectively in many places. Working with clinicians, local authorities and other stakeholders has been an important part of creating change. Where they focused on particular pathways or service areas, they appear to have been more likely to make a significant difference. Being close to the providers and communities – working collaboratively rather than in an adversarial way, while showing clinical leadership and being evidence based – also seem to have been important factors in this.
Implications
The new guidance envisages four stages of strategic commissioning:
- Understanding the context: build a dynamic, data‑driven picture of population needs, risks and current service performance.
- Developing a long-term population health strategy: set long‑term, evidence‑based plans and redesign pathways to create integrated neighbourhood care models
- Delivering through payer function and resource allocation: allocate resources, commission services and shape the market to deliver strategic population health outcomes.
- Evaluating impact: measure the impact of services and interventions using utilisation, risk, experience and outcome data.
Functions 1 and 4 are not particularly problematic, but if the intention is that the ICB will be doing most of the long list of tasks in numbers 2 and 3 itself then it seems likely that they will fail for the reasons already outlined. While the ICBs will have better data than previous incarnations of commissioning, they will be very much more distant from the areas they are responsible for and comparatively less well resourced. To succeed, the ICBs will need to develop their approach in new ways that learn from previous experience, while taking advantage of the emerging governance and delivery models developing at place and neighbourhood level.
There are strong arguments for a high level of devolution or delegation, given the information asymmetry problems and the extent to which the role involves working with communities, local authorities, places and neighbourhoods, and the difficulty that commissioners have when they are attempting to manage demand or volume risk. Many decisions about the design and delivery of services will be better taken closer to the communities being served. To succeed, ICBs will need to ensure there are strong place-based partnerships with the capability to contribute to ICB strategy and operationalise it.
History and international experience suggest there are limits to what can be achieved through payment systems and contracts. The logic of a devolved model for population health implies more use of capitation payments, combined with gain-sharing and performance management, rather than trying to micromanage pathways and individual behaviour.
Some areas where incentives for higher volumes are needed will benefit from using an activity-based payment model, but different approaches are needed for those where the objective is to have less or better care. As already noted, experience elsewhere is that it is the providers who will need to be doing much of the redesigning and implementation of pathways. The ICB’s job is to create the incentives to do this and the financial frameworks to manage the fallout, not to do it themselves.
Devolution will require clarity about what the ICB or the NHS centrally expects to be standardised across the system, and how this is expressed on a continuum from design principles and standards to detailed specification of areas such as pathways, policies or drug formularies.
Neighbourhoods and places will have to operate within a clear strategic framework and the ICB will need to be responsible for creating this. The guidance says that “each ICB should involve local stakeholders to determine the approach it uses to develop its plan”, while the population health improvement plan should take account of the neighbourhood health plans produced by health and wellbeing boards within the ICB. The logic of devolution and learning from past lessons strongly suggest that the plan should be much more co-produced than this suggests. The ICB will need to specify what is required from a top-down perspective, but to succeed there will need to be plenty of space for co-production from the bottom up.
The role of the ICB will need a clear articulation of decision rights, the population health and process indicators by which systems will be held to account, and efforts to ensure that the risk of these crowding out population health improvement objectives can be avoided. NHS England will need to work through these questions, and changes to their approach to performance management will be needed. An important lesson is to avoid overloading commissioning with too many objectives and different areas for action, not least because of the limited capacity that ICBs will have.
To support this tiered approach to commissioning, the ICB will need to be:
- challenging places to improve value, allocative and technical efficiency, and identifying opportunities to do this – there are range of methods for assessing allocative efficiency and making more value-based commissioning decisions
- constructing financial frameworks to allow money to shift across the system and to support innovation and double running – this will need some new mechanisms to deal with the stranded costs that may emerge for providers where there is a shift in activity
- developing the approach to primary care commissioning to support neighbourhood health, reduce inappropriate variation and to improve equity in resource allocation
- agreeing a consistent approach to decommissioning based on a sound ethical and methodological framework
- developing solutions for issues that require cross-place decision-making
- developing approaches to ensure that voluntary, community and social enterprise (VCSE) organisations and non-NHS providers are fully included in place- and neighbourhood-based services
- providing support to places where there are economies of scale or a need to pool effort – this would include primary care contracting, workforce strategy, estates and property, and other areas where specialist expertise is needed
- creating an environment that supports innovation and research.
Where applicable, the ICB will have a key role in working with mayors and combined authorities on the wider agenda relating to prevention and economic development.
The development of devolution to place and neighbourhoods needs to be set in a wider consideration of the level of devolution to ICBs – is this sufficient or are there some functions that would be better done at a supra-ICB or regional level? These would include areas where there are economies of scale and no particular need for a detailed understanding of local context.
The potential problems of misaligned policies and the tension produced by the different accountability lines and policy pressures will need to be addressed. Regions will need to ensure that providers are working with commissioners, and in particular that they are supporting the development of neighbourhood models of care. The current expectation that commissioners will intervene or get involved with performance or other problems in providers is not consistent with their role as a strategic commissioner.
Conclusion
For ICB-led strategic commissioning to succeed where previous models have struggled, significant reorientation is required. ICBs must resist the temptation to micromanage services through detailed contracts and specifications. They should instead embrace their role as system orchestrators – creating the conditions for innovation rather than attempting to design every detail themselves.
Action will need to encompass three interconnected areas. First, ICBs will need to delegate decision-making and operational commissioning to place and neighbourhood level, where proximity to communities and providers enables more effective responses to local needs. This devolution must be accompanied by clear strategic frameworks that define what needs standardisation and what requires local variation – establishing decision rights and accountability measures that focus on population health outcomes rather than process compliance.
Second, ICBs need to shift the balance of risk and responsibility to providers through capitated payment models combined with an approach that measures population health, value and outcomes. Providers, not commissioners, should lead pathway redesign and service innovation – they possess the clinical expertise, information and operational control necessary for successful transformation. ICBs should create the financial frameworks and incentives that enable this provider-led innovation while managing the system-wide implications of change, including support for double running and decommissioning.
Third, ICBs must build genuine co-production into strategic planning, moving beyond stakeholder consultation to authentic partnership with places, neighbourhoods, local authorities, and communities. This includes ensuring that VCSE organisations and non-NHS providers are fully integrated into neighbourhood models, and that patient and public involvement becomes substantive rather than ceremonial.
Critical to success will be resisting policy overload. ICBs have finite capacity and must concentrate on high-impact priorities: challenging places on value and efficiency, developing cross-place solutions, providing specialist support where economies of scale exist, and intervening where local systems require development support.
Regional teams and NHS England must align their performance management approaches with these neighbourhood-focused priorities – ensuring that providers are incentivised to support population health rather than to respond to conflicting accountability pressures.
The model ICB guidance provides the right components, but success depends on execution. ICBs that devolve operational authority, trust providers to lead clinical transformation, build genuine partnerships, and maintain disciplined focus on strategic priorities can break the cycle of commissioning failure. Those that attempt to replicate previous models at larger scale are likely to achieve similar disappointing results. The choice between system orchestration and centralised control will determine whether this latest reorganisation delivers meaningful improvement in population health.
Nigel Edwards is a Senior Advisor to PPL and to the European Observatory on Health Systems and Policies.
Please note that views expressed in guest articles on our website are the authors' own and do not necessarily reflect the views of the Nuffield Trust.