How should payment systems evolve in the new era of integrated care?

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.

It is difficult to have a conversation in the NHS currently without raising the issue of ‘accountable care’. Accountable care is a term that has been subject to broad interpretation, but is underpinned by the notion of coordination of care for a defined population within a predetermined budget. This contrasts with the prior dominant policy narrative of provider competition incentivised by payments for particular services or bundles of care (in England, this has been enforced by the ‘payment by results’ system introduced early this century). 

In the refreshed planning guidance for 2018/19, NHS England and NHS Improvement state that they “are now using the term ‘integrated care system’ as a collective term for both devolved health and care systems and for those areas previously designated as ‘shadow accountable care systems’. An integrated care system is where health and care organisations voluntarily come together to provide integrated services for a defined population.” In this long read we will therefore use the term ‘integrated care’. 

Capitation payments

Integrated care is often associated with ‘capitation’ payments – that is a risk-adjusted global payment per patient that covers many or all of that individual’s health services. When aggregated, this provides a ‘population’ budget that is not varied according to actual service use. Hence it is associated with a degree of financial risk being borne by the recipient of the budget and an incentive to manage the health and care proactively and cost effectively. 

Capitation budgets have long been used to fund general medical practice in the UK, and to set the overall commissioner budgets of clinical commissioning groups (CCGs) who are responsible for planning and purchasing care for local communities.

In truth, while capitation may be an important facilitator of integrated care, a suitable payment system will be far more complex and there is the potential for multiple payer relationships (and forms of payment) within an integrated care system. It is worth clarifying terms prior to any discussion of payment systems.

At the highest level, a clear distinction can be made between the ‘payer’ or funder of integrated care and the providers of the services that are delivered. In the English context, the payer is the CCG that takes on the responsibility of commissioning care and holding providers to account, and the provider is an organisation or group of organisations that provide care to the population in question. 

But it is also the case that the relationships between the providers will vary within different integrated care systems. At the simplest level, a single, integrated provider may be commissioned by the payer to provide all relevant services. However, it’s far more likely a coalition of providers will be required, and these providers may form a new joint venture, may loosely collaborate or may be led by a single one of their number who acts as a ‘prime’. It is immediately clear therefore that provider relationships may themselves be underpinned by a variety of different payment systems.

A further challenge comes in the form of boundary issues. In early thinking about integrated care, local systems were often conceived of as self-contained and serving distinct populations. While convenient for developing the ‘theory’ of integrated care, in practice this is rarely the case. 

In urban areas in particular, there are significant and complex overlaps in the catchment areas of providers. When specialist services are considered, this is a matter of more than local geography. There are many services that can only be provided effectively and efficiently on a regional or national scale, and an over-reliance on local capitated budgets could serve to fragment these services. 

Integrated care therefore cannot be considered a puzzle that is simply ‘solved’ by capitation funding. Capitation (particularly between payer and provider) may be part of the answer, but it is not sufficient in itself. 

Payment options

So what options are available for funding integrated care? Historically, there is a wide range of funding mechanisms that have been employed globally in health care systems, depending on the different policy goals. Each approach to funding has strengths and weaknesses, and needs to be considered in light of the goal or ‘blend’ of goals that are being pursued (for example, quality, access to services, cost containment and so on). The main approaches are explored here:

The dominant financial allocation system in use within the English NHS for acute care is a case-based payment system (‘payment by results’). This accounts for approximately 70% of NHS hospital funding. It is rooted in the historical desire to improve access and to facilitate competition between providers as a means of improving quality and efficiency of care. There is some evidence that this policy has been successful in those terms. 

However, while case-based payment systems may be suitable for incentivising access (and, less convincingly, quality), it also incentivises an episodic approach to care where providers compete for funding rather than collaborate to provide integrated care across organisational boundaries. It can encourage operational efficiency within silos, but not allocative efficiency across a system. 

Integrated care is founded on a different mix of policy objectives – better, and more cost-effective, population health through a range of mechanisms such as proactive care management, increased health promotion and disease prevention, and the development of care pathways that extend beyond organisational boundaries. These are the policy objectives that have emerged from the current environment of austerity.

If the development of integrated care is successful, it will significantly reshape the current provision. In particular, care is likely to shift from hospital settings to primary or intermediate care settings. The new care model vanguards have shown modest success, in that early evaluation findings suggest they have reduced the rate of increases in emergency admissions to hospital. Similarly, early pilots of integrated care suggest that the development of ‘care management’ in primary care can lead to a fall in overall hospital costs.

The implications of such changes to historical patterns of care are that the costs of care will also shift settings (and therefore likely shift across organisational boundaries). If more patients are treated out of hospital, it is reasonable to assume (and certainly it is the intention) that the acuity, and therefore the average cost, of those patients who do get treated in hospital will rise.  

Designing a payment system to support integrated care

So what might a payment framework look like that would incentivise the desired behaviours of providers within the system and ensure that the available resources are channelled effectively within that system? An effective payment system for integrated care is likely to embody a number of characteristics:

• Provides incentive to providers to collaborate rather than compete, where this generates greater benefit for the population.
• Incentivises a shift of care to the most appropriate setting from a cost and quality point of view – thereby delivering improved quality and cost efficiency.
• Recognises that the costs of treatment, and the point where those costs are incurred, will change as new clinical pathways are agreed.
• Avoids duplication of effort within a single system (in terms of both operational and capital spend).

• Stranded costs are overheads to providing a service that cannot be easily scaled back or released as a service shrinks or is decommissioned – for example, the ongoing costs of servicing debt used to purchase obsolete assets, continuing to purchase capacity from essential but underutilised staff, and maintaining now underutilised physical infrastructure.

  Provides a mechanism to remove ‘stranded costs’ from the system without overtly penalising the organisational holder of those stranded assets.

Stranded costs are overheads to providing a service that cannot be easily scaled back or released as a service shrinks or is decommissioned – for example, the ongoing costs of servicing debt used to purchase obsolete assets, continuing to purchase capacity from essential but underutilised staff, and maintaining now underutilised physical infrastructure.

It is unlikely that any one payment method that we have described here will alone be sufficient. As we have seen, all payment systems come with advantages and disadvantages. Instead, a blended payment system that incorporates a number of approaches is most likely to incorporate the characteristics above and should be preferred. 

While any such solution will need to be bespoke and grounded in the needs and structure of an individual local system, core elements that we believe should be part of the blend include:

• Block payments underpinned by transparent ‘whole-system accounting’: A substantial amount of the payment made to any provider within an integrated care system should be based on an estimate of the efficient cost of providing that organisation’s contribution to overall care. 
  This ‘revenue equals cost’ approach contrasts with the current case-based system (for hospitals), where national tariffs reflect an often out-of-date average or a normative cost of care unrelated to any specific provider. The block payments described would require ‘open book’ accounting between partners to both validate estimated costs and the impact of any in-year changes to plan. Payments will also need to include necessary transition costs to support the removal of stranded costs and other disruption caused by the implementation of new care pathways.
• Risk and gain share payments: In complex local systems with multiple providers (i.e. most of them), providers need to find a way to align the incentives of peers to their own and those of the payer. This can be achieved by making a portion of payment dependent on the system as a whole meeting its targets, such that when this is the case everyone is better off, and when it isn’t everyone is worse off.
• Payments based on quality and non-financial outcomes: skimping on either the volume or quality of care is a common market failure in health systems. Placing more financial risk on providers increases the chance of this type of behaviour. One way to counterbalance this is to make a portion of payments contingent on the achievement of quality and outcomes targets. 
  While the current NHS payment system features elements of outcomes-based payments (quality and outcomes framework, commissioning for quality and innovation [CQUINs]), these are often focused on process adherence within a single organisation, while in the context of integrated care the outcomes that affect payments should be increasingly the products of coordinated whole-system action.
• Continued use of case-based payments where appropriate: some element of case-based payment is likely to be both inevitable and desirable. The necessary ‘trading’ of care between neighbouring systems will require a ‘currency’ if it is to be managed effectively. This is needed to support patient choice (integrated care systems could appear as a restriction on choice if patients are unable to seek care outside of them) and to allow the development of specialist services on larger geographical footprints. 

In developing such blended approaches, there will be a temptation for systems to create increasingly complex combinations of incentives to drive ever more specific behaviours. This, however, would be a mistake. For a payment approach to successfully incentivise change, it must send easy-to-understand signals to the system and be aligned to participants’ core mission and values.

While now much maligned, it can be argued that payment by results was successful in achieving its main policy objectives of increased access and reduced waiting times, because it very clearly incentivised hospitals to increase capacity and compete for activity. Similarly, if blended payment approaches are to effectively support the objectives of integrated care, they will need to be crafted to achieve a similar and elegant simplicity. 

The move towards integrated care and population health management is a key component of NHS England’s Five Year Forward View. We are now more than half way through that plan and rapid progress is required if new payment approaches are to meaningfully contribute to its delivery. The use of blended payment approaches that build upon existing arrangements and infrastructure offer local systems a pragmatic means of making real progress in this important endeavour.