What should payment mechanisms achieve?

Improving health outcomes

Improving health outcomes (improving quality and length of life) is ICSs’ first purpose. The NHS Long Term Plan set out particular priorities to improve health outcomes for areas such as cancer, mental health, dementia, cardiovascular and respiratory conditions, among other areas of care and targeting groups such as children and young people. [44] Improving health outcomes will require improving care quality, care timeliness (covered by allocative efficiency, below) and core determinants of health (including supporting people to live healthier lives and make healthier choices). 

While NHS and local government partners can play an important role in addressing the wider determinants of health, many such determinants are shaped by other areas of public policy (such as welfare, housing and education services) outside of ICSs’ control. Addressing the wider determinants of health will therefore require an integrated approach from national government.

 

Improving care quality

Improving care quality across all care settings is vital to improving health outcomes and patient experience, including waiting times. Since 1989, there have been several attempts to improve health outcomes and to improve care for people with long-term conditions (such as the Quality and Outcomes Framework, QOF). While there have been some significant improvements in the management of some conditions such as cardiovascular disease in recent decades, there continues to be high levels of variation in quality of care in many areas, and considerable variation in resulting demand for acute care. ^iv  The Long Term Plan included targets to reduce certain waiting times, but the COVID-19 pandemic has inhibited progress. 

In particular, the rising number of patients presenting with long-term conditions is contributing to poorer patient outcomes and an increase in economic inactivity (a challenge government is looking at through the Major Conditions Strategy). [45,46,47] Earlier interventions are required to prevent conditions worsening, social prescribing to stop conditions from developing, and more healthcare at home and in the community to support people to live with their conditions. As one in four adults has at least two health conditions, rather than treating each condition separately, services need to be integrated so patients can be treated holistically. [48]

 

Reducing health inequalities

The 2010 Marmot Review drew attention to the issue of rising health inequalities – reducing health inequalities is now one of ICSs’ four core purposes. [49] As well addressing the wider determinants of health, reducing health inequalities requires improvement in access to services (including proximity of local services, opening times, transport to services) and analysis of population health data so resources can be targeted at groups most in need. NHS England has developed the Core20PLUS5 approach to support ICSs efforts to define target populations and accelerate improvement to reduce health inequalities. [50] Sir Michael Marmot’s work describes a continuous gradient of health inequality to which proportionate universalism is often an appropriate response.

 

Boosting allocative efficiency

Improving allocative efficiency is crucial to helping address many of these objectives, including improving outcomes and financial sustainability. This is central to the mission of ICSs which seek to provide their populations with the right care, in the right place, at the right time. Increasing length of stay in acute beds, and mental health and community inpatient units, is linked to a range of factors, including the provision of social and community care, and drives up costs with a negative impact on patient outcomes. [51,52,53] Too many people are attending A&E or being admitted to hospital who could be better cared for elsewhere. There are significant productivity gains from avoiding urgent care admissions and managing these in different settings. 

This contributes to the NHS spending an increasing proportion of its overall budget on acute care. [54] NHS spend on acute care increased by £10 billion in real terms (17 per cent) between 2020/21 and 2021/22, now accounting for over half (53 per cent) of total system spend. ^v  Since we know that spending on public health, primary and community care has a greater correlation with economic growth (every £1 spent correlated with £14 in gross value added, GVA, to the economy), than spending on acute sector healthcare services (correlate with £4 in GVA), supporting a leftward shift in resources may also help to improve economic growth. 

There has already been much positive progress in primary care to expand capacity and integrate care, with a growth in multidisciplinary team working and associated increases in consultations and consultation rates, and boost secondary prevention.55 However, despite the introduction of primary care networks, structural reform of primary care has been slow with still high variability in quality of care, access to care, resulting demand for acute care and the way practices are organised, with many as yet not able to deliver the scale of change set out in the Fuller stocktake of 2022. [56] More substantive change will be required – accompanied by increased proportion of spend on primary and community care – if improvements in health outcomes and associated reductions in acute activity are to be seen. 

Alongside this, ICSs also want to put more resources into primary prevention, that is taking action to reduce the incidence of disease and health problems within the population, either through universal measures that shape the wider determinants of health (such as lifestyle risks) or by targeting high-risk groups. This can include addressing health behaviours such as smoking cessation services, promoting physical activity and deploying health interventions such as effective vaccination and screening programmes, to prevent and identify ill health in the community. Some of these services are commissioned and delivered by the NHS, others by local government partners within systems. Some are going further and considering the NHS’s impact on air quality and the climate.

 

Boosting technical efficiency

Technical productivity remains a significant challenge for ICSs with considerable opportunity to improve processes to boost output. For example, operating theatre throughput remains highly variable, with considerable late starts and early finishes resulting in fewer patients being treated; A&E throughput has slowed despite increasing resources and flat demand (though complexity has increased), length of stay has increased considerably despite strong evidence demonstrating the negative impact of longer stays. ^vi

Similar challenges occur in community care, with variable numbers of contacts per whole-time equivalent and in community mental healthcare where despite significant increases in workforce, activity has not kept up and similarly lengths of stay in inpatient units are highly variable. [57] Together, this results in considerable waste with resources being tied up in existing models of care rather than being refocused on improving health through a greater focus on prevention and primary care. The 2016 Carter review estimated that by reducing unwarranted variation, particularly in clinical services, at least £5 billion of savings could be made annually in acute hospitals. [58] There are also significant opportunities from a big expansion in the use of virtual wards. [59]
 

Example models of care to incentivise

ICSs are already making progress to achieve these aims and there are examples of effective services and behaviours across the country.

Such services and behaviours need to be upscaled and spread. While ICS leaders are already trying to bring providers together to achieve these aims and avoid siloed working, change remains very slow. Financial integration and stronger financial incentives could help to integrate services for patients and lead to better outcomes and more efficient care. An ideal financial system would incentivise those changes need to accelerate these efforts. Currently, too often the current system stops them.
 

Five types of payment mechanisms

Before considering which payment mechanisms could be best suited to ICSs in England, we need to briefly consider the existing types of payment mechanisms used for healthcare services.

Block funding

Block funding is the payment of a lump sum to a provider for a defined service or services, at regular intervals, regardless of the population served or the amount of activity undertaken. Block payments are often linked to an anticipated volume of output activity. These block budgets are currently used in community and mental health care and were widely used across all care settings during the COVID-19 pandemic. Block funding can allow providers the flexibility to choose where to spend money between services, including to move services leftward/upstream as necessary and improve allocative efficiency, particularly when used across different care settings. The administrative burden and transaction costs are low and the capped funding provides consistency for both providers and commissioners.

Financial risk primarily sits with providers, who do not receive additional funding when rising demand for care leads to additional activity (such as urgent and emergency care), although this is often mitigated through cap and collar arrangements. In cases of controlled demand (such as with waiting lists), the risk is transferred from providers to commissioners who need to pay irrespective of levels of individual provider activity. However, block contracts offer few incentives to increase volume of activity, quality or technical efficiency in these settings. [67] Block funding for services in specific settings can also drive siloed working, with little incentive to collaborate with other providers.

Payment by activity (PbR) and fee for service (FFS)

Activity-based payments see providers paid for each activity, for instance a hip-replacement, they provide to a patient. The incentive for an organisation that does more work is further funding for that additional work, leading to a higher volume of care to more patients (assuming prices are set at the right level) and encourage greater technical efficiency, as providers keep marginal gains from doing more activity for a lower cost (although maximum activity levels can be capped). 

Payment by activity aligns with patient choice rules, as the money follows the patient to their choice of provider in the public or private sector (although there are disadvantages of this, for instance making it harder to for commissioners to prioritise patients with the greatest need and leaving public providers with the higher complexity cases with increased costs without additional income). 

Evidence from England and internationally indicates that payment by activity can lead to reduced length of stay, reduced waiting times and reduced unit costs. [68] In England, the introduction of activity-based payment in the acute sector in the early 2000s, introduced at a time of long waiting lists, helped to encourage activity in hospitals and has driven improvements in throughput and productivity.69 The introduction of PbR was accompanied by higher than historical average growth in the budget of the NHS, compared to more recent growth in investment (with the exception of spending during the COVID-19 pandemic). [70,71] It was also facilitated by wider changes – such as the creation of foundation trusts and a strong regulator (Monitor), with considerable financial incentives such as being able to keep their surpluses. 

However, without strong payment mechanisms to control demand (as increasingly exist in more active payor markets such as Germany and the US), there is concern that payment by results (PbR) incentivises additional low value activity – activity which may not be necessary or could be addressed in more efficient interventions. For example, it can incentivise providers to admit more people to hospital than is strictly necessary, carrying out lower value-added procedures when conditions could have been treated at a lower cost and with better health outcomes at an earlier stage. Payment for activity has not been adapted to encourage earlier intervention and prevention rather than more acute activity. [72,73,74,75,76] It does not necessarily incentivise coordination and integration with other service providers payment for activity therefore risks being a counter-vailing force against ICSs’ mission to shift resources to earlier, preventative, more cost-effective interventions. 

Annual budget cycles undermine the incentive for additional activity as extra resource that has been earned is taken away from the provider organisation at the end of the financial year. Payment by activity has comparatively high administrative costs, although only a small proportion of the productivity opportunity, and may only incentivise more profitable, low complexity activity which may not always correlate with patient need. [77] Over time, adjustments to PbR have made it difficult to evaluate its long-term impact and some of the initial incentives for individual organisations have similarly changed. PbR has not been tested in England as a payment for non-acute providers. 

To be successful, payment by activity needs strong commissioning skills to incentivise required activity, not low value activity. It also needs to be part of a wider strategy to move existing monies within a system. [78] Some argue there is still a role for PbR to eliminate the current elective care backlog, particularly given the substantive reductions in productivity and throughout (such as cases per theatre list, length of stay). However, it may not be effective without addressing wider system challenges (such as difficulties in discharging patients, particularly given capacity in the social care system), an increase in funding relative to demand and may be at odds with an emphasis on prevention. 

Fee for service (FFS) has also been used in primary care over the last 20 years, with a focus on additional or enhanced services alongside capitation-based and quality payments. FFS tends to incentivise specific preventative interventions, for instance vaccinations, with a good margin on the tariff at the expense of a population health approach or general preventative activity.79 However, many countries used FFS to drive activity in primary care – Denmark, France and Australia being the notable examples. [80] Experience of using fee for service in general practice in Denmark demonstrates the importance of balancing fee for service with capitation in order to mitigate incentivising unnecessary activity. 

Different international incarnations of PbR and FFS vary in the degree to which they differ or are similar, often in the extent to which they bundle or unbundle different the costs for each element of a care spell. However, both PbR and FFS have mainly been used to incentivise the number of existing ‘results’ and ‘services’. Both the names, however demonstrate that the financial incentives are for ‘results’ and for ‘delivering a service’. It is therefore possible to define a result or a service in a different way and pay for that. At the moment, PbR and FFS fund inputs rather than outputs and outcomes. It would be possible to define a ‘result’ or a ‘service’ through an output or an outcome rather than an input and incentivise those achievements rather than inputs.

Payments for quality and outcomes

Quality and/or outcomes payments incentivise improvements in the quality of care and patient outcomes. Outcomes-based payments can provide payments for different kind of results – not just outputs (ie hospital activity) but the outcomes (improvements in patients’ health) which are intended. This can allow providers to keep savings from making patients healthier when they spend less money on activity to achieve this result, incentivising both technical and allocative efficiency. It requires commissioners to be able to effectively measure agreed outcomes and use this to inform payments and future commissioning decisions. 

Existing use includes the Quality Outcomes Framework (QOF), a pay for performance scheme for primary care introduced in 2004, and Commissioning for Quality and Innovation (CQUIN) implemented in 2009, paying NHS trusts for specific deliverables. Best Practice Tariffs (BPTs) – with prices and payments reflecting the cost of best practice, rather than average costs of procedures and services – were also introduced in 2010 following the Darzi Review to try to reduce variation and incentivise best practice in four high-volume clinical areas. [81] 

QOF financially rewards GP practices for meeting a range of targets in the delivery of care. In 2023/24, these covered four domains: maintaining disease registers; clinical - improving the management of chronic diseases; public health - addressing key risk factors to health and improving vaccination rates; and quality improvement – improving workforce wellbeing and addressing capacity challenges. [82] This therefore includes both outcome and output payments. An evaluation of QOF found that it had incentivised general practice to have a more organised approach to chronic disease management and incentivised some secondary prevention. [83] It has embedded behaviours in general practice linked to its incentives. Conversely, an evaluation of the abolition of QOF in Scotland observed reductions in recorded quality of care for most performance indicators, suggesting changes to pay for performance should be carefully designed and implemented. [84] 

However, there is little evidence that QOF has driven a significant increase in primary prevention and public health activities in particular. Some GPs have criticised QOF for its increased administrative burden and its negative impact on providing person centre-care through the pursuit of a narrow set of process orientated, clinical quality indicators rather than patient outcomes. [85] To a certain extent, QOF may have led to an overly medicalised, mechanistic and transactional approach, which undermines integration. Additionally, QOF has often focused on output activity, not health outcomes, which limits innovation in service delivery to meet desired aims and the ability of general practitioners to consider the variety of drivers of health. The conclusion is process indicators in schemes such as QoF are useful only when they have been shown in randomised trials to be strongly and causally linked to important outcomes. [86] Of course, payment by outcomes relies on the availability of accurate data. This is already used to some extent for talking therapy services. 

Consistent tinkering since QOF’s introduction and additional payment measures, such as the Impact Investment Fund (IIF) and PCN Direct Enhanced Service (DES), have reduced the evidence-based nature of many of the indicators and added complexity and bureaucracy which has hindered its efficacy. Various sets of data collation and reporting requirements are required to enable payments. While there are likely to be changes to QOF and the IIF which are (at the time of writing) currently being discussed as part of the 2024/25 GP contract and PCN DES, a revision of QOF might focus on rewarding population health outcomes and incentivising performance improvements above an established baseline, minimising changes and complexity, re-establishing the use of evidence based indicators. [87] Simplicity is a key lesson to help payment mechanisms lead to the desired behaviours. The GP Contract consultation process also highlights the value of co-creation to ensure payment mechanisms are effective in incentivising intended behaviour. 

For CQUINs, evaluation suggests that the impact of CQUINs against desired results has been disappointing and the costs of the scheme outweigh the benefits.88 This is in part because the data needed to administer payments adds a significant administrative burden, particularly when the range of indicators for which payments are made proliferates. Potentially as a result, CQUINs are being scaled back in the proposed payment scheme for 2024/25. Using data from the established national clinical audit programmes may be option to consider. An early evaluation found higher support for BPTs than CQUINs among frontline clinicians who viewed it as evidence-based and with fairer payments.89 Evidence on its effects on quality and patient outcomes was mixed – the evaluation found some relative improvements but at the same time unintended consequences. 

A quality-based payment mechanism focused on waiting times is being trialled in West Yorkshire (see case study below) to incentivise waiting list reduction. Use of a block payment for acute elective services, which is modified by rewards and penalties for beating or exceeding waiting times targets, across the provider collaborative, enables clinical leads to innovative in service design and potentially allocative efficiency. Meanwhile, the payment element tied to waiting times, a metric of quality, incentivises delivery on a key national priority. However, although an evaluation is still pending, it is not clear that this approach necessarily helps to drive technical efficiency. 

Learning from local and international experimentation, any outcomes-based payment relies on the ability of commissioners and their providers to define and measure a set of clear indicators. This may involve a lengthy engagement process as well as an investment in the right tools to measure and analyse data and a transfer of risk from commissioners to providers (which should be backed up by appropriate reserves). These implementation costs should be considered but are likely to benefit from advances in data analytical systems now available. Additionally, any outcomes-based payments need to consider the size of the population cohort to avoid any cherry picking of patients with less complex conditions where outcomes targets could be more achievable. 

Outcomes-based payments should cover as much of the care pathway as possible, ideally the whole pathway, to incentivise the most effective interventions. Different providers then have to collaborate to focus resources on the interventions, interventions which will be out of the immediate control of some individual providers. This shift in commissioning across pathways, rather than particular services, goes beyond financing to the wider commissioning model. It can be a crucial enabler of supporting a leftward shift to earlier, more preventative interventions and is an important consideration going forward.