Feature / Allocation, allocation, allocation

In 2013, the new world of clinical commissioning will get under way – presuming the health bill is enacted. Before then, organisations need to be set up and, just as important, new relationships forged. Getting the right governance and accountability arrangements in place will be key to the success of the new system, as will ensuring the right level of support is available for new commissioners. But there is perhaps an even more fundamental piece of the new architecture to get in place – ensuring the commissioning groups get the right amount of money in their new budgets.

In 2011/12 £85bn of primary care trust allocations were informed by a weighted capitation formula. The formula uses population or capitation as the starting point for sharing out the overall pot of PCT allocations. But because not every member of the population needs or uses the same level of healthcare, each PCT’s population is weighted for age, gender, extra need and unavoidable costs (using the market forces factor):

Weighted population =

pop x age index x add need index x MFF index

Because the drivers of need for services are different – for example, for hospital care and prescribing – there are three components to the capitation formula: hospital and community health services or HCHS (79% of the formula); prescribing (11%); and primary medical services (10%).

The whole approach will change with the introduction of clinical commissioning groups, whose budgets will be built up from the commissioning budgets of their constituent practices. It is not simply a case of feeding practice-based populations into the current PCT formula, cranking the handle and then waiting for the budgets to drop out.

Need for new formula

A completely new formula is needed that can accurately set budgets for smaller populations (GP practices with an average list of 6,500) rather than the larger populations of PCTs (averaging 330,000).

The problem with the current weighted capitation formula is that it uses the characteristics of the population in an area, not the needs or characteristics of individuals. Yet people living next to each other could be registered with different GPs. So the area’s characteristics – levels of deprivation or standard mortality rates – are not necessarily appropriate for a practice-based population. At PCT level, these issues tend to even themselves out across the bigger numbers involved, but they can lead to problems at the practice level.

Fortunately the new formula did not have to start from scratch. A formula known as the person-based resource allocation (PBRA) has been under development for some time to support the development of practice-based commissioning. It was included in 2010’s PBC toolkit to inform PCTs’ setting of nominal budgets for practice-based commissioners.

The new formula, originally developed and now being refined by the Nuffield Trust, takes a new approach. Martin Bardsley, Nuffield’s head of research, describes it as a ‘more focused’ way of allocating resources. Rather than looking at the broad characteristics of a population in a wide area, it considers the needs and likely usage of services by individuals registered in an individual practice, which can be summed up to create commissioning group budgets.

The key difference in the model used is the additional use of ICD10 diagnosis codes to flag up individuals likely to incur a higher cost in the following year as a result of having been diagnosed, say, with diabetes or cancer.

Mr Bardsley says there has been much debate over the right casemix descriptors to use to provide the best predictor of future cost. In the end a ‘simple’ approach involving about 150 groupings of diagnosis codes was adopted (groupings already used by the NHS Information Centre). Healthcare resource groups would not have been appropriate.

‘Many HRGs are defined just in terms of process,’ says Mr Bardsley. ‘So you’d get a marker because you had an operation, not because you had the condition.’

The new formula only focuses on a portion of the current weighted capitation formula, looking to set budgets purely for hospital care – excluding maternity, mental health and prescribing. The formula is being looked at by the Advisory Committee on Resource Allocation. It will recommend whether the model is adopted this year with options for other services. The Commissioning Board will ultimately need to agree on the approach.

Size matters

As with all statistical modelling, it is not an exact science. While offering a better fit than the current weighted capitation formula, the person-based formula is also influenced by size. The larger the practice, the more likely the allocation is to be accurate. The ability to predict what happens to an individual is relatively poor – these models are not a crystal ball. But when you group together all the people in a practice or commissioning group, this statistical variability begins to average out. That’s why the size of a practice or a clinical commissioning group will be important.

The Nuffield Trust will suggest adjustments to the Department later this year. But time is of the essence. While the new system formally goes live in April 2013, shadow budgets – using the new formula if ACRA recommends it – will be needed for April 2012.

Even with a formula in place, decisions must be taken that will have an impact on the budgets received by clinical commissioning groups. Currently the weighted capitation formula sets a target allocation or fair share allocation for a PCT. The PCT is then moved from its baseline allocation (broadly its previous year’s actual allocation) to its target at a rate dictated by a pace of change policy.

Baseline setting

A similar approach is expected to be applied with the new formula for practices/commissioning groups. But having never received firm allocations before, commissioning groups do not have a baseline from which to start. A big bang approach – simply applying the formula in full to initial allocations – would lead to instability for some health economies.

Last month PCTs were required to submit a breakdown of current spending, broken down to practice level. In a letter from NHS chief executive Sir David Nicholson in August requesting the information, there was recognition that the exercise would ‘require some degree of estimation’. Not all PCTs will be able to accurately identify all expenditure down to practice level, particularly outside national tariff activities.

As Mr Nicholson’s letter pointed out, the development of a fair shares formula is only part of the story. ‘Efficient resource allocation also requires a good understanding of current spend at local level. This will allow the NHS Commissioning Board and the Department to assess the need or otherwise for a pace of change policy.’

Many argue it is the baseline exercise that is the biggest influence on an organisation’s allocation, certainly in the short term.

History suggests any pace of change policy is likely to be slow in times of limited overall growth. Previous pace of change policies have taken a differential growth approach – everybody receives some growth, while those furthest under target receive a bit extra.

However, there has been a very open policy of not making real cuts to those that are most over target, with the inevitable consequence  that gaps between the furthest under-target and the furthest over-target tend to close slowly.

Even with the formula in place and being used to set (or inform) allocations, other mechanisms will be needed to manage risk across health economies.

No formula can accurately predict exact resource usage by a small population. If one practice or commissioning group has an abnormally high level of high cost patients one year, it will create local pressures.

Sharing this risk through some form of local top-slicing or pooling arrangement will need to go hand in hand with the new approach to allocations.