GP and QOF expert Dr Gavin Jamie writes about the Global Sum/Carr Hill formula which is the baseline for GPs on GMS contracts - and where the money from the withdrawal of MPIG is going
It's over a decade since the name Carr-Hill became notorious to GPs. A new contract had been presented offering, among other things, a new way of funding general practice. Prior to this, practice income was based largely on the number of patients and GPs at the surgery with some adjustments for deprivation within the practice area.
The new contract proposed a whole new method of calculating practice income. This was to be based on evidence of practice workloads and would be responsive to the needs of a practice's population. Initially there were few details about how it was to be calculated. One fateful Monday practices received an estimate of their proposed income - the Global Sum. By lunchtime it was clear that there were big problems. Almost every practice looked set to lose money; in many cases these were large sums that would make the practice financially unviable. Internet forums buzzed as GPs compared figures and sent feedback to GPC negotiators.
By the middle of the week a solution had been produced. There were already transitional arrangements designed to smooth the path from one funding formula to the other. These were originally to last for three years but would instead become permanent - the Minimum Practice Income Guarantee (MPIG) would always be a top-up on whatever the Carr-Hill formula said. Around 90% of practices would receive something from this scheme.
Of course nothing is forever in the NHS and it was only a few years until the rules changed and the MPIG started to be eroded as the Global Sum increased.
This year things have gone further and the government plans to withdraw the MPIG altogether over the next seven years, putting the cash into the Global Sum pool. That pool becomes even more significant as QOF points have been withdrawn and over 200 of them moved into that pool.
The Carr-Hill formula, as the Global Sum calculation became known, has remained largely unchanged over the last decade although its popular name perhaps disguises its origins. Professor Roy Carr-Hill actually only produced some of the research and statistics which led to the formula being produced by negotiators. Whoever is responsible it is increasing in significance and it is now more important that practices understand how their income is calculated and how changes in practice populations can affect their income.
The formula works by applying a series of weightings based on the profile of patients in the practice. The lowest weight that a patient can have is one. All of the other factors that are applied serve to increase the weighting of individual patients. After each adjustment this will give a total weighted list that is larger than the actual number of patients registered with the practice.
A similar thing is happening at PCT level where a separate formula (the Weighted Capitation Formula) is used to calculate a weighted capitation at the PCT level. This is a more complicated and more negotiated system to try to work out how resources should be allocated to PCTs.
Practices have their weighted list cut back proportionately so that their total weighted lists equal the PCT’s overall weighted list. For instance if the total of practices after adjustment was two thousand and the PCT weighted total was one thousand patients then all practices would have their weighted list size halved.
The practical upshot of this adjustment is that practice income is dependent on how cash is distributed to the local area as well as their own practice profile. The cost to the NHS locally is fixed, whatever the formula calculates for individual practices. Two practices with identical populations may have different global sums depending on where in the country they are, although the differences are unlikely to be huge as both practice and PCT formulae share similar aims.
The largest adjustment is based around the age and sex of the patients registered with the practice. The group with the lowest weighting are men between the ages of five and fourteen. They carry weighting of one. At the other end of the scale a female patient aged eighty five or over carries a weighting of 6.72. Thus a practice will receive over six times as much after registering an elderly female patient as a ten year old boy. In general female patients carry a slightly higher weighting than male patients although this is most apparent between ages 15 and 44 years old where female patients carry twice the weighting. This is likely to be due to the workload effects of contraception and cervical cytology in women of this age.
The only age group where males outweigh females are children under five years old although the differences are fairly marginal.
As the largest component of weighting it is probably the one that has caused most difficulties for practices. University practices with their young populations have been particularly affected. It is not impossible for a university practice to receive half of the payment per patient of a practice covering an area popular for retirement. The differences between young and old are quite dramatic. It has been suggest that the current formula underestimates the workload on practices generated by young adults.
There is a further adjustment when patients live in a nursing or residential home. Each of these patients is worth 1.43 times a typical patient. Note that the weightings for each patient do not multiply together. Patients in residential care tend to be elderly and will attract a relatively high weighting for this but the effect of multiple weightings is more like addiction than multiplication.
If their weighting due to age is five that will not increase to seven if they are in a care home - it is closer to five and a half.
In fact the weighting of 1.43 will be applied to a typical patient at the practice, although you never know what a typical patient will be worth until the end of the calculation.
This is one areas where practices can act fairly simply to optimise their income. Unlike the other components of the global sum calculation residential institutions are not detected automatically by the NHS payment systems. A specific code needs to be entered into the appropriate field on the practice computer along with the registration information for each patient. Without this the practice will not receive the uplift - worth around £35 per patient. Finding patients resident in homes and making sure that the correct residential code is entered is vital. You local registration department should be able to supply a list of codes for nearby institutions.
There is a further adjustment for list turnover. Patients are more likely to attend shortly after they have registered with a practices and for this reason they have a weighting of 1.46. This will help practices as they register new patients. Similarly where practices have been subjected to list cleansing exercises and end up re-registering a lot of patients this may provide some compensation. However practices will not see the benefit if this is happening across the PCT area and all practices are similarly affected.
The other factors that are taken into account are more complex for practices to calculate themselves.
There is some account taken of morbidity and mortality using national indexes. The Standardised Limited Long-Standing Illness (SLLI) and the Standardised Mortality Ratio for those aged under 65 (SMR<65) are not figures that appear at the tip of GPs’ fingers. The former appears to exist almost entirely within the contract regulations. Both of these measures are calculated at the level of electoral wards - based on the patient’s registered postcode. If a practice does not have representative sample of patients from a particular ward then this adjustment will be less accurate. Electoral wards can be large and contain quite a bit of variation; a particular practice may cover only areas which are more or less deprived than average.
Rurality is a term that is little used outside the contract. This is an adjustment that is based on the average distance of patients from the practice. If you have a branch surgery it is the distance from the main surgery that counts. The extra weightings are not that large. A practice with patients an average of eight miles from the main surgery will only have a ten percent uplift over a surgery where patients are an average of one mile away. This might, however, be a small incentive to keep open a branch surgery with associated, more distant, patients.
Part of this adjustment is also based on population density. There is an increase in payment for areas which are less densely populated. The effect of this will depend on how much variation there is in density across the PCT area.
There are further adjustments for staff costs. The most obvious of these is London weighting but there are other weighting factors around the country which are calculated in the same way throughout the NHS. Once again there is relatively little that practices can do about this component.
It is worth practices checking the information that is supplied from their local area teams that is used to calculate their global sum. Although most of the figures will be difficult for practices to verify directly obvious errors should be questioned.
Direct knowledge of weightings may be more useful to practices as they consider changing their practice list. Adding or losing one thousand patients from an area of young families will produce a very different effect on both income and workload from an area with more elderly patients.
Whilst practices should not, and are contractually prohibited from, selecting patients within their practice area, decisions about how to develop the practice or deal with a need to reduce the practice area need to be considered in a way that will ensure the continued viability of the practice.
These calculations are unchanged from the time the contract was first introduced ten years ago. The weightings were informed by observations of practice behaviour which took place some time before that. At that time the observations were largely based on how long each patient’s record was opened on practice computer systems as a surrogate measure of workload.
Even over the time relatively short time of ten years primary care medicine has changed dramatically and the workload has increased, in large part due to increasing standards of monitoring and treatment of chronic disease. It is quite likely that the weightings from the early part of the millennium do not reflect current workloads.
The QOF allocated resources according to disease prevalence and tied resources directly to morbidity. The global sum, as we have seen, is based much more strongly on social and demographic details.
There is pressure to update the formula and the negotiators have declared an intention to do so. This has, however, been suggested before and eventually has come to nothing. Any change is likely to be disruptive to practices. Essentially the MPIG set practice income according to the pre 2004 contract. There will be a total of seventeen years of transition between the two schemes as MPIG is withdrawn.
Any change in the formula will be more likely to affect practices at the extremes the most. Practices whose list has a high weighting are more likely to see that fall. Practices with a low weighting are more likely to see a rise in income from any change to the formula.
Nothing is forever in the NHS.