Is health secretary Andrew Lansley’s flagship GP commissioning policy ever going to work? I mean; now the Treasury has got its hands on it?

We know the folk with the calculators and spreadsheets are concerned about spending controls.

But will the available fixes spell a de facto end to the radical change Mr Lansley wants to see: GPs incentivised by the prospect of keeping their surpluses to reinvest in patient care?

At the forefront of the Treasury’s collective mind is the question of controlling overspending. There is evidence to suggest this could be very significant if GPs were in charge: At least £1.2bn a year significant.

£1.2bn is the gross overspend sum which can be derived from research by the Nuffield Trust, due to be published soon.

Nuffield didn’t set out to find a hole in the policy. Rather, they have done detailed work on establishing a new GP commissioning allocation formula because the success of the policy rides, in part, on GPs believing the budget they are being asked to stick to is at least credible.

But in the course of their work, they looked at the amount 8,238 GP practices in England spend through their patients’ use of acute hospital services and they compared that to each practice’s target budget.

On their least extreme analysis they found 53 per cent of practices under or overspend by more than 5 per cent, with 26 per cent missing their budgets by more than 10 per cent and 11 per cent by more than 20 per cent. (A more extreme analysis – on a different iteration of the formula – found 70 per cent adrift by more than 5 per cent.)

HSJ has been told the proportion under and overspending is broadly the same; so the net position is breakeven.

What does that mean for gross overspending if the majority of the NHS budget were under the control of practice based commissioners?

On a £60bn total NHS commissioning budget, 13 per cent (half of the 26 per cent) of practices overspending at a rate of 10 per cent would result in an overspend of £0.78bn. Add another £0.4bn from 14 per cent overspending at a rate of 5 per cent and the total overspend reaches £1.2bn. This is a cautious calculation excluding those practices missing their budgets by more than 20 per cent.

It’s only fair to note here that these calculations are based on Actually Existing PBC - the shadow, “indicative budget” PBC that Mr Lansley says does not offer the same vivid incentives as the real, hard cash PBC he intends to establish. It’s his ability to offer those incentives that is now in doubt.

The Nuffield figures are based on an analysis only of the acute spending element of GP budgets, which makes up around 60 per cent of the total.

Yet the findings are replicated in data published by the DH, setting out each practices’ entire indicative budget for 2010-11 and their trends on NHS service use.

That data suggests under and overspending averages 8 per cent on either side. On a £60bn budget that would reach £2.3bn a year. 

The DH figures are a little limited because each practice’s allocation and use are presented in terms of their share of its primary care trust total, not as an absolute amount. Thus it is possible that in some cases, practice X’s “overuse” might really be practice Y’s “underuse”, because underuse by one practice increases the overall proportion used by another. 

But the absolute cash amounts in PCT accounts confirm the problem is real.

For example, data from Devon PCT covering its PBC accounts for 2007-08 shows that 55 of its 107 participating practices overspent their collective £163m budget by £6.7m: 4.1 per cent. Luckily for Devon, the remaining 52 practices under-spent their £122m budget at a similar rate, leaving £4.9m unspent.

Scaled to a £60bn NHS-wide commissioning budget, that level of gross under and over spending brings us back to £1.2bn.

In Devon’s case, the 4 per cent on either side pretty much balanced, leaving the PCT with a net overspend of £1.8m; just 0.6 per cent of the budget it devolved to GPs.

On a £60bn budget that 0.6 per cent would reach £360m: Annoying for the Treasury, but probably not enough to start contradicting the health secretary. 

But it’s not at all clear practice based commissioner gross under and overspends will be available to balance each other out in the same way, because Mr Lansley says the incentive for GP commissioners should be to keep any of the savings they generate to reinvest in care for their patients.

It’s then difficult to see GP overspending being treated in the same way as the NHS deficits of 2005-06 (and the scale of the gross overspends is similar). GPs are unlikely to accept top slicing and brokerage with the same good grace as PCTs.

That means the DH – and in turn the Treasury – could lose twice.

First, the £1.2bn of overspending gets spent.

Then the £1.2bn under-spending gets stored in discreet GP commissioning group balances to save up for something nice, like a polyclinic. And heaven help a health secretary who asks for any of that “back”, to prop up a deficit-ridden group elsewhere.

That then bounces PBC policy into the same difficult box as foundation trusts: GPs can accumulate surpluses, but if they choose to spend them in a year the rest of the NHS is in deficit, the DH breaches its spending limits. The Treasury’s cash flow could also struggle, if surplus commissioning “resource” (essentially Treasury IOUs) outstrips available bank notes.

The answer offered to this problem is for GPs to group into risk-sharing consortia, to internally balance out under and overspending.

But you don’t need to be over cynical about GPs to imagine surplus-generating practices preferring to consort with other surplus-generating practices, leaving the overspenders to form the GP consortia equivalents of bad banks.

Others might want to consort with practices not in the same geographical location as them – something the Right is particularly interested in as it could spell the start of a social insurance-based system.

Yet Mr Lansley has until now said he is against dictating the size and form of consortia, emphasising the importance of liberating clinicians from bureaucracy.

Academics have argued the ideal size consortium covers a population of 100,000, as that allows sufficient risk pooling to cope with rare high cost cases – as well as balance under and overspending.

A 100,000 population consortium would be akin in size to Hartlepool, Darlington and Bassetlaw PCTs now – at least in terms of patients registered on their GPs’ lists.

Others are a lot bigger; perhaps someone will have the bright idea of splitting them each down the middle?

With the inevitable requirement to manage GP expectations over how free they are really to spend those hard earned surpluses, it’s hard to see how the reinvention of another rather clunky commissioning bureaucracy is not on the cards

To adapt the phraseology of the now-infamous senior DH source: “PCTs are screwed”… Long live PCTs.


Another “solution” would be to ditch the fair shares concept, at least for now, and set all commissioning budgets at what they use THIS YEAR. It'd create a bit of a cruddy incentive for those who have been under using for years who wouldn't then see any benefit (surplus to reinvest). And would mean it was the biggest over-refers/spenders who would win - at least in the early years. It could also be the starting gun for wise GPs to start upping their referral rates asap – so as to inflate their baseline. Go!