How QOF pay weighting fell apart

Statistical errors left QOF pay weighting for 2009/10 in tatters. Stephen Robinson reports on what went wrong.

Dr Bailey: prevalence problem is ‘unintended and counter-intuitive’  (Photograph: JH Lancy)
Dr Bailey: prevalence problem is ‘unintended and counter-intuitive’ (Photograph: JH Lancy)

Practices across England have been deprived of thousands of pounds as statistical anomalies largely wiped out disease prevalence weighting on QOF pay.

Figures from the 2009/10 QOF show nine out of 10 practices received identical pay per point in eight clinical domains - worth 20 per cent of all QOF points.

In some domains, up to 99.7 per cent of practices received the same pay weighting.

As a result, practices with high prevalence that would normally receive a pay boost to deal with additional work were given the same rates of pay as low-prevalence practices.

GPC Wales chairman Dr David Bailey describes the problem as 'unintended and counter-intuitive'.GP leaders believe the 2010/11 framework has now rectified the issue.

But pay distribution for 2009/10 was hugely unfair. The unique situation arose from a perfect storm of unforeseen events.

In 2004, GP leaders agreed that QOF pay should be adjusted to account for prevalence. This recognised the added workload from targeting a large number of patients with a particular disease.

The effect was cushioned by uplifting practices with 5 per cent or less of the highest prevalence rate reported for each disease domain to protect them from losing out on QOF pay, and a square root adjustment to rein in high prevalence practices.

Diabetes worth 100 points in QOF
  Practice A Practice B
  2009/10 rules 2010/11 rules 2009/10 rules 2010/11 rules
Disease prevalence 7.1% 7.1% 0.2% 0.2%
QOF points earned 100/100 100/100 100/100 100/100
Prevalence after any adjustments 7.1% 7.1% 7.1% 0.2%
Prevalence factor (practice prevalence / national mean prevalence) 0.97 1.27 0.97 0.04
QOF pay for diabetes domain £12,285 £16,090 £12,285 £483
In 2009/10 QOF, two practices with very different diabetes prevalence receive the same pay (assuming equal list size and maximum points achievement). Under raw prevalence in 2010/11, practice B no longer receives the 5 per cent uplift, causing its pay to drop. Removal of uplift also lowers national average prevalence: therefore, practice A benefits with a larger prevalence factor, increasing its pay.

Unfair system
But some felt the system was unfair. They argues that practices with three times the prevalence had three times the work - and hence deserved three times the pay. So, in 2008, the GPC agreed to scrap the adjustments and switch to weighting pay according to raw prevalence.

For 2009/10, square rooting was dropped, but the 5 per cent uplift remained. At the same time, flaws in date collection meant dozens of practices recorded artificially high disease prevalence of up to 520 per cent.

These extreme outliers caused 'normal' practices with high prevalence to fall within the uplift and so receive identical pay per point as practices with almost no prevalence (see example above).

Similar effects may have occurred in other UK countries, but insufficient data are published to explore this in detail.

Dr Bailey defends the GPC's decision to scrap square rooting in 2009/10 and leave the uplift in place until April. The DoH was thought to have been in favour of scrapping both adjustments in one year.

'Originally it seemed most who would get hit would be hit by removing the 5 per cent uplift. So we gave them time to adjust,' he says.

Jon Ford, head of the BMA Health Policy and Economic Research Unit, admits the situation is 'bizarre'. 'The idea of getting more than 100 per cent prevalence is nonsense,' he says.With hindsight, the prevalence adjustments should never have been introduced, he adds. 'When the GMS contract was agreed, the fears about prevalence may have been overstated.'

Poor timing of data collection is largely responsible for last year's huge outliers. Practice list data is recorded on 1 January, the latest date available to allow QOF payments to be calculated at the start of April. But prevalence data are collected in March.

A practice with five patients on 1 January that gains 10 more with a disease by March would have a prevalence of 200 per cent. This was further exacerbated by the creation of rapidly-growing Darzi centres.

Patient care
The hit on practice income from 2009/10 is likely to affect this year's budgets. But RCGP clinical lead for diabetes Dr Brian Karet is confident these problems will not affect care. Practices will ensure important clinical work is done regardless, he says.

Dr Bailey says the new system of raw prevalence should eliminate these issues. 'It will be fairer; there won't be a repeat of the problem in the new system,' he says.

Not everyone will agree. In April, GP reported some university practices with low prevalence stand to lose six-figure sums (GP, 22 April). Moreover, those with artificially high prevalence will continue to earn inflated sums per QOF point.

Mr Ford admits last year's problems were unfortunate, but says the QOF is improving.

'QOF is regarded as an organic process: as we go through, it needs tweaking. Over the years it's proved its worth. It is better now than it was two years ago.'

QOF Prevalence Timeline

April 2004 GMS contract introduced; square root and 5 per cent uplift applied to moderate effect
of prevalence


October 2008 GPC announces details of switch to raw prevalence after concerns over fairness


April 2009 Square root mechanism dropped


April 2010 5 per cent uplift dropped


September 2010 GPC warns 'many practices have been adversely affected' by changes


October 2010 QOF figures reveal the extent of the damage


Have you registered with us yet?

Register now to enjoy more articles and free email bulletins

Register

Already registered?

Sign in

Before commenting please read our rules for commenting on articles.

If you see a comment you find offensive, you can flag it as inappropriate. In the top right-hand corner of an individual comment, you will see 'flag as inappropriate'. Clicking this prompts us to review the comment. For further information see our rules for commenting on articles.

comments powered by Disqus
Follow Us: