A flaw in the quality framework’s prevalence formula has stripped thousands of pounds from quality pay for practices in England.
The formula was designed to make sure that practices with high numbers of chronically-ill patients were rewarded fairly for the work this involves.
But an investigation by GP revealed that fewer than 3 per cent of practices received a positive prevalence weighting in the mental health domain in 2005/6. Weighting on stroke and COPD points was significantly affected.
The highest mental health prevalence recorded by an individual practice, in data published this month by the DoH, was 35.4 per cent, roughly 60 times the national average of 0.6 per cent.
As a result of the way practices’ prevalence pay weighting is calculated, this single high score cancelled out all positive mental health weighting for more than 8,100 English practices, out of a total of 8,406 participating in the quality framework.
Prevalence weighting on quality pay is calculated from the range of prevalences recorded by practices in each clinical domain of the quality framework. Practices in the bottom 5 per cent of the range are rounded up to the 5 per cent mark to ensure they receive some pay for the points achieved.
Once this has occurred, an adjusted national average is calculated. This average is square rooted, and each practice’s prevalence is square rooted, to narrow the total range.
Each practice is given an adjusted disease prevalence factor (ADPF), calculated by dividing their square rooted prevalence by the national square rooted prevalence.
Because of the high prevalence at the top of the range for mental health, 8,179 practices in England with prevalence ranging from zero to three times the national average received an identical ADPF of 0.99 in 2005/6.
This means they all received the same negative weighting on mental health quality pay, making the average point worth less than the basic £124. More than 3,000 practices with above-average prevalence fell into this category.
One practice in East Lincolnshire with three times the national average prevalence for mental health and almost exactly the national average list size achieved 41 points in 2005/6.
But it received an ADPF of 0.99, and was paid £122.76 per point, roughly £5,030 in total before list size was accounted for (see box, p15).
If the range of prevalence for mental health excluded the 35.4 per cent prevalence reported by the top practice, the East Lincolnshire practice would have been paid £7,626.
Stroke is the second most affected domain, with a prevalence of 23.8 per cent, 15 times the national average. Almost 3,000 English practices received the same ADPF for stroke of 0.86.
No practices with above average prevalence for stroke received negative weighting, but the range was skewed by some high-prevalence practices. Three practices in England had prevalence above 5.5 per cent for stroke.
In Northern Ireland and Scotland, there have been no outliers in the domains affected in England.