No reason for a QOF witch hunt
If exception reporting is an invitation to game, the vast majority of GPs appear to have declined, says Dr Tim Doran
As the NHS's great pay-for-performance adventure moves into its fourth year, the Department of Health is looking to modify the quality and outcomes framework in the light of new findings. One of its main concerns is that the provision to exception report patients is being exploited.
Exception reporting was intended to protect patients, by removing the temptation for GPs to subject patients to inappropriate treatment in their pursuit of financial incentives.
However, by exception reporting otherwise appropriate patients for whom they fail to achieve QOF targets, practices can boost their achievement rates and hence their rewards.
As the word 'fraud' can be highly emotive, such tactics are usually referred to as 'gaming'.For such an important part of the QOF – and one so open to abuse – relatively little is known about exception reporting.
From the outset, nobody had a clear idea of what an acceptable or appropriate level of exception reporting would be.
In the first year of the QOF, exception reporting data was not even collected nationally, leading to some extravagantly high estimates of rates.
Within the department this fuelled a fear of widespread gaming, and led to suggestions that exception reporting be removed from future contracts – drawing a predictably angry response from GPs.
There was therefore a great deal of interest in this month's Information Centre report1, which presented actual exception reporting data for the first time. The overall exception reporting rate of 5.6 per cent for 2005/6 did not exactly scream 'fraud'.
If exception reporting is an invitation to game, it is an invitation most practices appear to have declined. The highest rate of exception reporting was for CHD10 (the percentage of CHD patients treated with a ß-blocker), a treatment for which common contraindications exist.
As practices were only required to achieve this target for 50 per cent of patients, and as it only offered a modest financial incentive of £875 for the average practice, it was not an obvious candidate for gaming.
Conversely, BP5 (the percentage of hypertension patients with a BP of 150/90 or less) was a prime candidate, worth a more tempting £7,000 and having a more demanding maximum achievement threshold of 70 per cent. Yet BP5 was not one of the most exception reported targets.
Less drastic alternatives
Nevertheless, the Department of Health's fears are not entirely unfounded: examples can be found of practices that excluded over 90 per cent of patients from particular indicators, and 52 practices excluded more than 15 per cent of patients overall. Is this gaming? If it is, is it sufficient cause to remove the exception reporting safeguard, given the potential consequences for patients and practices?
Less drastic alternatives certainly exist. Of all the potential methods of gaming, inappropriate exception reporting is one of the easiest to detect, and PCTs can audit practices with suspicious figures.
Even if they were to set a more conservative threshold of 10 per cent exception reporting, the average PCT would only have two or three such practices on its patch to investigate. If exception reporting is retained, in addition to close scrutiny GPs may have to accept maximum thresholds of 100 per cent, as anything lower is difficult to justify when exception reporting is permitted and achievement for most indicators is universally high.
However, before informed decisions about the future of exception reporting can be made, more detail at individual practice level is required. The newly published data was a limited summary at the SHA level.
The relationships between exception reporting and practice and population characteristics, and the reasons why practices exclude patients, remain unknown.
This information would allow PCTs to profile local practices and detect low level gaming: what constitutes a high level of exception reporting for a single-handed practice serving an affluent population in a rural setting may differ from that for a large practice serving a deprived, urban population.
PCTs could also use this information more positively in partnership with GPs: to support best practice, rather than simply conduct anti-gaming witch hunts.
- Tim Doran is clinical research fellow at the National Primary Care Research and Development Centre and lead researcher on the QOF analysis project