Abstract
The Quality and Outcomes Framework (QOF), introduced as part of the General Practitioner contract in 2004, has been described as the boldest shift towards pay for performance in any health care system. 1 Around one quarter of general practice income is now derived through the achievement of quality targets in managing chronic diseases – such as diabetes and coronary heart disease – and risk factors – such as hypertension – through QOF. Early evidence suggests that this policy has resulted in a step-change in quality for some conditions such as asthma and diabetes but not for coronary heart disease. 2
Whilst the introduction of QOF has been associated with better quality care, it is important to assess whether these improvements have been equally distributed across population groups. A recent review 3 of mainly US studies has suggested that pay for performance programmes may worsen health care disparities if they encourage providers to ‘cherry pick’ healthier patients, exclude those not achieving targets from public reporting mechanisms or otherwise increase the resource gap between high and low performing health care providers. However, the review 3 found that there was almost no data on the effects of such programmes on disparities in healthcare delivery between racial or ethnic groups.
What has been the impact of QOF on health care disparities in the UK? Preliminary studies have generally found that practices situated in less deprived areas achieve higher QOF scores, 4, 5, 6, 7, 8 although variations in quality have tended to be fairly small. Furthermore, differences in attainment of QOF scores between affluent and poorer areas were reduced in the second year of the contract. 9 However, these studies are limited in that they did not use patient-level data and did not take into account ‘exception reporting’, whereby practices can exclude patients from performance reporting for reasons such as drug intolerances or patient refusal. At least one paper has found that there were higher ‘exception reporting’ rates in practices with higher deprivation levels. 10
Studies which have used patient-level information suggest that a number of disparities present prior to the introduction of QOF may have persisted. For example, lower prescribing of statins in black African relative to white British patients and lower prescribing of insulin in black African and south Asian patients were found to persist after the introduction of QOF in the Wandsworth Prospective Diabetes Study. 11 The study also found that whilst achievement of process of care measures was broadly equitable between ethnic groups, 12, 13 worse blood pressure and blood glucose control had persisted in black Caribbean patients. Similarly, other published data suggest that age and gender group differences in cholesterol control may have persisted in patients with coronary heart disease. 14 Longer term studies are required to determine whether these disparities persist or are attenuated with time.
A number of modifications to QOF may be required to produce more equitable management of incentivized conditions. First, QOF could be adjusted to provide better rewards to those practices serving deprived and ethnically diverse populations which tend to have a high prevalence of conditions such as diabetes and coronary heart disease. The current ‘square root transformation’ of prevalence has been criticized for providing insufficient rewards for practices with a high disease caseload. Second, quality indicators could be developed and expanded in clinical areas where suboptimal and inequitable care has been identified. For example, there is a case for further shifting the emphasis of quality assessment from process of care measures, where achievement is high and generally equitable, to quality indicators for prescribing and intermediate outcomes, where disparities have persisted. Third, disparities in care are more likely to be attenuated where thresholds for quality indicators are set high, hence these should continue to be increased over time. This is because low thresholds permit practices to earn maximum income without achieving adequate control of intermediate clinical outcomes in a large minority of their patients. For example, to achieve maximum QOF points for glycaemic control in diabetes, practices must ensure that 50% of their patients achieve HbA1c ≤7.5%.
In conclusion, QOF seems to have led to a continuing improvement in the quality of chronic disease management in the UK's NHS. However, some socioeconomic and ethnic disparities in quality of care remain that may require additional measures to address.
Footnotes
DECLARATIONS
Footnotes
Acknowledgements
None