The UK quality and outcomes framework

In the linked study (doi:10.1136/bmj.a2030), Ashworth and colleagues report on the relation between the quality and outcomes framework and health inequalities in general practice in the United Kingdom by assessing the effects of social deprivation on levels of blood pressure monitoring and control.1 Tackling health inequalities has been a consistent part of the political rhetoric in the UK for more than a decade, with primary care seen as a key player in improving life expectancy in areas with the worst health record and highest deprivation.2

Perhaps the most substantive government intervention in primary care in recent years has been the renegotiation of the general practitioner contract in 2003, which included—at its heart—a system of financial incentives for delivering quality care. This pay for performance scheme—the quality and outcomes framework—now links achievement on 129 indicators covering clinical domains; organisational domains; and additional services domains, such as contraception and patients’ experience, to the practice’s income. The framework was originally designed to improve health outcomes and not as a tool to tackle aspects of health inequality.3 In the first year of its implementation (2004-5), practices serving deprived populations achieved slightly lower quality scores and therefore received less financial remuneration than those in more affluent areas.4

Ashworth and colleagues provide a longer term perspective on the relation between the quality and outcomes framework and health inequalities.1 They used data collected during the first three years of the framework’s implementation, from more than 97% of practices in England, and they found that the small differences between values of blood pressure monitoring and control in the least and most deprived areas in 2005 were dwarfed by the overall improvement in values over the time period. Crucially, they also found that the achievement gap between practices in the least and most deprived areas had almost disappeared.

This important study supports the findings of a methodologically similar paper published recently in the Lancet,5 which examined the relation between socioeconomic inequalities and overall achievement in 48 of the clinical indicators of the quality and outcomes framework during the same three year time period. Doran and colleagues also found that median achievement scores increased across the board, and that the gap in median achievement narrowed from 4.0% to 0.8% between practices in the most deprived and least deprived areas. Interestingly, they also found that although performance in year 1 was associated with area deprivation, the increase in achievement was inversely associated with the practice’s performance in previous years and was not associated with deprivation.

The evidence therefore suggests that the quality and outcomes framework contributed to an improvement in the process of care and intermediate health outcomes for patients with a range of clinical conditions, although the size of the contribution is unclear.6 Low scoring practices in deprived areas also seem just as able to improve the quality of their care (as measured by the framework) as low scoring practices in more affluent areas. This offers the tantalising prospect that the quality and outcomes framework is a truly equitable public health intervention.7

However, several important caveats exist. For example, inequalities in health care might already have increased (as predicted by the inverse equity hypothesis8) in response to other quality improvement initiatives that pre-dated the introduction of the quality and outcomes framework. We know that the quality of care was improving in several clinical conditions before 2004, and this could have created a “ceiling effect” in terms of achievement for practices in more affluent areas. Inequalities could also have widened in unincentivised areas of health care, although two recently published studies in the United States and UK suggest that practices have delivered the same quality of care for conditions not included within a pay for performance scheme.9 10

This emerging evidence has several implications for policy makers and practitioners. In future, perhaps pay for performance schemes should be actively designed with health inequalities in mind.. For example, Massachusetts is considering including pay for performance targets for reducing ethnic inequalities in healthcare provision in their Medicaid programme.11 Perhaps future schemes should more directly reward absolute quality scores and improvement over time.12 Evaluations could also plan at the outset to consider and then monitor the consequences of pay for performance on health inequalities.

On balance, however, the message of Ashworth and colleagues’ paper is a hopeful one. High blood pressure is the most important risk factor worldwide for developing cardiovascular disease, a condition that contributes greatly to the gap in life expectancy between deprived and affluent areas. The problem of reducing blood pressure is now being tackled more effectively in practices across the land. Perhaps the greatest contribution that the quality and outcomes framework has made to changing practice will therefore be the largely unintended consequence of generating more equitable health care.