The Quality and Outcomes Framework: too early for a final verdict

An editorial in this Journal in 2002 suggested that the new GP contract could prove to be the requiem or renaissance for general practice.1 What has been learned 5 years on?

One obvious impact of the Quality and Outcomes Framework (QOF) has been the high quality scores achieved in the first year and the widely-publicised financial rewards. In retrospect, it is easy to see how GPs were able to score so well. Care was already improving rapidly in the years leading up to the new contract. So, for heart disease, for example, the percentage of patients with controlled blood pressure rose from 47% to 72% between 1998 and 2003, and the percentage of patients with cholesterol within recommended levels increased from 18 to 61 % in the same period.2 The roots for these improvements go back a decade or more. Audit was introduced as a compulsory part of the 1990 GP contract and seemed to have a modest impact at the time.3^,4 But what happened during that decade was that GPs gradually started using electronic records, they got used to comparing their care with others, and many GPs employed nurses to improve the care of chronic illness. So, when the QOF came along, much of the infrastructure for quality improvement was already in place, and GPs were able to respond rapidly to the new incentives.

Since the QOF was introduced, quality of care shows further improvement. For asthma and diabetes, care is now improving more rapidly than before the contract. For coronary heart disease, where care was already showing major change, the improvement has continued at the same rate.5 Care in relation to these three diseases has undergone definite if modest improvements over and above what was already being achieved. These trends should have some important impacts on health.6 Several commentators have doubted whether the improvements so far represent value for money in terms of £1 billion annual investment, but what has been achieved is a mechanism for an ongoing programme of quality improvement in new areas that is unique among national healthcare systems.

Mangin and Toop7 are unable to find evidence for many of the indicators in the QOF. This is hardly surprising. When a rigorous process was used to develop quality indicators for asthma, angina, and diabetes, only a quarter of indicators that GPs rated as ‘necessary to do and record’ were strongly evidence based.8 Quality indicators are always going to contain a large element of professional judgement, and that should be applauded, not derided. What is needed is a robust and transparent process for incorporating professional judgement in new indicators.

Any scheme which includes large financial losses and gains is potentially open to cheating. Cheating is hard to detect, but one aspect of the QOF that still concerns government is exception reporting. The rationale for exception reporting is that evidence-based guidelines were never intended to apply to every patient who sits down in front of his or her GP. Allowing the GP to say: ‘This indicator doesn't apply to my patient’, makes it easier to align managerial with professional incentives, and to avoid inappropriate distortions of care. So have GPs abused the ability to use exception reporting? On the whole, they have not. The median exception reporting rate was 6% in the first year of the contract,9 and 5.3% in the second year.10 One practice exception reported 86% of its patients in the first year, but this top figure for exception reporting has come down to 28% in the second year. Primary care trusts obviously have an inspection role for practices with high rates of exception reporting, but generally, there is little evidence of widespread abuse. Other forms of gaming are hard to detect. The suggestion that GPs recoded patients to diagnoses other than ‘coronary heart disease’ in the run up to the contract11 is cause for concern, although this could be legitimate cleaning of disease registers. However, it is a serious problem that the current payment system systematicaly penalises practices serving deprived populations with high morbidity.12 The payment formula needs to encourage casefinding in areas of high morbidity, not discourage it.

It is often suggested that incentives will widen health inequalities, because doctors will concentrate on patients who are easier to treat. When incentives were introduced for cervical cytology and immunisation in 1990, inequalities widened initially but over 6 or 7 years the gap narrowed so that there was an overall halving of inequalities between deprived and affluent areas.13^,14 With the QOF, things have been rather different. Several research groups have found that QOF scores are lower in deprived areas.15^–17 But the interpretations of this finding are very different. Our group's conclusion9 is that GPs in deprived areas achieved high scores without recourse to high rates of exception reporting, and the differences in scores between affluent and deprived areas are small and of relatively little clinical significance. This is a considerable achievement for practices in deprived areas.

Steele and colleagues report that care has not changed for conditions which were not included in the QOF,18 which was concordant with another study of care for un-incentivised conditions in the US.19 However, Steele and colleagues then go on to say that this means there is a risk that care for un-incentivised conditions will get worse. An alternative interpretation is that GPs have maintained standards for conditions not in the QOF despite all the time invested into meeting incentivised targets in the QOF. However, research evidence in this area is very sparse, and the issue is a crucial one for the future of general practice. Has the QOF turned GPs from health professionals interested in the patients in front of them to mere box tickers? Or will it increasingly do so in future? This is perhaps the most fundamental and insidious threat that the QOF presents.

In looking at the impact of the QOF on professional values, it is important to appreciate that there have been many other changes to primary care in the last 5 years. Young doctors now expect to work shorter hours, GPs no longer have 24-hour responsibility for their patients, and the public expects different things of health professionals. GPs and general practice have changed, and the differences are not all due to the QOF. In our research on the QOF, we found a wide range of views on the impact of the QOF on professionalism.20^,21 Many GPs are strong supporters of a system that they believe has helped them to deliver high quality care. Some believe that it has given them more time with patients, with more routine tasks delegated to nurses. Others believe that it has fundamentally removed holistic and caring aspects of the GP's role. Nurses appear to feel this change more acutely than GPs.21

What then is the future for the QOF? It is clearly here to stay. Many countries are seeking to emulate what the UK has done. In my view, there are three critical issues that need to be addressed as the QOF develops.

The first is whether the QOF is intended to resource and reward standard good practice, or whether it is a vehicle for changing practice. The original QOF indicators were largely based on existing national guidance (for example, National Service Frameworks). More recent developments have introduced indicators that are less familiar with the aim to change practice. These are two fundamentally different approaches, and their implications need to be thought through carefully. Past research suggests that external incentives are most likely to strengthen internal motivation where they support existing professional values, and may damage it when they don't.22^,23

The second issue is the need to minimise the distorting aspects of the QOF in relation to other aspects of care. At present, it is still too early to judge how important these are. GPs have devoted a lot of time to the QOF in its first 2 years, but much of that activity may now become routine. If it becomes clear that that QOF is having a damaging effect on other aspects of general practice care, then our professional negotiators should argue for a reduction in the proportion of GP income based on the QOF in future years.

Linked to this second point, more attention is needed on promoting the importance of inter-personal aspects of care. The motto of the Royal College of General Practitioners is Cum Scientia Caritas. This means that GPs need to combine scientific knowledge and skill with a caring approach to patients. For many doctors, caring is what the job is really all about, and consultations in which doctors feel they can make use of their relationship with the patient are the ones they find most satisfying.24

However, Denis Pereira Gray has suggested that the job of general practice has changed in recent years, and a reversal of roles has taken place between primary and secondary care, such that primary care is now the place where lives are saved.25 So, more than ever, the challenge for GPs is to combine high quality technical care with high quality interpersonal care. QOF has made an important start in supporting Scientia. We now need to think more about how to encourage Caritas. The link between the two is also crucial; that is, enabling patients to engage in meaningful decisions about their care, an area where the performance of British GPs is lower than GPs in other countries.26

In common with experience in other countries,27^–28 the use of financial incentives to improve care is not a panacea, but rather should be seen as an adjunct to other quality improvement initiatives. Future developments of the QOF should focus on areas where there is clear professional consensus that care needs to be improved; equaly important aspects of GPs' work and patients' perspectives also need to be considered.

• © British Journal of General Practice, 2007.