Intended for healthcare professionals

Editorials

Problems with performance related pay in primary care

BMJ 2007; 335 doi: https://doi.org/10.1136/bmj.39329.384329.BE (Published 13 September 2007) Cite this as: BMJ 2007;335:523
  1. David S Wald, consultant cardiologist and senior lecturer
  1. Wolfson Institute of Preventive Medicine, Barts and the London School of Medicine and Dentistry, London EC1M 6BQ
  1. d.s.wald{at}qmul.ac.uk

    Payments should be simplified and based more on disease treatment and prevention and less on risk factor measurement

    General practice in the United Kingdom has the largest healthcare pay for performance programme in the world—the quality and outcomes framework (QOF).1 By earning the maximum number of performance indicator “points,” an average sized practice can earn about £125 000 (€185 000; $252 000) in addition to its usual sources of income. In this week's BMJ, Guthrie and colleagues discuss the effectiveness of the system in relation to the management of cardiovascular disease. They show how general practices can earn many points and extra payments without this necessarily indicating a reduction in the risk of cardiovascular disease.2

    For example, a practice could receive nine points (each worth about £125) for generating a list of patients with hypertension. The completeness and accuracy of this list might be subject to external audit by the Primary Care Organisation. An extra 30 points would be earned if 90% or more of such patients have a record of risk factors (blood pressure and smoking history) in their notes, and 56 more points would be earned if 70% or more of such patients have a record of blood pressure lowered to below specified target values (150/90 mm Hg). Overall, 15% of payments, worth an estimated £200m across the approximate 11 000 general practices in the UK,3 arise from measuring cardiovascular risk factors (such as blood pressure and serum cholesterol) and recording whether they are below specified values.

    Whether or not doctors should receive financial incentives for providing medical care is debateable.4 5 Should police officers be paid extra for catching criminals and should firemen be paid incentives for putting out fires? A balance is needed between encouraging doctors to exercise independent professional judgment and paying them for carrying out specific tasks. The balance has moved too far towards payment per task done, and this is de-professionalising medical practice. The treatment and prevention of cardiovascular disease is becoming a series of isolated tasks predicated on financial rather than clinical value. Linking each task to the receipt of money means that money rather than medical judgment drives practice.

    In addition, the need to count cases and fill in forms requires extra resources and increases bureaucracy. Baroness Deech, head of the Office of the Independent Adjudicator, said in relation to the bureaucracy of postgraduate education, “We live in an age of over-regulation. I do think universities are over-regulated.”6 The same criticism can be applied to the National Health Service.

    A further problem with the QOF relating to cardiovascular disease is that many of the measurements documented are not worth documenting. If doctors are to be paid for performance it should be for treating and preventing disease. In vaccination, payments increase with numbers of children vaccinated, as all children are susceptible to infections. The same principle applies to cardiovascular disease—everyone is susceptible. Identifying people on the basis of a high risk factor cut-off value is inappropriate because relatively few events occur in people with high risk factor levels. Most events occur in the majority of people whose risk factor values are closer to the average.

    Blood pressure and serum cholesterol measurements are commonly used in screening because these important causes of coronary heart disease and stroke are thought to be useful for predicting who will and will not develop such an event. However, with certain exceptions (such as familial hypercholesterolaemia7), this is not the case. Aetiologically important risk factors are rarely useful as screening tests.8 It is often assumed that combining information on several cardiovascular risk factors will overcome the problem that individually they are poor predictors, but such an approach is only a little more precise than simply basing a person's risk estimate on age alone, and is not worth the extra cost and complexity.9 Most heart attacks and strokes (more than 90%) occur in people over the age of 55, which is why 55 has been proposed as a reasonable age above which to prescribe drugs to reduce cardiovascular risk.

    The QOF, now in its third year, has been useful in drawing attention to the importance of the treatment and prevention of cardiovascular disease, but not how best to do so. The QOF expert panel, assembled by the BMA and the NHS Employers is currently reviewing the QOF programme. This provides an opportunity to improve and simplify the system.

    Guthrie and colleagues argue for increased incorporation of treatment information into outcome indicators.2 This makes sense, because it is the treatment of risk factors that reduces risk, not their measurement. Performance indicators should not be based on the measurement of risk factor levels, but on the proportion of people with existing vascular disease or diabetes, or those above a given age who receive effective preventive treatment, in addition to encouraging sensible dietary and lifestyle measures (such as smoking cessation).

    The resources currently used to fund the management of risk factors and the QOF payments that follow them could be redirected into paying for the drugs used. The financial element would then be directly linked to treatment and prevention rather than the process. Much unnecessary medical activity and cost could be avoided—£200m from the existing cardiovascular disease specific QOF payments alone. Further savings would arise from better use of time in general practice, avoidance of risk factor measurement, and reduced administrative costs. Incentive payments would be better used sparingly to encourage selected effective interventions that need specialist examinations, such as screening for diabetic retinopathy in people with diabetes.1

    Such a revised QOF system would be simpler and would release valuable general practitioner time and resources. Greater importance would be attached to medical judgment, rather than to robotic tasks. The QOF expert panel is expected to deliver its recommended changes this autumn. Hopefully, it will have time to reflect on these matters and advocate a much simpler strategy for treating and preventing cardiovascular disease—one that is linked to more focused incentive payments, while protecting the independent professional status of doctors in the UK.

    Footnotes

    • ARTICLE
    • Competing interests: None declared.

    • Provenance and peer review: Commissioned; not externally peer reviewed.

    References