General Practice

Controversies in primary care Setting prescribing budgets in general practiceCapitation based prescribing budgets will not work

BMJ 1998; 316 doi: (Published 07 March 1998) Cite this as: BMJ 1998;316:748

Controversies in primary care Setting prescribing budgets in general practice

General practice prescribing costs have risen rapidly in recent years, and there are wide variations between practices in rates and costs of prescribing. Setting general practice prescribing budgets with a capitation based formula seems to offer a solution to these problems. However, capitation based formulas may unfairly penalise legitimate variations and increases in prescribing costs. We therefore asked Azeem Majeed and Trisha Greenhalgh to give their views on the subject.

Capitation based prescribing budgets will not work

  1. Azeem Majeed (a.majeed{at}, senior lecturer in general practice,
  2. Stephen Head, medical adviser
  1. a Division of General Practice and Primary Care, St George's Hospital Medical School, London SW17 0RE
  2. b North Derbyshire Health Authority, Chesterfield S41 7PF
  1. Correspondence to: Dr Majeed

    Many health authorities are considering introducing capitation based prescribing budgets for their general practices.1 There are two important factors driving this process. The first is cost containment. Drugs prescribed by general practitioners now account for 11% of all NHS spending (see table). Furthermore, general practice prescribing costs have been rising more quickly than both the average rate of inflation and the total NHS budget.2 Many NHS managers and treasury officials are unhappy with this rapid rate of increase and see considerable scope for savings in general practitioners' prescribing costs. For example, the Audit Commission estimated that prescribing costs could be reduced by about £425m if all general practitioners prescribed like the doctors in 50 general practices which the commission identified as being “good” prescribers.3

    Table 1

    Cost of drugs prescribed by general practitioners in United Kingdom from 1985 to 1995

    View this table:

    The second factor behind the increasing interest in capitation based budgets is the belief that such budgets will help to ensure that resources are allocated more fairly among general practices. There are wide variations in prescribing costs between general practices, and it is not clear whether these variations are clinically justified. To many people, these variations suggest that the prescribing of general practitioners is either inefficient or inappropriate.

    Capitation based budgets seem to offer a solution to tackling the dual problems of unacceptable variations in prescribing costs and increasing drug costs in general practice. It is assumed that capitation based budgets will encourage general practitioners (especially those with high prescribing costs) to examine their prescribing more critically, resulting in more cost effective and appropriate prescribing. Even where the introduction of capitation based budgets is not being considered, health authorities have been advised by the NHS Executive to consider giving general practices with above average costs a smaller increase in their prescribing budget than practices with below average costs.4 The implicit assumption is that, over a number of years, practices will move towards the average and that the variation in prescribing costs between practices will be reduced.

    Are capitation based budgets the best method of allocating general practice prescribing budgets and will such budgets be fairer than existing budgets, which are usually based on historical prescribing patterns?

    What formula should be used

    The first problem that proponents of capitation based budgets are faced with is what formula to use to allocate budgets. The NHS Executive does have a capitation based formula that it uses to help allocate prescribing budgets to health authorities. Health authorities could use the same formula (which contains weightings for age, sex, and chronic illness) to help set the budgets of their own general practices. However, there is no direct measure of chronic illness available at general practice level, only proxy measures derived from census data, which are of uncertain accuracy. 5 6 Health authorities could construct their own formulas using locally available data, assuming that they can competently use statistical techniques such as multiple linear regression and multilevel modelling. Unfortunately, when such methods have been used to examine the variation in prescribing costs between general practices only about 30-40% of this variation has been explained.79 What is not currently known is the cause of the remaining 60-70% of the variation in prescribing costs.

    If this unexplained variation is caused mainly by idiosyncratic prescribing by general practitioners, then it would be reasonable for health authorities to introduce capitation based budgets gradually and to start moving practices with high prescribing costs towards the average for the health authority. However, it is also possible that the unexplained variation in prescribing costs is a result of differences in the clinical characteristics of practice populations or because some general practices are better at identifying and treating groups of patients who need long term medication, such as those with asthma or ischaemic heart disease. If this is the case then using capitation based budgets may lead to unfair reductions in the budgets of some practices and excessively large increases in the budgets of others. Clearly, large and rapid changes in prescribing budgets would be disruptive for general practices and would be difficult to justify while the reason for the large variations in prescribing costs remains unknown.

    High and low cost prescribers

    An examination of general practices with low and high prescribing costs reveals some distinct patterns of prescribing (see box). There are low cost prescribers whose costs are low because of poor quality prescribing, usually due to inadequate identification and management of patients with chronic diseases. Giving such practices larger prescribing budgets will not correct their poor clinical practice and nor will it address the needs of their patients. Other low cost practices seem to be prescribing very effectively. They have a high rate of use of generic drugs, a low rate of use of drugs of limited therapeutic value, and make little use of new, more expensive drugs when there is an older and cheaper drug of similar effectiveness available.

    A tale of two practices

    Practice A

    This is a three doctor practice with a list size of 4400 patients in an inner city area. The practice's age weighted prescribing costs are currently 15% below the health authority average. The practice runs disease management clinics for conditions such as asthma, hypertension, ischaemic heart disease, and diabetes. The practice also takes on the prescribing of expensive drugs for conditions such as infertility and hormonal disorders. The practice has an above average rate of generic prescribing and a low rate of prescribing drugs of limited therapeutic value.

    A recent audit showed that the practice had no patients admitted for the emergency treatment of asthma during the past year. With a capitation based formula, the practice would receive a substantially larger prescribing budget. However, as the practice seems to meet the needs of its population with its current budget, it is not clear what it would do with these extra funds.

    Practice B

    This is a nine doctor training practice with 15 200 patients. Although situated in a relatively affluent health authority, the practice is located in one of the area's pockets of deprivation. The practice's age weighted prescribing costs are currently 12% above the health authority average. The practice provides high quality care and runs chronic disease management clinics, provides extra services such as anticoagulant monitoring and regularly audits its prescribing. The generic prescribing rate is above average. Despite this, the practice's prescribing costs remain high, and it has great difficulty in staying within its prescribing budget. With a capitation based formula, the practice would lose a substantial part of its current prescribing budget, and this would affect the practice's ability to meet the needs of its patients.

    When high cost prescribers are examined two main groups also appear. There are some high cost prescribers whose prescribing is of poor quality. They make little use of generic drugs and extensively use expensive drugs and drugs of limited therapeutic value. Other high cost prescribers seem to be practising high quality medicine. These practices usually run specialist disease management clinics, often serve populations with a high need for care, and have in place effective mechanisms for identifying and treating patients with chronic diseases. Because they offer a high quality service, they often attract patients with complex health needs, and this further increases their prescribing costs.

    A capitation based formula classifies practices only as low cost or high cost prescribers and tells us nothing about their quality of prescribing. This information can come only from a detailed analysis of practices' prescribing data combined with information collected directly from each practice on the use of formularies and practice guidelines and on local factors that increase or decrease the demand for drugs.

    Implications for general practitioners

    If health authorities are considering the introduction of capitation based budgets then general practitioners will need to prepare for this. At least one general practitioner in each practice should have a sound grasp of the practice's prescribing data. The practice should also collate information on factors that could increase the demand for drugs—such as the transfer of care from hospitals to general practice, a high prevalence of chronic disease in the practice population, and patients living in nursing or residential homes.

    Improving how drug budgets are allocated

    There are several actions that health authorities could take to improve how they allocate prescribing budgets (see box overleaf). For example, they could exclude patients who need high cost drugs, such as growth hormone and cyclosporin, before calculating prescribing budgets. Recent developments in the information supplied by the Prescription Prescribing Authority to health authorities (PACT data) make it fairly straightforward to separate the costs of these drugs from other drug costs.10

    Setting accurate drug budgets and improving the quality of prescribing

    • Remove the costs of expensive drugs (usually defined as drugs that cost more than £2000 a year) from practices' drug budgets

    • Weight appropriately for age and sex—the best current method of weighting is to use ASTRO-PUs (age, sex, and temporary resident adjusted prescribing units)

    • Discuss with practices special factors that could increase prescribing costs, such as patients in nursing and residential homes

    • Discuss with practices how they manage patients with chronic diseases such as ischaemic heart disease and asthma

    • Examine practices' prescribing, analysis, and cost (PACT) data to identify areas in which there is prescribing of ineffective drugs or in which costs could be saved through therapeutic substitution or increased generic prescribing

    • Discuss with practices whose prescribing is of poor quality why this is and how the quality of their prescribing can be improved

    • Provide regular feedback of prescribing data and reinforcement of good practice

    Unfortunately, it is more difficult to identify high cost patients (usually those with chronic diseases who are taking many different drugs) by means of routinely available information. Information on these patients can only be obtained directly from each general practice in an authority. Hence, health authorities need to start systematically collecting information from practices about patients with chronic conditions such as asthma or ischaemic heart disease or who are living in nursing homes. Health authorities will also need to know about the quality of care provided by general practitioners, particularly whether they have prescribing policies in place and how well they identify and manage patients with chronic diseases.

    Improving the quality of prescribing

    Health authorities need to develop better methods of monitoring and analysing the prescribing carried out by their general practices, and of feeding back the results of their analyses to general practices. Many health authorities are doing this by developing prescribing indicators for feedback to general practices.1113 All health authorities now have access to detailed information on the rates and costs of prescribing by their general practices. This information can be used to identify areas in which financial savings could be made or in which the quality of prescribing could be improved.

    Prescribing costs also need to be considered together with other NHS costs. For example, general practitioners could argue that, by increasing the prescribing of drugs such as cholesterol lowering drugs for people with ischaemic heart disease and of treatments for asthma, they will improve the health of these patients and reduce the likelihood of them suffering complications. This in turn will improve these patients' quality of life and reduce demand on hospital services. Even when problems are identified with general practitioners' prescribing, changes in prescribing practice will not be easy to achieve and will require various different educational methods and continual reinforcement of good practice. 14 15


    Capitation based formulas are currently very crude tools for determining general practice prescribing budgets and should be used only as a guide to setting prescribing budgets and not as their ultimate determinant. If health authorities apply capitation based budgets inflexibly this may lead to practices becoming reluctant to register patients with high prescribing costs.16 Hence, health authority pharmaceutical and medical advisers will still have to use their knowledge of local factors and their judgment when setting prescribing budgets. There are many problems with the methods currently used to allocate prescribing budgets, and these methods do need to be improved. General practitioners also need to ensure that their prescribing is appropriate and cost effective. Unfortunately, there are no easy solutions to these problems, only a lot of hard work for both health authorities and general practitioners.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    View Abstract