GPs' terms of service and fees are set out in the 'Red Book' first published in 1966 as a slim booklet. It now covers 82 sections and about 70 separate fee rates. Angela McCullagh and others say it is time to get rid of it

The government has announced its intention to tackle unacceptable variations in performance within the NHS. The national framework for assessing performance has put forward an initial set of high-level indicators intended to provide an overview of health authority performance; it ushers in a set of effectiveness indicators for primary care.1 At practice level, the degree of variation in the processes and outputs of primary medical care is substantial. These can cover well-publicised variations in consultation and prescribing rates as well as rates of referral to secondary care.

Attempts have been made to examine correlations with characteristics of practice populations (such as the age structure); characteristics of the practice (for example, single-handed or partnerships); and characteristics of individual general practitioners (for example, date of qualification). Examples include studies on particular conditions, treatments or client groups - for example the prescribing of non-steroidal, anti-inflammatory drugs, number of emergency asthma admissions or numbers of GP consultations with unwell toddlers.2,3,4

Little attention seems to have been paid to variations in what might be regarded as the controllable inputs into primary care - the level of NHS resources allocated to individual practices. GPs' responsibilities for their services to the HA are controlled by their national contract which lays down the terms of service, and the statement of fees and allowances (the Red Book) which controls how the GP claims reimbursement. The Red Book was first published in 1966 as a slim booklet, became a loose-leaf, ring-bound version in 1972, and is now available on floppy disk. It has become increasingly lengthy and complex and now covers 82 sections and about 70 separate fee rates.

Most of the current debate on resource allocation for primary care has centred on the arrangements for funding primary care groups for commissioning secondary and community care. This has obscured the equally difficult issues surrounding how primary care provision will be funded under the new arrangements.

By way of illustration, we conducted a brief exercise to examine the financial inputs into the 19 general practices on the Isle of Wight. Of these 15 were fundholding and four were not. Although the smallest HA in the country, the organisation of primary care is not atypical of other 'mature' areas. For the purposes of this exercise, the 'inputs' we defined excluded prescribing allocations but included non-cash limited remuneration and reimbursements including rent and rates, and cash-limited payments, including allowable practice staff costs. The results show that within the district, the total annual cost of providing GP-based primary care services varies from 51 to 76 per head of population. Weighting the practice populations for age and sex according to the national guidance for allocating prescribing and secondary care budgets does little to reduce the range.5

At present, there is no evidence to support the contention that increased financial inputs result in improved outcomes for patients. We found no correlation between financial inputs and a range of specified outputs which might relate to improved patient outcomes. These outputs included inadequate cervical smear rates; indicators of quality prescribing (the ratio of steroids to inhaled bronchodilators and the ratio of antidepressants to benzodiazepines) and referral rates to secondary care.

One possible justification for such variation in financial inputs could be that it reflects need, as measured by the levels of deprivation in practice populations. Examining a larger dataset, consisting of the 80 general practices in Portsmouth and south east Hampshire, we looked for evidence of a relationship between payments to practices for specified items of service (night visits, maternity, contraceptive payments, immunisation, minor surgery or child health surveillance) and levels of deprivation in the practice population as measured by the Under Privileged Area 8 ('Jarman') score. Again there was no correlation between any of these payments and the level of deprivation. Need, then, as proxied (??) by deprivation, does not explain the variations in the financial inputs we examined.

Our small studies need to be compared with results from other districts, but the variation in financial inputs without evidence of corresponding measurable differences in outputs will raise many questions. In the new NHS, it is difficult to justify a 50 per cent variation in the per capita cost of providing primary care when there is no clear evidence that it is targeted at those most in need or that it results in benefit to patients. It is another 'unacceptable variation' which needs to be tackled by policy makers. This is now particularly pertinent, when unified budgets are to be allocated to PCGs, supposedly on the basis of need. The current system of funding primary care, based on the arcane rules and regulations of the statement of fees and allowances, is neither equitable nor responsive to patients' needs. The attention focused on secondary care commissioning and provision must now also be directed towards these unacceptable variations in primary care. The Red Book should not survive in a modern and dependable new NHS.