An acute trust has worked with a PCG to produce a joint formulary in under a year. Facilitator Heather Walker outlines the project

Most hospitals already have a formulary: expenditure on drugs is increasing and there is evidence to demonstrate that formularies can reduce this expenditure without affecting the quality of patient care.

Since the government first restricted GP prescribing with the 'black list' in 1984, and the Department of Health recommended the development of formularies at a local level in 1990, many GPs have also become accustomed to the idea of prescribing to a limited list.

3There are a few examples of joint formularies, but most of these have been adapted from the hospital formulary and then modified for GP use.

4Homerton Hospital trust, an acute 400-bed unit, and City and Hackney primary care group, which consists of 147 GPs within 53 practices, decided to work together to produce a joint formulary from scratch. It was my job to facilitate the process. The aims were to promote rational, cost-effective prescribing across the locality, to ease the transfer of patients between primary and secondary care and to strengthen collaboration between the two organisations - a key requirement of the NHS plan.

Letters were sent to all GPs and consultants in the locality explaining that work would be starting on a joint formulary. Forty-six GPs and 20 consultants replied. A questionnaire was enclosed in an attempt to establish what prescribers wanted, who wished to be involved and to what extent.

Over 90 per cent of respondents wanted to be involved in their specialty areas and a fifth wanted to be involved in all aspects of production. From this, a core of 25 GPs and nine hospital consultants along with six hospital and community pharmacists was identified, which has became the formulary committee.

The questionnaire also asked what format the formulary should take. Most people wanted a hard copy but many also wanted an electronic version.

Preferences in terms of style and content varied, but there was a strong message not to 'rewrite' the British National Formulary.

At this stage a bi-monthly newsletter was started to keep everyone informed of progress and to highlight news that would be appropriate to both primary and secondary care prescribers.

Once the formulary was complete, it was suggested that the newsletter also be printed in A5 size so that it can be filed in the formulary binder. It is anticipated that it will include notification of all updates and changes.

A deadline of 1 August for completion was agreed, to tie in with the arrival of the new junior doctors in the hospital. Although this meant a heavy work schedule and the possibility that the end result might not be perfect, the general consensus was that it would keep the work focused. Once a completed document was available, energies could then be devoted to involving others in the discussion and implementation phase.

Because of the tight deadline and the difficulty of organising meetings with doctors working different hours in different organisations, it was decided that the formulary committee would be a 'virtual committee'. No attempt was made to get all 34 members round a table. Instead, the work was sent via e-mail or fax. This worked extremely well for those people that had easy access to e-mail.

Those that used fax did not have the easy exchange of ideas that e-mail affords, but it was still better than trying to get the group together.

The only disadvantage of this was that members did not meet each other and develop a relationship.

A first draft of each chapter was produced after initial input from a consultant specialist and a GP with an interest in that therapeutic area. This was circulated to the formulary committee, which was given a deadline by which it had to respond. If comments gave rise to major changes, a second draft was sent to the committee. If a final agreement could still not be reached, it had been proposed that the respective chairs of the hospital drug and therapeutics committee and the PCG prescribing sub-committee would meet to make a decision. In practice, such meetings were rarely needed.

The format and style of the formulary were agreed after circulating the first chapter. It is a standard A5 size in a ring binder. At the beginning of each chapter is a reminder of local and national guidelines available. Drugs are listed in BNF order with the usual adult doses and the cost of 28 days' primary care supply. The cost of printing was£13,000 for 1,000 copies.

Although the costs are liable to change, they were included as a general guide because it has been shown that most GPs perceive cost to be an important factor in drug selection but their actual knowledge is often inaccurate.

5Key notes and safety warnings are included, but these were kept to a minimum. The general principles that were used in making decisions on drug inclusion were evidence of clinical efficacy and safety, cost, patient acceptability, consultant preference and current usage.

The main controversial areas concerned differing opinions of 'best evidence'. Sometimes the specialists could not agree whether individual drugs from within a class were superior. There was also some compromise on the amount of detail included. Generally, consultants wanted all the information their juniors would need, whereas GPs found that too specialised.

Implementation is a crucial part of the project. Evidence shows that dissemination by itself is unlikely to effect changes in behaviour.

Therefore a detailed programme of activities has been planned, including seminars, discussion groups and encouraging feedback.

5The formulary is an advised list of first or second-line drugs for use in routine situations.

Drugs not listed in the formulary are not banned. But junior doctors within the hospital will be expected to adhere to the list, whereas consultants can prescribe from outside the formulary provided the hospital's drug and therapeutics committee has approved the choice.

GPs may use part or all of the formulary as they see appropriate. Formulary choices are discussed at visits from the PCG pharmaceutical adviser.

The production of the formulary has been an educational experience for all those involved.Not least, it has enabled prescribers to gain some insight into the reasons behind therapeutic choices made in both primary and secondary care.

So far feedback has been very encouraging. The real test will be whether the formulary is used to change prescribing practice. Complete production took 10 months. This rapid turn-around reflects the success of the 'virtual formulary committee'.

Work on the project continues. An electronic version is already on the hospital intranet and we are hoping to have it on the PCG intranet in the next six months. Extra chapters for nurse prescribers are also planned.We hope we have created a sound basis for ongoing discussion and debate about drug choices for all prescribers in the area.

REFERENCES

1 Hill-Smith I. Sharing Resources to Create a District Drug Formulary: a countywide controlled trial. Br J Gen Pract 1996; 46: 271-275.

2 Collier J, Foster J.Management of a Restricted Drugs Policy in Hospital: the first five years' experience. Lancet 1985; i : 331-33.

3 Department of Health. Improving Prescribing - the implementation of the GP indicative prescribing scheme. Section 13.Department of Health, 1990.

4 Garvey Getal. Grampian Health Board's Joint Formulary. Br Med J 1990; 301: 851-852.

5 Ryan Metal. Do Physicians' Perceptions of Drug Costs Influence their Prescribing? Pharmacoeconomics 1996; 9 (4): 321-331.

6 NHS Centre for Reviews and Dissemination. Getting Evidence Into Practice. Effective Health Care 1999; 5: Number 1.