The quest to cut the NHS drugs bill has met with spectacular success in the elderly care department of one acute hospital. David Griffith and Mark Robinson explain how

The NHS now spends over£6bn a year on drugs - 14 per cent of its costs. Various options, such as limited list formularies, have been used to try to achieve more rational prescribing, but ample scope for improvement remains. In 1992 we instituted a system of regular audit of prescribing in our elderly care department which we believe could helpfully be implemented elsewhere.

Cutting costs

We have identified the drugs that are the cheapest of a group of preparations with similar properties and efficacy, and encouraged switching to them. These measures have contributed to cost containment (see table). The audit scheme was introduced in 1993, and in 1997 our expenditure was virtually the same. Pharmacy inflation of 11.2 per cent during this period means that the predicted cost for 1997 based on 1993 levels would be£105,500, so an effective saving of£11,500 has been achieved for this year alone. If allowance were made for a 26 per cent increase in admissions over this period, it is likely that this figure would be even greater, although more complex factors come into play in assessing this. Multiply these savings across other departments and the reduction in expenditure can be significant.

We have addressed several quality issues in prescribing.

We have reduced the use of sleeping tablets: addiction to them may be a problem at various ages, and in older people it is particularly dangerous because it increases the likelihood of falls. Following the introduction of audit, monthly issues for the most commonly used sleeping tablet fell from 2,392 to 734, and discharge prescriptions fell from an annual rate of 300 to 156.

Reducing drug use

British Thoracic Society guidelines state that combinations of two drugs should only be used in treating certain respiratory problems if single drugs have been tried but failed to give adequate relief. We have reduced the use of a second drug from 4,470 doses to 1,980 a year over a four-year period.

Low-dose aspirin is now recognised as beneficial for a number of vascular problems and we try to ensure its use whenever indicated. Overall, discharge prescriptions per year have increased from 295 to 696. A recent retrospective survey of 200 patients showed that virtually all patients in whom aspirin was indicated had been started on this drug.

The use of multiple drugs is a constant threat and problem in older people. We have tried to restrict discharge drugs to those considered essential or highly desirable. As more and well-proven treatments have become available for some conditions, there are pressures to increase the numbers of drugs used. Despite this, we have achieved a modest reduction from 4.03 to 3.28 discharge items per person in a three-year period. The cost per discharge for drugs has come down from£7.69 to£5.41 per patient.

A few details on our particular local characteristics may be helpful before describing the audit process. Within a 690-bed district general hospital, the elderly care department medical staff - four consultants, two specialist registrars and eight senior house officers - look after 2,700 admissions a year. Most patients are looked after on wards where all patients are under the care of a single consultant team. This is helpful in monitoring the prescribing practices of individual consultant teams.

Frequent departmental meetings take place, so monthly meetings with the principal pharmacist fit easily into this framework. There is an openness about scrutiny of colleagues' working practices, a willingness to accept criticism or suggestions, and a will to achieve better outcomes.

The relatively small size of this group is probably particularly helpful in creating an appropriate atmosphere. In the meetings, discussions take place about new evidence on drug treatment, literature reviews and good prescribing practice. Departmental policies are drawn up and agreed. In some instances, these are formally set down and incorporated into induction notes for all doctors joining the department, but in other instances they are built much more informally into routine practice through ward rounds. These changes are monitored and reinforced through data collection by pharmacy with presentation of the results at the regular meetings.

The pharmacy uses a JAC computer system with a standard financial and clinical reporting suite. It provides information on the cost, volume and range of dispensed medicines.

The system described here relies on one 45-minute departmental meeting per month, plus the time taken by the pharmacist to develop a topic and marshal the figures - about four hours per month. We have found that frequent feedback has worked. And this system of audit has produced continuing effects in prescribing practice over several years.

Our approach could make a useful contribution to improving prescribing practice as it is simple, does not involve expensive or complex audit packages, relies on information that is likely to be widely available from pharmacies, and is effective.