Back issue listing | Subject Index | Journal listing by subject | ImpAct Back issue listing

Promoting interest in evidence-based practice in primary care

Why was the initiative launched?
What was done?
Persuading parents
Did it work?
Were parents listening?
What was the effect on prescribing levels?
Can we build on our success?
Tips for success
For more information contact

Learning from an initiative at Manor View Practice in Bushey to reduce the use of antibiotics in children with ear infections

Why was the initiative launched?

In May 1997 a Cochrane review published in the BMJ questioned the use of antibiotics as the initial management of acute ear infections (otitis media) in children. At the time, Chris Cates, a partner in a training practice in Bushey was becoming involved in the work of the Cochrane Collaboration. The practice served about 11,000 patients with few from ethnic backgrounds. Information about prescribing in the practice suggested that about half of the antibiotics given to children were for acute otitis media.

What was done?

The six partners agreed to expnére ways to use the results of the Cochrane review : could the practice reduce their antibiotic prescribing level ? After some discussion they decided to continue to offer prescriptions to parents - but ask them to delay using it for a couple of days. They foresaw that refusal of a prescription at a first appointment might simply prompt more appointments. A deferred prescription might give parents confidence to wait and see if their child improved without the use of antibiotics.

They chose to tackle the initiative allowing each GP to adopt the new policy at their own pace. They decided not to write their new policy as a guideline or protocol: this would require agreement on a definition of otitis media, not felt to be feasible in a rapidly changing condition. There was also a feeling that the preparation of a guideline might simply encourage partners to avoid the diagnosis of otitis media and use a related condition as the argument for prescribing antibiotics. The discussion had prompted awareness of the evidence and the issue: this was argued to be sufficient.

Persuading parents

Chris Cates prepared a draft handout for parents to persuade them of the merit of the new approach. It explained the results of the research and the recommended treatment, ie the use of paracetamol to reduce the pain and antibiotics if the symptoms persist. The handout was modified following suggestions from partners and made available for the use in July 1997. GPs were given a supply of the handout and encouraged to use it. They were not then asked to make any other changes to their approach, and in particular not required to keep separate records of their actions.

The new policy was soon seen to be having an effect. Responses from parents were favourable. The change in policy was not causing any problems. The practice decided to look for ways to measure the impact of their efforts and find out what was happening to the deferred prescriptions. In October 1997, Roseanne Whitfield, a registrar attached the practice needed a project as part of her training programme. It was agreed that she should contact a sample of parents to check whether the prescription had been used. Also a local practice agreed to act as a control and allow information about their prescribing levels (PACT data) to be analysed. The practice is about the same size with a similar group of GPs and patients and, like Manor View, uses amoxycillin as the antibiotic of choice in children with otitis media. The control practice, although aware of the evidence, was not planning specific action to change their prescribing policy.

Did it work?

The effort involved had an encouraging impact on prescribing levels and has given the practice a real sense of achievement. Quality of care has improved and parents support the changes. Success with the antibiotic initiative was reflected in the award of Beacon status in 1999. More significantly the work has prompted important changes in the way the practice looks at the evidence for the care it provides.

Were parents listening?

All the partners were asked to keep records for two weeks of patients who consulted with otitis media and whether an antibiotic prescription, for deferred or immediate use, was given. The GPs saw 24 children over the two-week period. Patents of 19 of these could be contacted one week later. Seven parents advised to use the prescription immediately and all had done so. Two of the eight parents given a prescription but advised to hang on to it used it immediately: the other six waited and did not need to use it. Parents of four children were not given a prescription. No children returned for further treatment. Although small numbers were involved the results were encouraging. Parents seemed to be heeding the advice given.

What was the effect on prescribing levels?

Support from the health authority and the PPA was enlisted to help compare prescribing levels between the practice and the control practice. A search of practice records had shown that the majority of prescriptions for amoxycillin were for otitis media, the use of other antibiotics for otitis media balanced the use of amoxycillin for other conditions. Prescribing levels were compared for the twelve months before the use of the handout for parents was available and with the two following years.

The initial impact was marked. Over the first six months the total number of antibiotic prescriptions for children was about 20% lower than in the same six months the previous year. To help them see more clearly the impact of their work over a longer period and allow for seasonal variations in prescribing levels the practice calculated monthly and annual rates (using suitable statistical techniques) - see Table 1 and Figure 1. More detailed graphs to illustrate the rates of change are included on the ImpAct website. These show an encouraging downward trend. There was a big difference in the first year of the intervention but prescribing in the control practice has also reduced. Anecdotal evidence suggests a 'playgroup' influence; word seems to have been spread by parents about the merits of the new policy. Additionally the control practice has set out to see if they could also achieve a similar reduction when they saw the results of the new policy.

Figure 1:


Table 1: Prescriptions for amoxycillin suspensions
  Before the change in policy 1996/7 First year 1997/98 Second year 1998/99
Median number of prescriptions per month  
Manor View Practice 75 47 35
Control practice 72 66 39
Total number of prescriptions per year  
Manor View Practice 988 639 448
Control practice 991 835 501

Figure 2:


Figure 3:


Can we build on our success?

Progress with the initiative has encouraged the practice to look for other ways to improve the quality of their care and treatment of patients. Time is now set aside for a half-hour meeting every other week to address topical issues. Initially the meetings involved only the GPs in the practice but invitations are now being extended to nursing staff.

The agenda is drawn up and a range of issues addressed. Clinical issues include the treatment of sinusitis (based on a Cochrane review like the work on antibiotics), the use of post-coital contraception, the use of a new spirometer in the asthma clinic. Other issues include ways to improve appointment systems and the displays in the waiting room.

Partners and nursing staff adopt topics and agree to investigate the background and provide presentations to their colleagues. This task of preparation and presentation is proving to be a useful personal development opportunity for those who have not been used to such work. The meetings are popular and productive. After reviewing the evidence the use of the Yuzpe method of post-coital contraception has been abandoned in favour of Levonorgestrel alone.

Tips for success

√ Ensure that all clinicians are involved from the outset: don't leave out colleagues who can reinforce the messages to parents, such as health visitors
√ Avoid making meetings the key stepping stones to progress. Make full use of informal opportunities to talk, such as over coffee, and in the corridors
√ Use posters in surgeries and other leaflets to get the message over to parents.
√ Look for ways to use graphical illustrations in handouts for patients: don't rely only on words alone
√ Look for ways to demonstrate progress like quick local surveys of patients' reactions.
√ Keep the process simple and don't add to the burden of record keeping.

For more information contact

Dr Chris Cates
General practitioner
Manor View Practice
Bushey health centre
Hertfordshire WD2 2NN

Telephone 01923 225224
Fax 01923 213270


The handout for parents can be downloaded from Chris Cates' website and be adapted freely for your own use.

ImpAct bottom lines

Be openly incremental: make plans to build on successes and encourage further development


previous or next story in this issue