Identifying the sore
that needs antibiotics
GPs are often berated for over-prescribing antibiotics for sore throats Dr Nick Francis pinpoints how to target prescribing by confidently identifying cases of strep throat
Negotiations around antibiotic prescribing are not easy, but rationalising the use of
antibiotics is now seen as being of utmost
importance at both national and international levels. Antimicrobial resistance is no longer simply a hospital problem and there is increasing evidence that high levels of prescribing are associated with raised levels of resistance in both populations and individuals.
There are also very good practical reasons to consider investing in these consultations: rationalising antibiotic prescribing for sore throat has been shown to result in reductions in future consultations for the problem.
Being aware of the evidence for the effectiveness of antibiotics and developing the communication skills to explore patients' expectations and concerns can transform the consultation and may be a key component in our efforts to address the increasing problem of antimicrobial resistance.
Further, using clinical prediction rules and delayed prescribing strategies can also play an important role and near-patient tests, while not proven to be of help yet, may play a greater role in the future.
The consequences of the overuse of antibiotics are now well known. But what role do they have in the management of sore throat? And how effective are they?
The best answers come from a Cochrane review including 26 studies and more than 12,500 patients. This review showed
85 per cent of patients who do not take antibiotics will be symptom free one week after consulting.
Antibiotics do confer some symptomatic benefit, with the greatest seen three days after consulting. However, the number needed to treat (NNT) for one extra person to be symptom free at this point (for those who had no throat swab taken) was 14.5. This, of course, means that for every 14-15 people you treat with antibiotics only one of them gets better any quicker.
For those patients who had a swab taken and a culture positive for streptococcus, the NNT was 3.5. The review failed to find evidence of a protective effect against acute glomerulonephritis, but did show a reduction in the incidence of acute otitis media, rhinosinusitis, and quinsy in those treated with antibiotics.
The reviewers pointed out, however, that these complications are rare in Western society and therefore the NNT to prevent one complication is very high.
For example, nearly 200 people would have to be treated with antibiotics to prevent one case of otitis media. Likewise, there was a protective effect for rheumatic fever.
However, the only included trials that showed any cases of rheumatic fever were those reported before 1961. This complication has now become so rare in Western society that the NNT would be extremely high.
The optimal duration of antibiotic treatment for sore throat is not clear. A number of studies have shown an improvement in 'microbiological cure' with 10 days of treatment compared with shorter durations.
This has led to many sources recommending a 10-day course on the grounds that
microbiological cure is important to prevent relapse and the spread of the infection. But there is no evidence from these studies for differences in the resolution of symptoms.
Further, up to 40 per cent of individuals carry GABHS, but are asymptomatic. This may lead some patients with sore throat and a positive throat culture to be treated unnecessarily with antibiotics, since the true cause of symptoms may still be a viral infection.
A trial that compared treatment with penicillin V for seven days versus three days found that symptomatic cure rates were significantly higher for those taking antibiotic for seven days and on the basis of this evidence, antibiotic treatment is recommended for at least seven days. The Health Protection Agency recommends penicillin V 500mg BD-QDS (QDS for more severe infections) for seven to 10 days.
The Cochrane review makes it clear that antibiotics confer little benefit overall, but are more likely to be of benefit if targeted at those with a GABHS infection. So how do we best target those who are likely to have a streptococcal infection?
Aids to diagnosis
Diagnosing streptococcal infection on clinical grounds alone is not easy and the greater the clinical uncertainty, the greater the likelihood of unnecessary antibiotic use. Several researchers have developed clinical prediction rules that use a combination of symptoms and signs to try to increase the diagnostic accuracy.
The one that has been most validated for use in primary care was developed by Centor in 1981 and was modified in 2000 by adding age as a variable. This 'Modified Centor Score' is shown opposite. It has been found to have a sensitivity of 85 per cent and a specificity of 92.1 per cent for detecting GABHS in patients presenting with sore throat.
Two studies that assessed the use of this system in the community found using the score to inform prescribing decisions would have reduced antibiotic prescriptions for sore throat by approximately 50 per cent.
Other aids to diagnosing streptococcal sore throat include throat swabs and rapid antigen detection tests (RADTs). Throat swabs are used extensively in other countries, but the delay in obtaining results and the fact that up to 40 per cent of people are asymptomatic carriers limits their usefulness in primary care and they are not recommended for routine use.
RADTs have the potential to provide the clinician with a result within minutes but their widespread use has not been adopted in this country because of concerns over poor sensitivity. Newer RADTs have now been developed that are reported to have much higher sensitivities and specificities, and may now prove to be valuable tools for use in primary care.
A large-scale evaluation of these tests is being commissioned by the National Co-ordinating Centre for Health Technology Assessment and will hopefully provide us with a much clearer understanding of the role of these tests within a few years.
Nick Francis is a GP in the South Wales valleys and MRC research fellow in the department of general practice, University of Cardiff, with a special interest in common infections in primary care
Find the full version of this article in The Practitioner, free with your May 25 copy of Pulse
Modified Centor sore throat score
Score 1 point for each of the following:
Absence of cough
Tender anterior cervical adenopathy
Tonsillar swelling or exudates
Subtract 1 point if:
Total score Chance of streptococcal infection
0-1 2-6 per cent
2-3 10-28 per cent
=4 38-63 per cent