1. Little PS, Gould C, Williamson I, Warner G, Gantley M, Kinmonth AL. Reattendance and complications in a randomised trial of prescribing strategies for sore throat: the medicalising effect of prescribing antibiotics. BMJ 1997; 315: 350-352[Abstract/Free Full Text]. (9 August.)
2. Prior A, Montgomery P, Mitchelmore I, Tabaqchali S. The microbiology and antibiotic treatment of peritonsillar abscesses. Clin Otolaryngol 1995; 20: 219-223[Medline].
3. Hibbert J. Acute infection of pharynx and tonsils. In: Stell PM, ed. Scott-Brown's otolaryngology. 5th ed. Vol 5. London: Butterworth, 1987:76-98.   

Authors' reply

EDITORSimo et al have misunderstood the criteria for including patients in our trial. Patients with quinsy were not eligible; the complications quoted relate to the follow up period.¹

How common are serious complications such as quinsy, and can they be prevented by prescribing antibiotics for patients who are not very ill? Assessment of how common cases of quinsy are must take account of the population denominator. Only one case of quinsy occurred in 434 patients who had no or delayed treatment. On the assumption that this was not a chance finding and that half of all cases of quinsy are preventable (generous assumptions), then roughly 868 patients would need treatment to prevent one case of quinsy. If we assume that 33%, 20%, or 10% of subsequent quinsy is preventable then 1302, 2170, or 4340 patients respectively would need to be treated to prevent one case. A "mega-trial" would be needed to estimate the true effect.

Computerised data from 16 000 patients in our two surgeries (an inner city surgery and a market town surgeryabout 7000-8000 presentations of sore throat²) document 23 patients with quinsy or pre-quinsy, of whom only nine presented with a prior sore throat (six received antibiotics, which did not prevent the quinsy). Thus most quinsy in otolaryngology departments probably presents de novo and is not "preventable." Simo et al list other complications, which are also rarefor example, a general practitioner has roughly a 1 in 5 lifetime chance of preventing a case of rheumatic fever or glomerulonephritis.²

These rare complications must be balanced against the complications of prescribing which do not present toand thus may be "invisible" tootolaryngology departments; these include rash, diarrhoea, recurrence, and antibiotic resistance (all common); anaphylaxis in 1-2 in 5000 cases; and death from anaphylaxis in 1 in 50 000 (similar to the likelihood of preventing rheumatic fever if antibiotics are given).² A major "side effect" is the medicalising effect of prescribing: one patient will return to the surgery for every nine patients treated, with a larger long term effect due to reinforcement.

General practitioners must balance the large, quantified medicalising effects of prescribing in the climate of increasing demand and antibiotic resistance with the poorly quantified possibility that a few patients may benefit from reduced complications if they do prescribe, taking into account the serious side effects of prescribing. Thus until better evidence is available for clinical targeting to minimise complications for patients who are not very ill, either not prescribing or delaying prescribing is likely to be the most effective and efficient management.