Acute sore throat—diagnosis and treatment in general practice

Abstract

This study tested how general practitioners diagnose streptococcal infection on clinical grounds alone, in patients who presented with sore throats.

Four hundred and fifty-two patients were admitted to the study. A clinical diagnosis, prognosis and follow-up was completed in each case and the clinical assessment was checked by throat swabbing at first contact and a week later.

The doctors were inaccurate in predicting streptococcal infection, but better than might be expected if prediction were a matter of pure guesswork. Colds and influenza implied negative prediction, tonsillitis a positive prediction, and pharyngitis was doubtful.

In this series negative prediction for pharyngitis was 85·2 per cent and positive prediction 31·5 per cent accurate. The equivalent figures for tonsillitis were 61·5 per cent and 38·9 per cent respectively. There was a general tendency to overpredict streptococcal infection which was most marked in acute follicular tonsillitis, but this led to few false negatives. The tendency to overpredict streptococci was most marked when the patient was an adolescent female.

There were differences between the urban and rural patterns. During the same period, influenza (and similar illnesses) was recorded less often in the country, whereas urban practitioners were more likely to predict streptococcal infection. Rural practitioners were more accurate in prediction because they were less prone to implicate streptococcal infection than their urban colleagues; there was a higher proportion of cases with proven streptococcal infection in the town and there is a disproportionately high number of adolescent females among the urban patients.