Clinical assessment of pharyngitis in general practice.

The present study investigates the feasibility of the clinical differentiation between patients with beta-haemolytic streptococcal pharyngitis from those with pharyngitis caused by other agents, based on the patients' symptoms, symptom duration, and the clinical findings. Twenty-four general practitioners recruited 225 patients for the study. Fifty-six patients had positive and 169 patients negative group A beta-haemolytic streptococcal throat cultures. Twenty-two patients in the streptococcal group and 76 patients in the non-streptococcal group were initially correctly diagnosed based on an overall clinical assessment (sensitivity 0.39, specificity 0.55 and accuracy 0.51). This is as accurate as "flipping a coin". Similar figures were found with regard to the individual symptoms (accuracy 0.38-0.68) and clinical findings (accuracy 0.36-0.65). Taking the duration of symptoms into account and applying discriminant function analysis did not significantly improve these figures. In order to obtain a sufficiently accurate diagnosis in general practice, the use of the new rapid agglutination test for streptococcal identification is recommended.