One hundred ninety-six individuals, 86 with clinically overt pharyngotonsillitis and 110 of their clinically negative contacts were studied to evaluate the sensitivity and the specificity of quantitative saliva cultures for group A β-hemolytic streptococci. We also compared this technique with semiquantitative throat cultures as a means of isolating group A streptococci and of differentiating the streptococcal carrier state from patients with bona fide streptococcal upper respiratory tract infection as defmed by the presence of an antibody response. The data indicate that the throat culture is a more reliable means of identifying group A β-hemolytic streptococci in the upper respiratory tract than is the saliva culture. The converse is true for non-group A β-hemolytic streptococci; the saliva culture is a much better means for isolating these organisms. In individuals positive by both techniques we found good correlation between the degree of positivity of the saliva culture and the degree of positivity of the throat culture. Furthermore, while there was a definite trend for individuals with strongly positive cultures to demonstrate more often an antibody rise in either antistreptolysin O and/or antideoxynibonuclease B—indicating bona fide infection—this relationship was not sufficiently constant to provide a clear differentiation. This study also indicates that discordance (one positive, one negative) of simultaneous duplicate semiquantitative throat cultures is much more common among individuals who do not show an antibody response ("carriers") than among those with an antibody response (bona fide infection). This study confirms our previous observations suggesting that the presence of C-reactive protein in the serum of patients with a positive culture for group A streptococci and clinical signs and symptoms of pharyngitis is often an indication of true streptococcal upper respiratory tract infection, and that even with a positive saliva culture at the initial visit, a negative C-reactive protein is only infrequently (25%) associated with an antibody response.
- Received January 22, 1979.
- Accepted April 27, 1979.
- Copyright © 1979 by the American Academy of Pediatrics