to the editor: The article¹ on pharyngitis by Vincent and colleagues raises several questions about their approach to diagnosis and treatment of group A beta-hemolytic streptococcal pharyngitis.

They report the sensitivity and specificity of throat culture as 97 and 99 percent, respectively, for diagnosing group A beta-hemolytic streptococci, citing McIsaac’s article,² which seems to be higher than reported in the literature. McIsaac and colleagues² compared the sensitivity and specificity of a score-based approach to the diagnosis of sore throat in their population as compared to throat culture, but did not provide the sensitivity and specificity of throat culture in their article. Perhaps a more accurate estimate comes from the study by Gerber and colleagues.³ The sensitivity and specificity of culture is reported as 78 and 99 percent, respectively. Similarly, in an article published in American Family Physician, Hayes and colleagues⁴ state that under ideal conditions, the sensitivity of throat culture for group A beta-hemolytic streptococci is only 90 percent; in office settings, the sensitivity ranges from 29 to 90 percent.⁴ The specificity of throat culture is 99 percent under ideal conditions and can be anywhere from 76 to 99 percent in office settings.⁴

The authors also recommend obtaining cultures if symptoms do not improve, citing the article by Attia.⁵ Attia and colleagues sought to determine the performance of a predictive model for group A beta-hemolytic streptococcal pharyngitis and did not recommend follow-up cultures if symptoms do not improve.⁵ We fail to understand the reasons for obtaining cultures in these patients as shown in Figure 1 of the article¹ by Vincent and colleagues because some of these patients have already been treated based on rapid strep test results. It is unclear whether they intend to identify bacteriologic failures, relapse, or reinfection.

in reply: Drs. Singh and Dolan raise the question of the sensitivity of throat culture to diagnose streptococcal pharyngitis and reinforce the importance of obtaining the specimen under “ideal conditions.” In our article,¹ we recognized the significance and described the proper technique for obtaining a throat culture under the subsection “Laboratory Evaluation” and state in Table 1 a sensitivity of “97 percent results dependent on the technique, medium, and incubation.”² Several of the references that support a sensitivity greater than 90 percent of throat culture for the diagnosis of group A beta-hemolytic streptococcus report a sensitivity of 90 to 97 percent when performed properly.^2–5 A question also is raised about the necessity of obtaining a throat culture in the subset of patients who do not improve following treatment. Bisno and colleagues² state that these patients require reculture to identify treatment failures, reinfection, or relapse.