Mitchell Levine wonders how often physicians in the study missed cases of group A Streptococcus infection that would have been caught had the score approach been used. These data were omitted from the final version of the article to meet the word limit requested by CMAJ's editors. We did, however, note that the physicians missed substantially more cases of streptococcal infection in children (20%) than if they had used the score approach (6%, p = 0.006) [Research].²

In the study, physicians identified 85 of 102 cases of streptococcal infection (83.1%).² The false-negative rate of 16.9% for physician judgement is not less than the 15% rate for the score. In addition, this estimate for physician sensitivity is somewhat higher than the 50-75% estimate generally reported in other studies.^3,4,5 However, all family physicians in the present study were provided with an article about the sore throat score and a laminated pocket version of the score for quick reference; this may have affected their performance.

In the original study, in which no information about the score was provided, the sensitivity of usual physician care was 69.4% compared with 83.1% for the score (p = 0.06) [Education].⁶ This result is more in keeping with published reports and suggests that physicians miss 25%–50% of cases of group A Streptococcus when they rely on their clinical judgement. As a result, front-line practitioners can be reassured that they are likely to miss fewer cases of group A Streptococcus when they use the score approach than when they rely on their clinical judgement.