TABLE 4

Spectrum Bias of RADT and Office BAP Culture

CategoryEstimate (95% CI), %P Comparing Sensitivitiesb
SensitivitySpecificityAccuracy
All patients (30% criterion-positive)
    RADT (n = 1843)69.6 (65.6–73.4)97.8 (96.8–98.5)89.3 (87.9–90.7)<.001
    Office BAP culture (n = 1842)80.8 (77.3–84.1)96.8 (95.7–97.7)92.1 (90.9–93.3)
    Combined (n = 1839)a85.1 (81.8–87.9)94.7 (93.4–95.8)91.8 (90.5–93.1)n/a
McIsaac scores of 0–2 (19% criterion-positive)
    RADT (n = 817)49.4 (41.3–57.4)97.8 (97.6–99.5)89.2 (87.1–91.3)<.001
    Office BAP culture (n = 815)64.7 (56.7–72.2)97.3 (95.7–98.4)91.0 (84.8–97.2)
    Combined (n = 814)a70.5 (62.7–77.5)96.2 (94.4–97.5)91.3 (89.1–93.1)n/a
McIsaac scores of 3–5 (38% criterion-positive)
    RADT (n = 1026)77.7 (73.3–81.7)96.7 (95.0–97.9)89.4 (87.5–91.3)<.001
    Office BAP culture (n = 1027)87.2 (83.5–90.4)96.4 (94.6–97.7)92.9 (91.3–94.5)
    Combined (n = 1025)a90.9 (87.6–93.5)93.2 (90.9–95)92.3 (90.5–93.9)n/a
  • The concurrent hospital laboratory BAP culture was used as the criterion standard, and results for the 6 pediatric offices were combined.

  • a Combined indicates the use of current guidelines that consider positive RADT results as diagnostic of GAS pharyngitis and recommend confirming negative RADT results with a BAP culture.

  • b McNemar's test was used. Specificities and accuracies of the RADT and office BAP culture also were compared but did not differ statistically.