Which is the best method to trace group A streptococci in sore throat patients: culture or GAS antigen test?

Objective: To compare an antigen detection test (GAS antigen test) with the results from combinations of two various bacteriological test media in general practice patients with sore throat. Furthermore to assess the diagnostic properties of the chosen GAS antigen test and to compare semi-quantitative results of this test with the bacterial load found in the throat culture.
Setting: Two Norwegian general practices in Stokke and Kongsberg communities.
Subjects: 306 patients with sore throat lasting less than 7 days; 244 were adults, 62 were children under 10 years old, mean age 23.9 years (SD 15.0), 40% were men.
Main Outcome Measures: Results from GAS antigen test, and distribution of bacteriological findings in throat cultures, compared with the results of our GAS antigen test; semi-quantitative results of the GAS antigen test compared with the bacterial load by culture.
Results: In the primary culture 110 patients harboured group A streptococci (GAS) infection, while the second culture identified another 17, giving a total of 127 patients. Some 33 patients harboured large-colony groups C and G. The GAS antigen test used had a sensitivity of 97% and specificity of 95% regarding GAS when compared with the two cultures. We found a significant correlation between the bacterial loads by culture and the semi-quantitative results of the GAS antigen test.
Conclusions: By using a second, different set of bacteriological media, we identified an additional 17 patients with GAS infections. This raises the question of validity of frequently used reference standards in studies related to streptococcal infections. Compared with the combined results of the two throat cultures, the GAS antigen test used showed high sensitivity and specificity. Semi-quantitative evaluations of the rapid immunological test may also be of clinical value.