Comparison of three methods for detection of pneumococcal antigen in sputum of patients with community-acquired pneumonia.

In a prospective study of 249 patients with community-acquired pneumonia, three tests for the detection of pneumococcal antigen in sputum were compared: a coagglutination test for detecting capsular antigens (Cap-CoA), a sandwich enzyme immunoassay (PnC-EIA) and a coagglutination test (PnC-CoA), both the latter detecting the pneumococcal C-polysaccharide common to all pneumococcal types. Sixty-three patients had culture-positive pneumococcal pneumonia, 45 pneumonia caused by other bacteria and 141 pneumonia of viral or unknown etiology. The sensitivity of Cap-CoA (63%) and PnC-CoA (65%) was somewhat higher than that of PnC-EIA (49%), but not significantly so. The specificity was 96-98% for all three methods. Using PnC-CoA 66 patients with possible pneumococcal infection were detected, the diagnosis being verified by culture in 41. Using Cap-CoA 59 such patients were detected, the diagnosis being verified in 40, and using the PnC-EIA 47 such patients were detected, the diagnosis being verified in 31. Antigen was found almost as often in non-purulent as in purulent samples, and as often in washed as in non-washed purulent samples. However, antibiotic treatment before the sputum sample was obtained resulted in significantly lower sensitivity of both PnC-CoA and Cap-CoA. This study confirms the high sensitivity and specificity of methods for pneumococcal antigen detection in sputum. Since CoA is easier and quicker to perform, and cheaper than the EIA, either PnC-CoA or Cap-CoA would seem to be the technique of choice for detection of pneumococcal antigen, whereby all sputum samples, including non-purulent samples, can be used.