RESUMEN
The performance of the recently commercialized Uni-Gold™ Streptococcus pneumoniae test for the detection of pneumococcal antigen in urine was studied in a multicenter study. First, we studied the interassay agreement between Uni-Gold™ and the BinaxNOW® S. pneumoniae urinary antigen test on 337 consecutive urine samples sent to the laboratory for the detection of pneumococcal antigen. The two tests performed similarly (κ = 0.82): both tests positive in 27 cases, both tests negative in 299 cases, and with divergent test results in 11 cases. Secondly, the tests were run on urine samples from 203 patients with bacteremia, including 51 patients with pneumococcal bacteremia. The sensitivities and specificities were 67 and 86 % for Uni-Gold™, and 57 % and 94 % for BinaxNOW®, respectively. The false-positivity rate was significantly higher for Uni-Gold™ compared with BinaxNOW® in patients with Escherichia coli bacteremia (15 vs. 2.1 %, p = 0.04), and tended to be higher in patients with bacteremia with alpha-hemolytic streptococci (32 vs. 11 %, p = 0.13). When cases with E. coli and alpha-hemolytic streptococci were excluded from the analysis, the overall false-positivity rate was 9/85 (11 %) for Uni-Gold™ and 6/85 (7.1 %) for BinaxNOW®. In conclusion, the study showed that Uni-Gold™ was not inferior to BinaxNOW® for the detection of pneumococcal urinary antigen in patients with pneumococcal bacteremia. The specificity of Uni-Gold™ was suboptimal due to false-positive results in cases with E. coli and alpha-hemolytic streptococci bacteremia. However, in patient populations usually subjected to testing for pneumococcal urinary antigen, such as pneumonia and meningitis patients, bacteremia with these pathogens is uncommon. The diagnostic usefulness of the Uni-Gold™ test should be further evaluated.