A double-blinded comparative evaluation of three media for chromophore testing with viridans and nutritionally...

A double-blinded prospective comparison of chromophore testing among 109 clinical isolates of alpha-hemolytic gram-positive cocci and 48 strains of nutritionally variant (deficient) streptococci (NVS) for its sensitivity, specificity, and predictive value under stimulated clinical laboratory conditions was performed after growth in three different media--Todd-Hewitt broth supplemented with 10 micrograms/mL of pyridoxal (THBP), THBP with 0.8% (w/v) yeast-extract (THBP + YE), and a semisynthetic chemically defined medium with Todd-Hewitt dialyzate (CDMT). The chromophore detection rates of Streptococcus mitis were 60.7, 67.9, and 78.6% after growth in THBP, THBP + YE, and CDMT, respectively. After growth in THBP there was significantly lower chromophore detection than either THBP + YE or CDMT for NVS (P less than 0.001), with a trend toward significantly less detection for S. mitis (0.15 greater than P). For Streptococcus sanguis II the detection rates were 45.7, 57.1, and 54.3%, respectively, for each medium. There was no significant difference between the detection rates after growth in THBP + YE or CDMT. One NVS was persistently chromophore negative, the first so described. There was a similar number of presumably false positive chromophore S. sanguis I strains (3 of 12) detected after growth in THBP + YE or CDMT. The sensitivity of chromophore detection for NVS (pooling results in THBP + YE and CDMT) was 95.8% and specificity was 59.6%. For a nonvariant streptococcus, the sensitivity was 61.9% and specificity was 93.5% for an isolate to be either S. mitis or S. sanguis II.