Optimizing methods to detect colonization with fluoroquinolone-resistant viridans group streptococci prior to hematopoietic cell transplantation

Patients undergoing hematopoietic cell transplantation (HCT) who receive fluoroquinolone prophylaxis are at high risk for bacteremia due to fluoroquinolone-resistant viridans group streptococci (FQ-R VGS), particularly Streptococcus mitis. Detecting colonization with FQ-R VGS may identify neutropenic patients at risk for streptococcal bacteremia during FQ prophylaxis, but the optimal specimen screening site and microbiologic screening method are unknown. We therefore simultaneously collected buccal, oropharyngeal, and perianal swabs from patients upon admission for HCT, and performed selective culture with both direct plating and broth enrichment. FQ-R VGS were detected in 62 (90%) of 69 participants. Direct plating resulted in a positive percent agreement (PPA) of 79% (95% confidence interval [CI]: 67%–88%) for detecting FQ-R VGS from buccal swabs, and a PPA of 86% (95% CI: 74%–93%) from oropharyngeal swabs compared to detection by any specimen or method. Broth enrichment yielded a PPA of 97% (95% CI: 89%–100%) for both buccal and oropharyngeal swabs. Perianal swabs had low PPA. For detecting FQ-R S. mitis group, the direct plating PPA was 74% (95% CI: 60%–85%) for buccal and 72% (95% CI: 58%–84%) for oropharyngeal swabs, while the broth enrichment PPA was 82% (95% CI: 69%–91%) and 58% (95% CI: 43%–72%), respectively. Combining direct plating and broth enrichment of buccal swabs had a PPA of 96% (95% CI: 86%–100%) for detecting FQ-R S. mitis group. Buccal swabs are a reliable specimen to detect FQ-R VGS colonization when using broth enrichment, and combining direct plating and broth enrichment may optimize detection of FQ-R S. mitis group.