Evaluation of patient- versus provider-collected vaginal swabs for microbiome analysis during pregnancy

Abstract

Abstract Objective We aimed to evaluate if patient- and provider-collected vaginal swabs in pregnant women reflect similar bacterial community characteristics. Pregnant patients performed a self-collected vaginal swab, then underwent a provider-collected swab via speculum exam. DNA pyrosequencing of the 16S rRNA gene V1V3 and V3V5 variable regions was performed. Relative abundance of taxa, alpha diversity, and beta diversity of patient- and provider-collected swabs were compared. Results Ninety-four vaginal swabs from 47 women were analyzed. On non-metric multi-dimensional scaling plots, paired patient- and provider-collected swabs clustered closely. The median Pearson correlation coefficient was 0.993 (interquartile range 0.951–0.999) for V1V3 and 0.987 (interquartile range 0.902–0.999) for V3V5. Among paired V1V3 and V3V5 sequences, 83.0% and 73.9% showed strong Pearson correlation (> 0.9), respectively, between patient- and provider-collected swabs; V1V3 and V3V5 sequences with weaker Pearson correlation (< 0.9) had correlation coefficients 0.57–0.89 and 0.49–0.89, respectively. No taxa were preferentially detected by sampling method, with relative abundance of taxa highly conserved. No significant difference in Shannon diversity for V1V3 (p = 0.22) and V3V5 (p = 0.11) sequences among paired samples was seen. We demonstrate that bacterial communities defined from patient- and provider-collected vaginal swabs in pregnant women are similar, validating utilization of patient-collected swabs for vaginal bacterial microbiome sampling during pregnancy.

Keywords

• 16S rRNA
• Patient-collected sample
• Pregnancy
• Vaginal microbiome