Effect of Genital Sampling Site on the Detection and Quantification of Species with Quantitative Polymerase Chain Reaction during Pregnancy.

Objective. This study aimed to compare the qualitative and quantitative reproducibility of quantitative PCR (qPCR) for Ureaplasma species (Ureaplasma spp.) throughout pregnancy and according to the genital sampling site. Study Design. Between 5 and 14 weeks of gestation (T1), vaginal, fornix, and two cervical samples were taken. Sampling was repeated during the 2nd (T2) and 3rd (T3) trimester in randomly selected T1 positive and negative women. Qualitative and quantitative reproducibility were evaluated using, respectively, Cohen's kappa (κ) and interclass correlation coefficients (ICC) and repeated measures ANOVA on the log-transformed mean number of DNA copies for each sampling site. Results. During T1, 51/127 women were positive for U. parvum and 8 for U. urealyticum (4 patients for both). Sampling was repeated for 44/55 women at T2 and/or T3; 43 (97.7%) remained positive at the three timepoints. κ ranged between 0.83 and 0.95 and the ICC for cervical samples was 0.86. Conclusions. Colonization by Ureaplasma spp. seems to be very constant during pregnancy and vaginal samples have the highest detection rate. [ABSTRACT FROM AUTHOR]