Methods
This was a retrospective study of pregnant individuals who had Ureaplasma cervical culture at an academic institution from January 2014 to December 2020. Our Institutional Review Board approved this study. In 2014, we started a quality improvement project in which we routinely obtained Ureaplasma cervical culture for all pregnant individuals with risk factors for preterm birth (history of preterm births or PPROM, history of recurrent pregnancy loss, pregnancy with multiple gestations, or pregnancy requiring cerclage). The American College of Obstetricians and Gynecologists (ACOG) defined preterm birth as delivery less than 37 weeks’ gestation. PPROM was defined as rupture of membranes prior to the onset of labor less than 37 weeks’ gestation. Recurrent pregnancy loss is defined as 2 or more consecutive early pregnancy losses. Indications for cerclage included history indication, ultrasound indication due to the short cervix, or physical exam indication. These indications were documented in the medical record.
We excluded individuals who had Ureaplasma cervical culture for preterm contraction, multiple gestation, or unknown indications. Therefore, we only analyzed pregnant individuals with singleton pregnancies who had Ureaplasma cervical culture that was obtained for a history of preterm births, a history of recurrent pregnancy loss, or cervical cerclage. We also limited analyses to pregnant individuals who had Ureaplasma cervical culture at or prior to 20 weeks’ gestation, since we wanted to assess the effectiveness of early intervention. In addition, we excluded pregnant individuals if they delivered within two weeks of the initial Ureaplasma cervical culture since these individuals would not have enough time to receive the treatment. Pregnant individuals were categorized according to the initial Ureaplasma cervical culture results and whether they received appropriate antibiotics (negative; positive and received treatment; and positive but did not receive treatment [including those who did not receive appropriate antibiotics]).
Ureaplasma cervical culture was obtained at the first prenatal visit by inserting a vaginal speculum and removing the excess mucus from the cervical opening, using a cotton swab. The specimen collection swab was then placed within the external cervical os and gently rotated for 30 seconds for appropriate sampling. The swab was then removed avoiding contact with the vaginal walls and immediately placed into the transport tube which was securely capped. Aptima® swab was commonly utilized for this procedure. This test detects Mycoplasma genitalium, Mycoplasma hominis, and Ureaplasma species through nucleic acid amplification. If Ureaplasma cervical culture was positive, we prescribed Azithromycin for pregnant individuals and Doxycycline for sexual partners (Box 1). If pregnant individuals received antibiotics, we obtained Ureaplasma cervical culture 4 weeks after the initial treatment and treated them again if the second cervical culture was positive. Alternative antibiotics regimens in case of drug allergy are presented in Box 1.
Our primary outcome was gestational age at delivery. The estimated due date (EDD) was based on the date of the last menstrual period confirmed by first-trimester ultrasound. Secondary outcomes included Ureaplasma cervical culture positive 4 weeks after the initial treatment (treatment failure), preterm birth less than 37 weeks’ gestation, spontaneous preterm birth less than 37 weeks’ gestation, spontaneous preterm birth less than 34 weeks’ gestation, chorioamnionitis, PPROM, pregnancy loss less than 22 weeks’ gestation, neonatal intensive care unit (NICU) admission, neonatal respiratory distress syndrome (RDS) or transient tachypnea of newborn (TTN), and stillbirth or neonatal demise.
Ureaplasma treatment failure rates were assessed according to the indications for Ureaplasma cervical cultures. For the analysis of the treatment failure rate, we only included pregnant individuals who had a positive Ureaplasma cervical culture, received treatment, and had a repeat Ureaplasma cervical culture after the initial treatment. Because some individuals had several indications for Ureaplasma cervical cultures, we classified mutually exclusive categories for indications using the following hierarchy. First, if individuals had cerclage, the indication was classified as “cerclage.” Second, if individuals had a history of preterm births, the indication was classified as “history of preterm birth.” Third, if individuals had a history of recurrent pregnancy loss, the indication was classified as “recurrent pregnancy loss.” The hierarchy was maintained if individuals had more than one indication, with the highest-order indication prioritized to assign the classification. For example, if an individual had a history of preterm birth and underwent cervical cerclage, this individual was classified as “cerclage” group.
We calculated the sample size based on the following assumptions. We assumed that 90% of individuals with a positive Ureaplasma culture would receive treatment. To obtain an alpha of 0.05 and the power of 80% and detect a Hazard ratio of 0.6, we would need 224 individuals with a positive Ureaplasma culture (201 with appropriate antibiotics and 23 without appropriate antibiotics).
Descriptive statistics were calculated for all study variables. Chi-square test, Fisher’s exact test, Student’s t-test, Wilcoxon rank sum test, or Kruskal-Wallis tests were performed as appropriate. A P-value <0.05 was considered significant. We plotted the Kaplan-Meier curves to investigate the association between the gestational age at delivery and Ureaplasma culture results (negative; positive and received treatment; or positive but did not receive treatment). Log-rank test was performed to obtain P-value to compare Kaplan-Meier curves. A Cox proportional regression model was used to calculate Hazard ratio (HR) with 95% confidence intervals (95%CI), controlling for variables with a P<0.05 based on bivariable analyses. For secondary outcomes, multivariable logistic models were used to calculate adjusted P-value and adjusted odds ratios (aOR) with 95% confidence intervals (95%CI), controlling for variables with a P<0.05 based on bivariable analyses (Ureaplasma negative as a referent). A simple logistic regression was used to examine the association between the treatment failure rate and cervical culture indications (history of preterm birth as a referent). All statistical analyses were performed using Stata/SE 17.0 (StataCorp, College Station, TX).