Objective
We sought to evaluate the contributions of vaginal bleeding and cervical length to the risk of preterm birth.
Study Design
This was a secondary analysis of a cohort study designed to study predictors of preterm birth. The study included 2988 women with singleton gestations. Women underwent midtrimester transvaginal ultrasound assessment of cervical length and were queried regarding first- and second-trimester vaginal bleeding.
Results
There was a significant second-order relation between cervical length and preterm birth ( P < .001, P = .005). Women with vaginal bleeding were at higher risk of preterm birth (odds ratio, 1.5; 95% confidence interval, 1.3–2.0). There was a significant interaction between cervical length and vaginal bleeding ( P = .015). After accounting for cervical length and interaction, the adjusted odds ratio for vaginal bleeding and preterm birth was 4.8 (95% confidence interval, 1.89–12.4; P = .001).
Conclusion
The magnitude of risk of preterm birth associated with sonographic cervical length depends on a woman’s history of first- and second-trimester vaginal bleeding.
Preterm birth accounts for >12% of all births and is a major cause of neonatal morbidity and mortality. A known risk factor for preterm birth is a sonographic short cervix.
First- and second-trimester bleeding alone is also associated with an increased risk of preterm birth. This risk increases as bleeding persists throughout the first and second trimesters. Multiple previous studies have identified this relationship. In a recent prospective study of 2802 women with singleton pregnancies, the relative risk of preterm birth in women with bleeding in both the first and second trimesters was 2.4 (95% confidence interval [CI], 1.6–3.6). This increase in risk remained after accounting for maternal age, drug use, and prior adverse pregnancy outcomes.
However, no data exist regarding the interaction between a sonographic short cervix and vaginal bleeding during pregnancy. We sought to evaluate the independent and interactive contribution of vaginal bleeding and cervical length to the risk of spontaneous preterm birth.
Materials and Methods
This study is a secondary analysis of a large multicenter cohort study, the Preterm Prediction Study of the National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network. In the original study, >3000 women who were receiving routine prenatal care were recruited from 1992 through 1994 to identify risk factors for preterm birth. The current study was exempt from the institutional review board process and approval because it was a secondary analysis of the Preterm Prediction Study.
In the current analysis, women with singleton pregnancies with known data regarding cervical length, vaginal bleeding history, and pregnancy outcome were included. Women were excluded if there were major fetal anomalies, multifetal gestations, current cervical cerclage, human immunodeficiency virus positive status, or a placenta previa. The primary outcome was spontaneous preterm birth at <36 completed weeks.
Midtrimester assessment of cervical length was obtained by transvaginal ultrasound with methods as published by Iams et al. The median gestational age at this study visit was 23.9 weeks with a range of 20.8–28 weeks’ gestation. This corresponds to the first study visit in the original study.
A standardized interview at the same study visit was used to assess vaginal bleeding during pregnancy. Specifically, first- and second-trimester bleeding was assessed. Women were queried whether any vaginal bleeding or spotting had occurred in the first and/or second trimester with either yes or no responses. Any bleeding episode was included in this analysis regardless of number of episodes or whether it occurred in the first or second trimester or both trimesters.
Both assessment of cervical length and vaginal bleeding history occurred contemporaneously during 1 study visit and outcomes were ascertained prospectively. The women enrolled in the study and their health care providers were blinded to the cervical length ultrasound results and interview regarding bleeding history.
Statistical analysis was performed with software (STATA 10; Stata Corp, College Station, TX). Logistic regression was used to assess the relation between cervical length and vaginal bleeding to the risk of spontaneous preterm birth at <36 completed weeks. Covariates considered as possible confounders or effect modifiers of these relationships were maternal age, maternal race, marital status, cigarette use, and history of preterm birth. Covariates were retained in the final logistic regression model if they were associated with the outcome ( P < .10) or if their inclusion in the model changed the beta describing the relation between the primary exposures and the outcome variable by ≥10%. For all primary hypothesis tests, P < .05 was considered statistically significant.
Results
Of the 3073 women enrolled in the original study, 2988 women were included in the current study. The women excluded from the study did not have complete follow-up data. The median gestational age at entry into the study was 23.9 weeks. Demographic data collected during the study visit are shown in Table 1 . The women excluded from the study did not differ significantly from the study population (data not shown).
