Cervical length for prediction of preterm birth in women with multiple prior induced abortions

Authors

  • J. Visintine,

    Corresponding author
    1. Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Thomas Jefferson University, Philadelphia, PA, USA
    • Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Thomas Jefferson University, 834 Chestnut St., Suite 400, Philadelphia, PA 19107-5127, USA
    Search for more papers by this author
  • V. Berghella,

    1. Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Thomas Jefferson University, Philadelphia, PA, USA
    Search for more papers by this author
  • D. Henning,

    1. Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Thomas Jefferson University, Philadelphia, PA, USA
    Search for more papers by this author
  • J. Baxter

    1. Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Thomas Jefferson University, Philadelphia, PA, USA
    Search for more papers by this author

Abstract

Objective

To determine whether transvaginal sonographic cervical length predicts preterm birth in women with multiple prior induced abortions.

Methods

This was a retrospective cohort study using the Thomas Jefferson University Prematurity Database. Patients with a singleton pregnancy and a history of more than one induced abortion were identified. Exclusion criteria were cerclage and indicated preterm birth. Subjects were followed with transvaginal ultrasound measurement of the cervix between 14 and 24 weeks' gestation and grouped into those with and those without a short cervix; a cervical length of < 25 mm was considered short. The primary outcome was spontaneous preterm birth at < 35 weeks.

Results

Fifteen of the 65 (23%) women with more than one induced abortion included in the study had a short cervix. The demographics and risk factors were similar between those with and those without a short cervix. The overall incidence of preterm birth was 21.5% (14/65); in women with a short cervix the incidence was 47% (7/15) and in women without a short cervix it was 14% (7/50). The sensitivity, specificity and positive and negative predictive values of a short cervix in the prediction of preterm birth were 50%, 84%, 47% and 86%, respectively. The relative risk of a short cervix for spontaneous preterm birth was 3.3 (95% CI, 1.4–7.4).

Conclusion

A cervical length of < 25 mm on transvaginal ultrasound is predictive of preterm birth in women with more than one prior induced abortion. Women with multiple prior induced abortions and a short cervix have a 3.3-fold greater chance of spontaneous preterm birth compared with those with a cervical length of ≥ 25mm. Copyright © 2008 ISUOG. Published by John Wiley & Sons, Ltd.

Introduction

Preterm birth is a significant public health concern around the world. In the United States, 7.7% of all live births are delivered prior to 35 weeks' gestation. Preventive efforts have focused on women at high risk of preterm birth, such as those with a prior spontaneous preterm birth. For women deemed to be at high risk for preterm birth based on risk factors, cervical length as obtained by transvaginal ultrasound has emerged as an important prognostic tool1. The ability of cervical length to predict preterm birth has been evaluated in specific high-risk populations, such as in women with a prior preterm birth, twin gestations or uterine anomalies2–5.

A history of multiple induced abortions has been implicated as a risk factor for preterm birth6–8. However, the role of cervical length to predict preterm birth has not been studied specifically in women with a history of multiple induced abortion. The objective of this study was to evaluate the predictive accuracy of a short cervix for preterm birth in women who have had multiple induced abortions.

Methods

This was a retrospective cohort study using the Thomas Jefferson University Prematurity Database and approved by The Thomas Jefferson University Hospital institutional review board. Included in this database are all women from 1995 onwards presenting for prenatal care at Thomas Jefferson University Hospital or one of its affiliates who are found to have risk factors for preterm birth (prior preterm birth, multiple gestation, prior cervical conization, multiple induced abortions and Mullerian abnormalities). Women included in this database were followed with transvaginal ultrasound examination of the cervix approximately every 2 weeks. Our study included women entered in the database during the period 1995 to 2005 who had more than one prior induced abortion, who had cervical length data for the current pregnancy from 14 to 23 + 6 weeks, and for whom outcomes were known. Induced abortion was defined as the voluntary termination of a viable pregnancy at ≤ 13 weeks. We excluded those with history-indicated or ultrasound-indicated cerclage, multiple gestations, medically indicated preterm births and fetal deaths.

All transvaginal ultrasound cervical length measurements were performed by experienced sonographers using standard techniques. The patient's bladder was emptied prior to visualization of the cervix. Only the minimum pressure necessary was used to obtain a clear image of the cervical canal in the mid-sagittal plane. At least three cervical measurements were obtained, and the shortest measurement was used in this study; all data analysis was based on the shortest cervical length obtained between 14 and 23 + 6 weeks.

Patients were divided into two groups, according to cervical length. Short cervix was defined as a cervical length < 25 mm. The primary outcome was spontaneous preterm birth at < 35 weeks' gestation. Statistical analysis was performed using the SPSS 13.0 statistical package (SPSS Inc., Chicago, IL, USA). Chi square, Fisher's exact and Student's t-tests were used as appropriate.

Results

We identified 133 patients with more than one induced abortion, of which 68 were excluded, leaving 65 for data analysis (Figure 1). Of the 68 excluded patients, 25 were excluded for a history-indicated cerclage, 22 for an ultrasound-indicated cerclage, 16 secondary to multiple gestation, three for indicated preterm birth, one due to a second-trimester termination for a lethal fetal anomaly and one due to a second-trimester fetal demise.

Figure 1.

Flow diagram showing delivery outcome of women involved in the study.

A short cervix was noted in 23% (15/65) of all patients identified with more than one prior induced abortion. The overall incidence of spontaneous preterm birth was 21.5% (14/65). Demographic characteristics and risk factors for preterm birth in the patients with a cervical length < 25 mm and those with a cervical length ≥ 25 mm are given in Table 1. The incidence of spontaneous preterm birth in patients with a cervical length < 25 mm was 47% (7/15). For patients with a cervical length ≥ 25 mm the incidence was 14% (7/50). The sensitivity, specificity and positive and negative predictive values of a cervical length < 25 mm in the prediction of spontaneous preterm birth were 50%, 84%, 47% and 86%, respectively. The relative risk of a cervical length < 25 mm for spontaneous preterm birth was 3.3 (95% CI, 1.4–7.4).

Table 1. Demographic characteristics of women in the study group according to cervical length
CharacteristicCervical lengthP
< 25 mm (n = 15)≥ 25 mm (n = 50)
  • There were no cases with Mullerian anomaly.

  • *

    Student's t-test.

  • Chi-square test.

  • Fisher's exact test.

Maternal age (years, mean (SD))28 (5.2)30 (5.4)0.267*
Ethnicity (n (%))  0.055
 African American14 (93)28 (56) 
 Caucasian013 (26) 
 Other1 (7)3 (6) 
 Unknown06 (12) 
Smoker (n (%))2 (13)10 (20)0.562
Prior preterm delivery (n (%))6 (40)20 (40)1.00
Prior cervical conization (n (%))02 (4)1.00
Women with prior second-trimester induced abortion (n (%))02 (4)1.00
Prior first-trimester induced abortions (n per woman (mean (SD)))2.7 (1.1)2.5 (0.86)0.478*

Discussion

We have demonstrated that for women with more than one induced abortion, a cervical length < 25 mm obtained by transvaginal ultrasound between 14 and 24 weeks is predictive of spontaneous preterm birth. In order to determine if any prior study had investigated this specific topic, Medline and PubMed literature searches were performed in May 2006 using the key words: induced abortion, pregnancy termination, transvaginal ultrasound, preterm birth and cervical length, both individually and as phrases. We did not identify any articles which specifically evaluated the use of transvaginal ultrasound in women with multiple induced abortions for the prediction of preterm birth.

In the general population, a short cervical length (< 25 mm at 16–24 weeks) is associated with an increased risk of preterm birth, but the usefulness of the test in the general population is limited, with a positive predictive value of only 18%9. However, for women identified as having a higher likelihood of preterm birth based on risk factors, the positive predictive value of a short cervical length ranges from 35 to 75%4, 5. Our findings suggest that for women with a history of more than one induced abortion, cervical length according to transvaginal ultrasound is predictive of preterm birth, with a positive predictive value of 47%. These data are consistent with those of previous studies of women at increased risk for preterm birth.

Since the legalization of induced abortion in the United States, multiple epidemiological studies have been conducted to determine the effect of induced abortion on future pregnancies. The majority of induced abortions in the United States are performed mechanically, by cervical dilatation and suction curettage. There has been concern that mechanical dilation may result in injury to the cervix10, which may increase the risk of cervical insufficiency and preterm birth. The majority of early cohort studies evaluating subsequent pregnancy outcome for women who had an induced abortion in their first pregnancy demonstrated no significant increase in risk for preterm birth11. Levin et al.12, however, demonstrated that women with a history of more than one induced abortion had a two- to three-fold increased risk of delivery at < 28 weeks. Recent large European epidemiological studies support the association between multiple induced abortions and preterm birth6–8. Moreau et al.7, in the EPIPAGE study, showed that a history of more than one induced abortion was a risk factor for very preterm birth, defined as delivery between 22 and 32 weeks. In this multicenter study, which included 1943 births at < 33 weeks, women with a history of one induced abortion had an adjusted odds ratio for very preterm birth of 1.3 (95% CI, 1.0–1.8), while women with a history of more than one induced abortion had an adjusted odds ratio for very preterm birth of 2.6 (95% CI, 1.1–5.9)7. It appears that the risk of preterm birth is increased in women who have had more than one induced abortion.

While our study is the first to assess the predictive value of sonographic cervical length in this particular population, there are shortcomings. A large number of patients, those with multiple gestation or with cerclage, were excluded from our original cohort. In addition, patients included in our study population had other potentially confounding risk factors for preterm birth, such as prior spontaneous preterm birth and smoking, in addition to a history of induced abortion. These were included in order to make our population as representative as possible and therefore our results more amenable to extrapolation to the general population.

There are no published studies evaluating interventions such as cerclage, progesterone, bed rest and tocolytics for women with more than one induced abortion who have a short cervical length (< 25 mm) in the second trimester. While our study suggests that a cervical length of < 25 mm is predictive of preterm birth in this population, until a proved treatment to prevent preterm birth in these patients is available, routine screening for short cervical length cannot be recommended.

Ancillary