Preterm birth remains a leading cause of perinatal morbidity and mortality.1
Fortunately, the rate of preterm birth in the United States has dropped for the sixth consecutive year to a 15-year low of 11.5% in 2013.2 This achievement may be at least partially attributable to the identification of at-risk patients using second-trimester ultrasound evaluation of the cervix along with the use of interventions to prevent preterm birth such as vaginal progesterone and in patients with a history of prior preterm birth or cerclage.3,4 For these reasons, many institutions have adopted a universal transvaginal cervical length screening protocol (Khalifeh A, Quist-Nelson J, Berghella V. Current implementation of universal cervical length screening for preterm birth prevention in the United States  [abstract]. Obstet Gynecol 2016;127(suppl 1):7S).
Although antenatal corticosteroids are not an intervention to reduce preterm birth, the administration of antenatal corticosteroids before preterm birth significantly reduces associated neonatal morbidity and mortality.5 Timing of corticosteroids appears to be a critical factor in their efficacy. Although reduction of neonatal death is seen at less than 24 hours after a first dose, studies have failed to show demonstrable mortality or morbidity benefits at more than 7 days from the first dose.6 We hypothesized that health care provider concern for impending preterm labor in patients with an incidental finding of a short cervix might influence hasty steroid dosing, potentially putting these patients at risk for suboptimal antenatal corticosteroid timing.
The aim of this study was to assess the relationship between patients identified as having a short cervix as part of a universal transvaginal cervical length screening protocol and the timing of administration of antenatal corticosteroids in relation to the timing of birth.
MATERIALS AND METHODS
Our institution adopted a universal transvaginal cervical length screening protocol in October 2012 for all women with a singleton gestation undergoing a routine anatomy ultrasonogram between 17 0/7 weeks and 23 6/7 weeks of gestation. After obtaining Montefiore Medical Center institutional review board approval, we retrospectively analyzed the electronic records of all women who experienced spontaneous preterm delivery between 24 0/7 weeks and 34 0/7 weeks of gestation after the initiation of universal transvaginal cervical length screening (October 2012) at our institution until August 2015.
Patients with major fetal anomalies, multiple gestations, iatrogenic preterm deliveries resulting from maternal or fetal indications, and patients who did not deliver at our institution were excluded.
Maternal demographics and medical and obstetric history data were obtained through electronic medical records. Gestational age and cervical length at the initial transvaginal ultrasonogram and on any follow-up ultrasonograms were recorded. Short cervix was defined as 2.0 cm or less in the absence of prior spontaneous preterm birth or less than 2.5 cm if there was a history of prior spontaneous preterm birth.7 Neonatal data including gestational age at delivery, birth weight, gender, admission to the neonatal intensive care unit, respiratory distress syndrome (RDS), and neonatal demise were also recorded. Each maternal and neonatal record was individually reviewed by study investigators.
All episodes of antenatal corticosteroid dose administration were identified. We considered one dose to be an incomplete initial steroid course, two doses a completed course, three doses an incomplete rescue course, and four doses as a complete rescue course. The interval of days to delivery from antenatal corticosteroid administration was recorded. In the event a patient received a rescue antenatal corticosteroid course, timing from antenatal corticosteroid administration to delivery was based on the patient's most recent (ie, rescue) antenatal corticosteroid course. Optimal steroid timing was considered a steroid to delivery interval of 24 hours to 7 days (allowing for a complete steroid course).
The primary outcome was delivery within 7 days of the first dose (initial or rescue) of antenatal corticosteroid. Secondary outcomes included delivery 24 hours to 7 days after the initial steroid injection, antenatal corticosteroid to delivery interval in days, the incidence of delivery after a complete steroid course, neonatal survival, neonatal intensive care unit length of stay, and RDS.
We tested normality using the Shapiro-Wilk test and Q-Q plots. We used unpaired Student t test or Wilcoxon rank-sum test for between-group comparisons, as appropriate. χ2 test and Fisher exact test were used for inferences on proportions.
A multivariable logistic regression model was used to quantify the association between optimal timing of antenatal corticosteroids and a short cervix. The model building was done according to Hosmer and Lemeshow.8 Variables identified as potential confounding factors and those with a P value of <.25 in the univariable analysis were included in the multivariable logistic regression model. Using a backward elimination approach, the likelihood ratio test comparing the model including the variable with the nested model excluding it was used to assess whether the variable contributed significantly to the model (P<.20). The Hosmer–Lemeshow goodness-of-fit test was used to assess calibration of the model.
Among patients with known short cervix, the potential association of cerclage and the use of vaginal progesterone on the timing of antenatal corticosteroid administration was also studied. Data are expressed as mean (standard deviation), median (interquartile range), or odds ratio (95% confidence interval), unless otherwise stated. A two-sided P value <.05 was considered significant. All analyses were performed with STATA/SE 12.1.
A total of 1,253 patients had spontaneous preterm birth before 34 weeks of gestation within the study period, October 2012 to August 2015. Of those, 987 were excluded from the analysis for multiple gestation, fetal anomalies, iatrogenic preterm birth resulting from secondary causes such as preeclampsia and growth restriction, or missing data. Of the remaining 266 patients, 69 (25.9%) had a short cervix at the time of their universal transvaginal cervical length screen and 197 (74.1%) did not (Fig. 1).
Patients in the short cervical length group had a median cervical length of 1.5 (0.8–1.9) cm at the time of the second-trimester transvaginal ultrasonography (Fig. 2) compared with 3.9 (3.3–4.5) cm in the no short cervical length group (P<.001) and were more likely to deliver at an earlier gestational age (P=.006; Table 1).
During the study period, 64 of 69 (92.8%) of patients with a short cervix and 176 of 197 (89.3%) without a short cervix received at least one steroid injection before delivery (P=.411). At least one dose of steroids was given within 7 days of delivery in 33 of 69 (47.8) patients with a short cervix compared with 126 of 197 (64%) patients in the no short cervix group (P=.015). When the optimal interval of 24 hours to 7 days was considered, there was no difference between groups (30/69 [43.5%] of those with a short cervix compared with 87/197 [44.2%] of those without; P=.922).
The results of the univariable and multivariable logistic regression analysis are shown in Tables 2 and 3, respectively. The adjusted odds ratio of receiving antenatal corticosteroid within 7 day before delivery was 0.51 (95% confidence interval 0.29–0.9) when comparing the short cervical length group with the no short cervical length group. The median interval between the antenatal corticosteroid administration and delivery was significantly shorter in the no short cervical length group compared with the short cervical length group (3 days compared with 8 days, P<.001; Table 4).
The median injection to delivery interval for patients who delivered less than 7 days but before receiving a complete steroid course was 6.9 (1.1–9.9) hours. For the seven patients in the short cervical length group, the median dose to delivery interval was 11.5 (1.6–18.1) hours and for those in the no short cervical length group (n=39), the median dose to delivery interval was 5.7 (1.1–9.6) hours (P=.293).
Within the short cervical length group, a cerclage procedure was performed in 23 (33.3%) patients and vaginal progesterone was used by 29 (42%). In the subgroup analyses, no differences were seen in the optimal timing of antenatal corticosteroid administration and the presence of cerclage or use of vaginal progesterone (P=.610 and P=.164, respectively).
Although neonates from women with short cervical length had a longer neonatal intensive care unit stay compared with women without a short cervical length (47 compared with 34 days; P=.011), this difference was not significant after controlling for gestational age at delivery (P=.149). Similarly, there was no difference in neonatal survival or RDS between the groups (Table 4).
In this study, we found that patients identified with a short cervix on universal transvaginal cervical length screening and who subsequently delivered prematurely had a significantly longer interval from antenatal corticosteroid administration (8 days compared with 3 days) and fewer of them received antenatal corticosteroid within 7 days of birth compared with patients with spontaneous preterm birth who were not identified as having a short cervix by universal transvaginal cervical length screening. However, a similar percentage of patients in both groups delivered within the optimal interval of 24 hours to 7 days. Neonatal outcomes were similar between these two groups; however, our study was not powered to assess differences in neonatal outcomes.
This study highlights the clinical dilemma for physicians caring for patients with an incidentally identified short cervix trying to balance the risk of administering corticosteroids too early with the risk of administering them too late. Although available data suggest that a timely yet incomplete course of antenatal corticosteroids may confer greater benefit over a remote yet completed course,6 robust data are lacking. This dilemma extends to rescue steroids: Makhija et al9 showed that the option of giving a rescue course did not translate into an increased number of patients receiving optimally timed steroids.
Our results are in contrast to the findings of Adams et al10 who identified a cervical length less than 2 cm as a factor that was associated with optimal timing of antenatal corticosteroids. Our findings do, however, align with the results of Makhija et al.9 They noted that patients with a short cervical length had the lowest rates of optimally timed steroids among their entire cohort (only 2.8% compared with an average of 25% for other groups delivered within the optimal window).
The significance of suboptimally timed steroids is yet to be fully clarified. In 1972, groundbreaking work by Liggins et al5 compared unexposed patients with those who had received antenatal corticosteroid less than 24 hours, 24–48 hours, 2–7 days, and then at greater than 7 days exposure to delivery intervals and found a significant improvement in the rate of RDS only in the 2- to 7-day group. Subsequent studies comparing neonates who received antenatal corticosteroid with optimal compared with suboptimal timing have been conflicting.11–14 In a meta-analysis, Crowley found that neonates who were delivered after optimally timed steroids had an odds ratio of 0.35 for RDS, whereas those who were delivered less than 24 hours or greater than 7 days after receiving steriods did not demonstrate a statistically significant reduction in RDS.15 Unfortunately, all of these studies have all been limited by sample size and have focused primarily on the prevention of RDS.
Our study was not without limitations. We had outcome data only for women who delivered in our institution, and they comprised the minority of women who underwent universal transvaginal cervical length screening. Our relatively small sample size limited our statistical power to detect differences in neonatal outcomes. Management decisions made about corticosteroid administration were not standardized. We did not include patients with a short cervix who went on to deliver after 34 weeks of gestation. Including those patients would have likely contributed to a longer steroid to delivery interval for patients in the short cervix group.
In summary, a policy of universal transvaginal cervical length screening may increase the likelihood that corticosteroids for fetal maturation are administered less than optimally.
1. Behrman RE, Butler AS. Preterm birth: causes, consequences, and prevention. Washington, DC: National Academies Press (US); 2007.
2. Mathews TJ, MacDorman MF, Thoma ME. Infant mortality statistics from the 2013 period linked birth/infant death data set. National Vital Statistics Reports. Hyattsville (MD): National Center for Health Statistics; 2015.
3. Iams JD, Goldenberg RL, Meis PJ, Mercer BM, Moawad A, Das A, et al. The length of the cervix and the risk of spontaneous premature delivery. National Institute of Child Health and Human Development Maternal Fetal Medicine Unit Network. N Engl J Med 1996;334:567–72.
4. Hassan SS, Romero R, Vidyadhari D, Fusey S, Baxter JK, Khandelwal M, et al. Vaginal progesterone reduces the rate of preterm birth in women with a sonographic short cervix: a multicenter, randomized, double-blind, placebo-controlled trial. Ultrasound Obstet Gynecol 2011;38:18–31.
5. Liggins GC, Howie RN. A controlled trial of antepartum glucocorticoid treatment for prevention of the respiratory distress syndrome in premature infants. Pediatrics 1972;50:515–25.
6. Roberts D, Dalziel S. Antenatal corticosteroids for accelerating fetal lung maturation for women at risk of preterm birth. The Cochrane Database of Systematic Reviews 2006, Issue 3. Art. No.: CD004454.DOI: 10.1002/14651858.CD004454.pub2.
7. Prediction and prevention of preterm birth. Practice Bulletin No. 130. American College of Obstetricians and Gynecologists. Obstet Gynecol 2012;120:964–73.
8. Hosmer DW, Lemeshow S. Applied logistic regression. New York (NY): John Wiley & Sons; 2000.
9. Makhija NK, Tronnes AA, Dunlap BS, Schulkin J, Lannon SM. Antenatal corticosteroid timing: accuracy after the introduction of a rescue course protocol. Am J Obstet Gynecol 2016;214:120.e1–6.
10. Adams TM, Kinzler WL, Chavez MR, Fazzari MJ, Vintzileos AM. Practice patterns in the timing of antenatal corticosteroids for fetal lung maturity. J Matern Fetal Neonatal Med 2015;28:1598–601.
11. Waters TP, Mercer B. Impact of timing of antenatal corticosteroid exposure on neonatal outcomes. J Matern Fetal Neonatal Med 2009;22:311–4.
12. Sehdev HM, Abbasi S, Robertson P, Fisher L, Marchiano DA, Gerdes JS, et al. The effects of the time interval from antenatal corticosteroid exposure to delivery on neonatal outcome of very low birth weight infants. Am J Obstet Gynecol 2004;191:1409–13.
13. Peaceman AM, Bajaj K, Kumar P, Grobman W. The interval between a single course of antenatal steroids and delivery and its association with neonatal outcomes. Am J Obstet Gynecol 2005;193:1165–9.
14. Vermillion ST, Bland ML, Soper DE. Effectiveness of a rescue dose of antenatal betamethasone after an initial single course. Am J Obstet Gynecol 2001;185:1086–9.
15. Crowley PA. Antenatal corticosteroid therapy: a meta-analysis of the randomized trials, 1972 to 1994. Am J Obstet Gynecol 1995;173:322–35.