- Email: [email protected]
Does the clinical presentation of a prior preterm birth predict risk in a subsequent pregnancy?
Research
ajog.org
OBSTETRICS
Does the clinical presentation of a prior preterm birth predict risk in a subsequent pregnancy? Daphnie Drassinower, MD; Sarah G. Obican, MD; Zainab Siddiq, MS; Danielle Heller, BA; Cynthia Gyamfi-Bannerman, MD, MPH; Alexander M. Friedman, MD OBJECTIVE: The objective of the study was to determine whether risk
of recurrent preterm birth differs based on the clinical presentation of a prior spontaneous preterm birth (SPTB): advanced cervical dilatation (ACD), preterm premature rupture of membranes (PPROM), or preterm labor (PTL). STUDY DESIGN: This retrospective cohort study included singleton pregnancies from 2009 to 2014 complicated by a history of prior SPTB. Women were categorized based on the clinical presentation of their prior preterm delivery as having ACD, PPROM, or PTL. Risks for sonographic short cervical length and recurrent SPTB were compared between women based on the clinical presentation of their prior preterm birth. Log-linear regression was used to control for confounders. RESULTS: Of 522 patients included in this study, 96 (18.4%) had
prior ACD, 246 (47.1%) had prior PPROM, and 180 (34.5%) had
prior PTL. Recurrent PTB occurred in 55.2% of patients with a history of ACD compared with 27.2% of those with PPROM and 32.2% with PTL (P ¼ .001). The mean gestational age at delivery was significantly lower for those with a history of ACD (34.0 weeks) compared with women with prior PPROM (37.2 weeks) or PTL (37.0 weeks) (P ¼ .001). The lowest mean cervical length prior to 24 weeks was significantly shorter in patients with a history of advanced cervical dilation when compared with the other clinical presentations. CONCLUSION: Patients with a history of ACD are at an increased risk of having recurrent preterm birth and cervical shortening in a subsequent pregnancy compared with women with prior preterm birth associated PPROM or PTL.
Key words: advanced cervical dilation, recurrent preterm birth, recurrence risk
Cite this article as: Drassinower D, Obican SG, Siddiq Z, et al. Does the clinical presentation of a prior preterm birth predict risk in a subsequent pregnancy? Am J Obstet Gynecol 2015;213:686.e1-7.
S
pontaneous preterm birth is a leading cause of perinatal morbidity and mortality. Long-term sequelae of preterm birth include cerebral palsy, cognitive and developmental abnormalities, and chronic respiratory disease; increased risks for these conditions are particularly high when delivery occurs at very early gestational ages.1-3 Because From the Department of Obstetrics and Gynecology, Columbia University Medical Center, New York, NY. Received April 22, 2015; revised June 15, 2015; accepted July 18, 2015. The authors report no conflict of interest. Presented in poster format at the 34rd annual meeting of the Society for Maternal-Fetal Medicine, New Orleans, LA, Feb. 3-8, 2014. Corresponding author: Daphnie Drassinower, MD. [email protected] 0002-9378/$36.00 ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2015.07.029
of the neonatal risks associated with preterm birth, extensive research has focused on improving screening for preterm birth risk and optimizing the interventions that reduce preterm birth rates.3-6 Although the causes of preterm birth are only partially understood, preterm delivery is thought to occur by multiple mechanisms and may manifest in a variety of clinical presentations.7,8 Spontaneous preterm labor (PTL) with intact membranes, preterm premature rupture of membranes (PPROM), and advanced cervical dilatation (ACD) may all represent distinct pathways resulting in early delivery.7,8 The specific presentation of a prior preterm birth may be of clinical importance in a subsequent pregnancy, both in terms of guiding interventions and in terms of prognosticating risk. In particular, clinical management may differ for ACD; interventions such as cerclage are
686.e1 American Journal of Obstetrics & Gynecology NOVEMBER 2015
supported specifically for a history of this condition.9 However, there are few data comparing subsequent obstetric outcomes based on clinical presentation of prior preterm delivery, both in terms of subsequent preterm birth rates and whether related risk factors, such as short transvaginal sonographic cervical length, are present. Given the importance of elucidating preterm birth risk factors, both for guiding interventions and for patient surveillance and counseling, the purpose of this study was to compare subsequent pregnancy outcomes of women who had a prior preterm birth based on whether they had a history of PTL, PPROM, or ACD. We hypothesized that patients with a history of preterm delivery because of advanced cervical dilatation are at higher risk of recurrent preterm birth and cervical shortening in a subsequent pregnancy compared with those with a history of PPROM or PTL.
Obstetrics
ajog.org
TABLE 1
Demographic and clinical characteristics by prior preterm birth presentation Demographic
ACD (n [ 96)
PPROM (n [ 246)
PTL (n [ 180)
P value
Maternal age, y (SD)
33.0 (5.7)
32.3 (5.9)
32.0 (5.8)
.37
Parity, median (IQR)
2 (1-3)
2 (1-2)
2 (1-3)
< .01
Prior full-term births, median (IQR)
0 (0-1)
0 (0-1)
1 (0-1)
< .01
Prior preterm births, median (IQR)
1 (1-2)
1 (1-1)
1 (1-2)
< .01
Government-insured clinic, n (%)
34 (35.4)
128 (52.0)
89 (49.4)
Race/ethnicity, n (%)
.02 .04
White
25 (26.0)
40 (16.3)
34 (18.9)
African American
17 (17.7)
26 (10.5)
14 (7.8)
Hispanic
18 (18.8)
53 (21.5)
44 (24.4)
Other/unknown
36 (37.5)
127 (51.6)
88 (48.9)
24.4 (6.2)
28.1 (5.9)
29.5 (6.0)
< .01a
Mean GA of earliest PTB, wks (SD) History-indicated cerclage, n (%)
42 (43.8)
11 (4.5)
9 (5.0)
< .01
History of LEEP, CKC, or D&E, n (%)
16 (16.7)
26 (10.6)
14 (7.8)
.07
17OHP-C, n (%)
77 (80.2)
206 (83.7)
155 (86.1)
.60
Comparison of continuous variables was performed using an analysis of variance. Nonparametric data were compared using Kruskal-Wallis and are presented as medians and interquartile ranges (IQR). Categorical variables were compared using the c2 test. ACD, advanced cervical dilatation; CKC, cold knife cone; D&E, dilation and evacuation procedure; GA, gestational age; LEEP, loop electrosurgical excision procedure; 17OHP-C, 17-alpha-hydroxyprogesterone caproate; PPROM, preterm premature rupture of membranes; PTB, preterm birth; PTL, preterm labor. a
The P value represents the difference between the ACD group and both the PPROM and PTL groups; the P value for the Student t test comparing the mean GA at delivery of the earliest PTB between the PPROM and PTL groups was .02. All other P values represent the difference between the ACD group and both the PPROM and PTL groups; however, the PPROM and PTL groups were similar.
Drassinower. Type of spontaneous preterm birth and risk of recurrence. Am J Obstet Gynecol 2015.
M ATERIALS
AND
M ETHODS
This retrospective cohort study evaluated women with a current singleton pregnancy and an obstetric history significant for a prior spontaneous preterm birth. Patients were identified from an ultrasound database from a single tertiary center (Columbia University Medical Center, New York, NY). This database includes midtrimester transvaginal ultrasound cervical length measurements because cervical length screening is standard practice at our institution for all women with prior spontaneous preterm birth <37 weeks. Cervical length measurements take place every 2 weeks between 16 weeks onward or more frequently if short cervical length is detected. Patients were included if they received care between 2009 and 2014.
Patients with multiple gestations, major fetal anomalies, abdominal cerclage, physical examinationeindicated cerclage, missing delivery data, a history of iatrogenic preterm birth, or evidence of placental abruption at the time of delivery or by placental pathology were excluded. Additionally, patients were excluded if documentation of the clinical presentation of the prior preterm birth was unclear and thus could not be categorized. This included patients who presented for evaluation of PPROM but were found have cervical dilation because we considered the potential for misclassification among these patients. Approval for this study from the Columbia University Institutional Review Board was obtained. Review of the electronic medical record was performed to obtain individual
Research
demographic, obstetric, and medical information. Documentation of receipt of vaginal or intramuscular progesterone administration was abstracted, as was whether a patient received a history or ultrasound-indicated cerclage. Women were categorized based on whether their prior preterm birth presentation was most compatible with 1 of 3 specific diagnoses: advanced cervical dilation, PPROM, or preterm labor. This designation was based on at least one of the following sources: review of inpatient hospital records for the index preterm birth, obstetric history obtained at a first prenatal visit, and/or a maternal-fetal medicine consultation. Patients were categorized as having a history of advanced cervical dilation if they reported that in the absence of contractions they were found to have a dilated cervix on examination for spotting, increased pelvic pressure, or increased discharge or were otherwise incidentally diagnosed with cervical dilatation. Patients were considered to have a history of PPROM if they reported preterm rupture of membranes in the absence of contractions. Patients were allocated to the preterm labor group if they reported presenting for evaluation for painful contractions or reported painful contractions with concomitant rupture of membranes at the time of presentation to labor and delivery. Whether patients did or did not have additional prior term deliveries was also abstracted. The shortest cervical length measurement between 16 and 24 weeks’ gestation was determined for each patient. Based on our institution’s protocol, cervical length measurements take place every 2 weeks for women with a history of prior preterm birth or more frequently if short cervical length is detected. Each transvaginal cervical length assessment was obtained using the technique published by Iams et al,6 wherein sonographers record the shortest cervical length for each examination that clearly displays the internal and external os with equivalent thickness of the anterior and posterior cervix. Sonographer training at our site was performed in conjunction with a
NOVEMBER 2015 American Journal of Obstetrics & Gynecology
686.e2
Research
Obstetrics
Maternal-Fetal Medicine Units Network study by Grobman et al10 and involved a didactics series and image submission. Additionally, the obstetrical ultrasound guidelines at Columbia University Medical Center require 3 measurements of the cervix, including at least 1 assessment while the patient performs the Valsalva maneuver. The shortest of the 3 cervical length values is reported clinically, and this measurement was recorded into our database for each visit. Sonographic assessments were performed for clinical use; therefore, the gestational age at the initiation of cervical length surveillance as well as the frequency of surveillance was variable between patients. During the study period, ultrasound machines from multiple manufacturers were used to collect clinical data including General Electric (Fairfield, CT), Philips (Andover, MA), Medison (Seoul, South Korea), and Acuson (Mountain View, CA) ultrasound machines. Transvaginal probes (5e9 mHz) were used to obtain cervical length measurement. The primary outcome for this study was preterm delivery <37 weeks. Secondary outcomes included mean gestational age at delivery, preterm delivery <34 weeks, preterm delivery <28 weeks, shortest mean cervical length, and the presence of a short cervix <25 mm prior to 24 weeks. Categorical variables were analyzed with the c2 test, whereas continuous variables were compared using an analysis of variance. A Kruskal-Wallis test was used to compare the median differences for nonparametric data. Pregnancy duration was analyzed with KaplanMeier curves and the log-rank test. All analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC). A value of P < .05 was considered statistically significant. Adjusted analyses were performed using log linear regression and included the following factors: age, race/ ethnicity, the number of prior full-term deliveries, the number of prior preterm deliveries, gestational age of last preterm delivery, and a history of cervical procedures (loop electrosurgical excision procedure, cold-knife cone, and/or
ajog.org
TABLE 2
Outcomes by presentation of prior preterm birth Prior ACD Prior PPROM Prior PTL (n [ 96) (n [ 246) (n [ 180) P value
Outcome
All patients (including patients with history-indicated cerclage) Mean GA at delivery, wks (SD)
34.0 (5.8)
37.2 (3.8) 37.0 (3.5)
< .01
Preterm delivery <37, n (%)
53 (55.2)
67 (27.2)
58 (32.2)
< .01
Preterm delivery <34 wks, n (%)
34 (35.4)
23 (9.4)
21 (11.7)
< .01
Preterm delivery <28 wks, n (%)
17 (17.7)
11 (4.5)
6 (3.3)
< .01
Mean shortest CL prior to 24 wks, mm (SD)
24.0 (13)
35.3 (11)
34.7 (12)
< .01
36 (14.6)
28 (15.6)
< .01
Short cervix prior to 24 wks, n, (%)a 48 (50)
Subgroup analysis excluding patients with history or ultrasound-indicated cerclage Mean GA at delivery, wks (SD)
33.6 (6.0)
37.3 (3.4) 37.3 (3.0)
< .01
Preterm delivery <37 wks, n (%)
21 (60)
60 (27.4)
48 (29.5)
< .01
Preterm delivery <34 wks, n (%)
14 (40)
19 (8.7)
15 (9.2)
< .01
7 (2.8)
3 (1.8)
< .01
Preterm delivery <28 wks, n (%)
6 (17.1)
Mean CL prior to 24 wks, mm (SD) Short cervix prior to 24 wks, n (%)
a
24.6 (13.0)
37.0 (9.9) 36.0 (11.2) < .01
16 (46)
16 (7)
15 (9)
< .01
ACD, advanced cervical dilation; CL, cervical length; GA, gestational age; PPROM, preterm premature rupture of membranes; PTL, preterm labor. a
Short cervix was defined as a transvaginal sonographic cervical length of <25 mm. P values represent the difference between the ACD group and both the PPROM and PTL groups, but the PPROM and PTL groups were similar.
Drassinower. Type of spontaneous preterm birth and risk of recurrence. Am J Obstet Gynecol 2015.
prior dilation and evacuation procedure). We chose a log linear model over a logistic regression model because the former permits estimating the true relative risks instead of odds ratios. Because the design of this study is a retrospective cohort study, we estimated the relative risks directly using a log linear model.11
R ESULTS Of 595 cases potentially eligible for inclusion, 44 patients were excluded because the etiology of the prior preterm birth could not be established, 18 patients were excluded because of missing delivery data, 6 patients were excluded because of a history of suspected placental abruption, and 5 patients were excluded because of major fetal anomalies, resulting in 522 patients remaining in the analysis. Of patients included in the analysis, 96 (18.4%) had a history of advanced cervical dilation, 246 (47.1%) had a history of PPROM, and
686.e3 American Journal of Obstetrics & Gynecology NOVEMBER 2015
180 (34.5%) had a history of preterm labor. Demographic characteristics and risk factors for preterm delivery are summarized in Table 1. Race and insurance status differed significantly between the groups. Additionally, patients with a history of preterm labor were significantly more likely to have had a full-term birth, and patients with history of advanced cervical dilation were more likely to have had multiple prior preterm births and more likely to be treated with a history-indicated cerclage. The mean gestational age of the prior preterm delivery was significantly earlier in the prior ACD group compared with the prior PPROM and PTL groups. Use of 17-alpha-hydroxyprogesterone caproate was similar among all three groups. Table 2 demonstrates primary and secondary outcomes. Women with a history of ACD were significantly more likely to have recurrent preterm
Obstetrics
ajog.org
Entire cohort
Patients with cerclage excluded
displays a Kaplan-Meier curve that demonstrates that time to delivery (survival) was significantly decreased for women with prior ACD.
Unadjusted relative risk (95% CI)
Unadjusted relative risk (95% CI)
C OMMENT
TABLE 3
Unadjusted relative risks for primary and secondary outcomes
Outcome Delivery <37 wks Prior PPROM
Referent
Referent
Prior ACD
2.0 (1.5e2.7)
2.2 (1.5e3.1)a
Prior PTL
1.2 (0.9e1.6)
1.1 (0.8e1.5)
a
Prior PPROM
Referent
Referent
Prior ACD
3.8 (2.4e6.1)a
4.6 (2.6e8.3)a
Prior PTL
1.3 (0.7e2.2)
0.6 (0.2e2.0)
Referent
Referent
Delivery <34 wks
Delivery <28 wks Prior PPROM
a
Prior ACD
4.0 (1.9e8.1)
5.3 (1.9e15.0)a
Prior PTL
0.7 (0.3e2.0)
0.6 (0.2e2.2)
Prior PPROM
Referent
Referent
Prior ACD
3.4 (2.4e4.9)a
5.3 (3.1e9.1)a
Prior PTL
1.1 (0.7e1.7)
1.1 (0.6e2.1)
Short cervix
ACD, advanced cervical dilation; PPROM, preterm premature rupture of membranes; PTL, preterm labor. a
Research
Statistically significant difference.
Drassinower. Type of spontaneous preterm birth and risk of recurrence. Am J Obstet Gynecol 2015.
delivery <37 weeks compared with women with a history of PPROM or PTL (55.2% vs 27.2% vs 32.2%, respectively; P < .01). The mean gestational age at delivery for women with prior ACD was 34.0 weeks compared with 37.2 for women with a history of PPROM and 37.0 weeks for women with a history of preterm labor (P < .01). Delivery at <34 weeks and 28 weeks was also significantly higher in the prior ACD group. The mean cervical length prior to 24 weeks was significantly shorter in patients with a history preterm delivery because of advanced cervical dilation when compared with those with a history of preterm birth because of PPROM or preterm labor; these patients were also more likely to have a short cervix <25 mm prior to 24 weeks. A subgroup analysis was performed excluding patients treated with cerclage because this treatment may have affected the obstetric outcome (Table 2). For all of the
primary and secondary outcomes, results were similar to the initial analysis. In an unadjusted analysis, patients with a history of ACD had a relative risk for preterm birth <37, <34, and <28 weeks of 2.0, 3.8, and 4.0 respectively, with prior PPROM as the referent (Table 3). This elevated risk remained statistically significant in a sensitivity analysis eliminating patients with cervical cerclage. In a multivariate analysis including such factors as the number of prior preterm births, the gestational age of prior preterm birth, and demographic factors, ACD continued to be associated with significantly elevated risk for recurrent preterm birth at <37 weeks, <34 weeks, and <28 weeks (Table 4). In contrast, the most important predictors of short cervical length were African American and Hispanic race, and ACD was not associated with significantly increased risk for short cervix <25 mm at <24 weeks. The Figure
In this cohort, the overall rate of recurrent preterm delivery was 34%, consistent with previously published data.2,5 Patients with a history of advanced cervical dilation were at an increased risk for recurrent preterm birth compared with those with a history of PPROM or PTL, a relationship that persisted in multivariable analysis adjusting for several demographic and obstetric risk factors. A subgroup analysis excluding patients with a history-indicated or ultrasoundindicated cerclage revealed similar risk between preterm birth risk and prior ACD. Although it is recognized that different clinical preterm birth presentations may represent different syndromes and pathogenic pathways,7,8 the impact of the clinical presentation of the prior spontaneous PTB on the risk of recurrence has not been previously addressed in the literature. Furthermore, studies on interventions for prevention of recurrent spontaneous preterm birth have generally not differentiated patients based on the initial presentation in a prior pregnancy.5 This analysis had several findings consistent with previously reported data.2 It demonstrated that the gestational age of the prior preterm delivery, the number of prior preterm deliveries, and black race were significant risk factors for recurrent preterm birth. This analysis also found that short cervix occurred in a higher proportion of patients with a history of advanced cervical dilation, although in an adjusted analysis, this association was not significant when controlling for race and the number of prior preterm births. We offer 2 possible explanations for this finding. Patients with a history of advanced cervical dilation were more likely to get a cerclage. It is possible that cerclage prevents cervical shortening by interfering with the natural progression of cervical insufficiency. Another possible explanation is that patients with
NOVEMBER 2015 American Journal of Obstetrics & Gynecology
686.e4
Research
Obstetrics
ajog.org
TABLE 4
Log-linear model for primary and secondary outcomes Outcome
Preterm delivery <37 wks
Preterm delivery <34 wks
Preterm delivery <28 wks
Short cervix prior to 24 wks
Referent
Referent
Referent
Referent
Obstetric history Prior PPROM
a
2.77 (1.58e4.84)
a
2.33 (1.02e5.35)
a
1.11 (0.69e1.74)
Prior ACD
1.81 (1.32e2.48)
Prior PTL
1.19 (0.89e1.59)
1.21 (0.70e2.11)
0.71 (0.27e1.92)
0.40 (0.20e0.80)a
Prior full-term birthb
0.88 (0.76e1.03)
0.99 (0.80e1.23)
1.08 (0.79e1.48)
0.99 (0.70e1.40)
White
Referent
Referent
Referent
Referent
African American
1.04 (0.67e1.63)
0.88 (0.45e1.72)
1.50 (0.52e4.34)
2.49 (1.09e5.68)a
Hispanic
1.15 (0.76e1.73)
1.02 (0.52e2.0)
0.93 (0.29e3.0)
2.74 (1.30e5.76)a
Government-insured clinic
0.83 (0.63e1.10)
0.69 (0.42e1.15)
1.02 (0.50e2.08)
0.96 (0.50e1.84)
Cervical procedure
0.86 (0.59e1.27)
0.82 (0.46e1.48)
0.55 (0.22e1.38)
1.18 (0.52e1.84)
History-indicated cerclage
0.99 (0.71e1.39)
0.98 (0.56e1.70)
1.01 (0.47e2.19)
0.62 (0.30e1.31)
Race
Log-linear regression model was also adjusted for the following continuous variables: number of prior preterm births and gestational age of prior preterm delivery. ACD, advanced cervical dilation; PPROM, preterm premature rupture of membranes; PTL, preterm labor. a
Statistically significant difference; b Prior full-term birth indicates the number of prior full-term births (continuous variable).
Drassinower. Type of spontaneous preterm birth and risk of recurrence. Am J Obstet Gynecol 2015.
a history of ACD delivered at earlier gestational age or may have had ACD in the index pregnancy and therefore stopped receiving cervical length measurements. Cervical cerclage is an intervention that has been shown to be effective in preventing preterm delivery in select patient populations.12-14 However, there are currently no studies specifically addressing the efficacy of cerclage in patients with a history of advanced cervical dilation. Despite this lack of evidence, many providers offer a historyindicated cerclage to patients suspected to have had prior ACD (based on a history of preterm delivery because of advanced cervical dilation), an approach supported by the American Congress of Obstetricians and Gynecologists.9 An alternate strategy is to perform serial cervical length surveillance with the addition of an ultrasound-indicated cerclage in the presence of cervical shortening <25 mm.15 Our data show that patients with a history of advanced cervical dilation are at significant risk of recurrent PTB mediated by cervical shortening,
particularly in the setting of risk factors such as African American race and multiple prior preterm births. We suggest that it is therefore reasonable to consider a history-indicated cerclage for women with a history of advanced cervical dilation who may be at particularly high risk for preterm birth. Furthermore, future research is warranted to determine the relative benefits of prophylactic versus ultrasound-indicated cerclage for extremely high-risk women. There are several factors to consider in interpreting the validity of this study’s findings. All patients included in this study had a history of prior preterm birth, and the vast majority of patients received progesterone to reduce the risk of recurrent preterm birth. Although the analysis of the modulating effects of this intervention is beyond the scope of this analysis because of the ubiquity of progesterone use in preterm birth prevention strategies, this will likely be a limitation of subsequent preterm birth research as well and makes our results generalizable. Whereas our analysis was not powered to determine the effect of cerclage on outcomes, a sensitivity
686.e5 American Journal of Obstetrics & Gynecology NOVEMBER 2015
analysis excluding patients with cerclage yielded similar results, suggestive that our findings are applicable to populations with and without this intervention. Other limitations in this study are inherent to its retrospective nature. Cervical length measurements were performed for clinical purposes, and therefore, there is heterogeneity in the number and frequency of cervical length measurements performed in each patient. Additionally, patients with ACD who disproportionately delivered at earlier gestational ages no longer contributed cervical length measurements, and women remaining pregnant could provide a skewed representation of cervical length throughout pregnancy; this makes our findings potentially biased toward underestimating the prevalence of a short cervix in women with a history of ACD. However, given the large proportion of women who developed a short cervix with a history ACD determination of a relationship was still possible. Another limitation of this study is that the etiology of the prior preterm birth
Obstetrics
ajog.org
FIGURE
Association between prior preterm delivery and subsequent pregnancy
Research
in chorioamnionitis, PPROM, or PTL. This potential for misclassification of diagnosis could lead to biased results. Although we attempted to minimize the risk by careful review of medical records and excluded cases if the etiology of the prior preterm birth could not be reliably determined (for example, by excluding patients who presented for evaluation of PPROM but were found to be dilated on examination), given the complexity of diagnosis we cannot exclude it completely. The difficulty in differentiating the clinical presentation that preceded the prior preterm delivery may be a reason that its impact on the risk of recurrent preterm birth has not been elucidated further in research literature. In conclusion, our study showed that the clinical presentation that led to a prior preterm birth as assessed by patient history is a highly significant predictor of recurrent preterm delivery. We propose that future investigation and riskreducing interventions should consider these important factors in optimizing research strategies and patient care. -
REFERENCES
Kaplan-Meier curve showing the association between a prior preterm delivery because of advanced cervical dilatation and a shorter gestation in the subsequent pregnancy compared with PPROM and preterm labor. Log rank rest, P ¼ .001. ACD, advanced cervical dilation; PPROM, preterm premature rupture of membranes; PTL, preterm labor. Drassinower. Type of spontaneous preterm birth and risk of recurrence. Am J Obstet Gynecol 2015.
was, in part, self-reported. In some cases, the prior preterm birth occurred at our institution in which case the medical records for that delivery were available for review and the etiology was established with a high degree of certainty. In other cases, our classification on the etiology of the prior preterm birth was based on a review of the initial obstetrical visits, which in some situations relied on a patient’s report of her history. Although this is a limitation of our study,
it also makes our results generalizable because often providers must base their approach to a current pregnancy on the patient’s report of a previous pregnancy when review of medical records is not possible. Finally, we recognize that diagnosing cervical insufficiency is challenging because advanced dilation is often asymptomatic and may go unrecognized for an indefinite period of time, leading to exposed membranes that may result
1. Lockwood CJ, Kuczynski E. Risk stratification and pathological mechanisms in preterm delivery. Paediatr Perinat Epidemiol 2001;15(Suppl) 2:78-89. 2. Spong CY. Prediction and prevention of recurrent spontaneous preterm birth. Obstetr Gynecol 2007;110:405-15. 3. Mercer BM, Goldenberg RL, Moawad AH, et al. The preterm prediction study: effect of gestational age and cause of preterm birth on subsequent obstetric outcome. National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network. Am J Obstet Gynecol 1999;181:1216-21. 4. Fonseca EB, Celik E, Parra M, Singh M, Nicolaides KH; Fetal Medicine Foundation Second Trimester Screening Group. Progesterone and the risk of preterm birth among women with a short cervix. N Engl J Med 2007;357:462-9. 5. Meis PJ, Klebanoff M, Thom E, et al. Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate. N Engl J Med 2003;348:2379-85. 6. Iams JD, Goldenberg RL, Meis PJ, 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.
NOVEMBER 2015 American Journal of Obstetrics & Gynecology
686.e6
Research
Obstetrics
7. Creasy RK, Resnik R, Iams JD. Creasy and Resnik’s maternal-fetal medicine: principles and practice, 6th ed. Philadelphia, PA: Saunders/ Elsevier; 2009. 8. Iams JD, Cebrik D, Lynch C, Behrendt N, Das A. The rate of cervical change and the phenotype of spontaneous preterm birth. Am J Obstet Gynecol 2011;205:130.e1-6. 9. American College of Obstetricsw and Gynecologists. Cerclage for the management of cervical insufficiency. ACOG Practice bulletin no. 142. Obstet Gynecol 2014;123:372-9. 10. Grobman WA, Thom EA, Spong CY, et al. 17-alpha-hydroxyprogesterone caproate to
ajog.org prevent prematurity in nulliparas with cervical length less than 30 mm. Am J Obstet Gynecol 2012;207:390.e1-8. 11. Spiegelman D, Hertzmark E. Easy SAS calculations for risk or prevalence ratios and differences. Am J Epidemiol 2005;162: 199-200. 12. Berghella V, Rafael TJ, Szychowski JM, Rust OA, Owen J. Cerclage for short cervix on ultrasonography in women with singleton gestations and previous preterm birth: a metaanalysis. Obstet Gynecol 2011;117:663-71. 13. Owen J, Hankins G, Iams JD, et al. Multicenter randomized trial of cerclage for
686.e7 American Journal of Obstetrics & Gynecology NOVEMBER 2015
preterm birth prevention in high-risk women with shortened midtrimester cervical length. Am J Obstet Gynecol 2009;201:375.e1-8. 14. Final report of the Medical Research Council/Royal College of Obstetricians and Gynaecologists multicentre randomised trial of cervical cerclage. MRC/RCOG Working Party on Cervical Cerclage. Br J Obstet Gynaecol 1993;100:516-23. 15. Berghella V, Mackeen AD. Cervical length screening with ultrasound-indicated cerclage compared with history-indicated cerclage for prevention of preterm birth: a meta-analysis. Obstet Gynecol 2011;118:148-55.
ajog.org
OBSTETRICS
Does the clinical presentation of a prior preterm birth predict risk in a subsequent pregnancy? Daphnie Drassinower, MD; Sarah G. Obican, MD; Zainab Siddiq, MS; Danielle Heller, BA; Cynthia Gyamfi-Bannerman, MD, MPH; Alexander M. Friedman, MD OBJECTIVE: The objective of the study was to determine whether risk
of recurrent preterm birth differs based on the clinical presentation of a prior spontaneous preterm birth (SPTB): advanced cervical dilatation (ACD), preterm premature rupture of membranes (PPROM), or preterm labor (PTL). STUDY DESIGN: This retrospective cohort study included singleton pregnancies from 2009 to 2014 complicated by a history of prior SPTB. Women were categorized based on the clinical presentation of their prior preterm delivery as having ACD, PPROM, or PTL. Risks for sonographic short cervical length and recurrent SPTB were compared between women based on the clinical presentation of their prior preterm birth. Log-linear regression was used to control for confounders. RESULTS: Of 522 patients included in this study, 96 (18.4%) had
prior ACD, 246 (47.1%) had prior PPROM, and 180 (34.5%) had
prior PTL. Recurrent PTB occurred in 55.2% of patients with a history of ACD compared with 27.2% of those with PPROM and 32.2% with PTL (P ¼ .001). The mean gestational age at delivery was significantly lower for those with a history of ACD (34.0 weeks) compared with women with prior PPROM (37.2 weeks) or PTL (37.0 weeks) (P ¼ .001). The lowest mean cervical length prior to 24 weeks was significantly shorter in patients with a history of advanced cervical dilation when compared with the other clinical presentations. CONCLUSION: Patients with a history of ACD are at an increased risk of having recurrent preterm birth and cervical shortening in a subsequent pregnancy compared with women with prior preterm birth associated PPROM or PTL.
Key words: advanced cervical dilation, recurrent preterm birth, recurrence risk
Cite this article as: Drassinower D, Obican SG, Siddiq Z, et al. Does the clinical presentation of a prior preterm birth predict risk in a subsequent pregnancy? Am J Obstet Gynecol 2015;213:686.e1-7.
S
pontaneous preterm birth is a leading cause of perinatal morbidity and mortality. Long-term sequelae of preterm birth include cerebral palsy, cognitive and developmental abnormalities, and chronic respiratory disease; increased risks for these conditions are particularly high when delivery occurs at very early gestational ages.1-3 Because From the Department of Obstetrics and Gynecology, Columbia University Medical Center, New York, NY. Received April 22, 2015; revised June 15, 2015; accepted July 18, 2015. The authors report no conflict of interest. Presented in poster format at the 34rd annual meeting of the Society for Maternal-Fetal Medicine, New Orleans, LA, Feb. 3-8, 2014. Corresponding author: Daphnie Drassinower, MD. [email protected] 0002-9378/$36.00 ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2015.07.029
of the neonatal risks associated with preterm birth, extensive research has focused on improving screening for preterm birth risk and optimizing the interventions that reduce preterm birth rates.3-6 Although the causes of preterm birth are only partially understood, preterm delivery is thought to occur by multiple mechanisms and may manifest in a variety of clinical presentations.7,8 Spontaneous preterm labor (PTL) with intact membranes, preterm premature rupture of membranes (PPROM), and advanced cervical dilatation (ACD) may all represent distinct pathways resulting in early delivery.7,8 The specific presentation of a prior preterm birth may be of clinical importance in a subsequent pregnancy, both in terms of guiding interventions and in terms of prognosticating risk. In particular, clinical management may differ for ACD; interventions such as cerclage are
686.e1 American Journal of Obstetrics & Gynecology NOVEMBER 2015
supported specifically for a history of this condition.9 However, there are few data comparing subsequent obstetric outcomes based on clinical presentation of prior preterm delivery, both in terms of subsequent preterm birth rates and whether related risk factors, such as short transvaginal sonographic cervical length, are present. Given the importance of elucidating preterm birth risk factors, both for guiding interventions and for patient surveillance and counseling, the purpose of this study was to compare subsequent pregnancy outcomes of women who had a prior preterm birth based on whether they had a history of PTL, PPROM, or ACD. We hypothesized that patients with a history of preterm delivery because of advanced cervical dilatation are at higher risk of recurrent preterm birth and cervical shortening in a subsequent pregnancy compared with those with a history of PPROM or PTL.
Obstetrics
ajog.org
TABLE 1
Demographic and clinical characteristics by prior preterm birth presentation Demographic
ACD (n [ 96)
PPROM (n [ 246)
PTL (n [ 180)
P value
Maternal age, y (SD)
33.0 (5.7)
32.3 (5.9)
32.0 (5.8)
.37
Parity, median (IQR)
2 (1-3)
2 (1-2)
2 (1-3)
< .01
Prior full-term births, median (IQR)
0 (0-1)
0 (0-1)
1 (0-1)
< .01
Prior preterm births, median (IQR)
1 (1-2)
1 (1-1)
1 (1-2)
< .01
Government-insured clinic, n (%)
34 (35.4)
128 (52.0)
89 (49.4)
Race/ethnicity, n (%)
.02 .04
White
25 (26.0)
40 (16.3)
34 (18.9)
African American
17 (17.7)
26 (10.5)
14 (7.8)
Hispanic
18 (18.8)
53 (21.5)
44 (24.4)
Other/unknown
36 (37.5)
127 (51.6)
88 (48.9)
24.4 (6.2)
28.1 (5.9)
29.5 (6.0)
< .01a
Mean GA of earliest PTB, wks (SD) History-indicated cerclage, n (%)
42 (43.8)
11 (4.5)
9 (5.0)
< .01
History of LEEP, CKC, or D&E, n (%)
16 (16.7)
26 (10.6)
14 (7.8)
.07
17OHP-C, n (%)
77 (80.2)
206 (83.7)
155 (86.1)
.60
Comparison of continuous variables was performed using an analysis of variance. Nonparametric data were compared using Kruskal-Wallis and are presented as medians and interquartile ranges (IQR). Categorical variables were compared using the c2 test. ACD, advanced cervical dilatation; CKC, cold knife cone; D&E, dilation and evacuation procedure; GA, gestational age; LEEP, loop electrosurgical excision procedure; 17OHP-C, 17-alpha-hydroxyprogesterone caproate; PPROM, preterm premature rupture of membranes; PTB, preterm birth; PTL, preterm labor. a
The P value represents the difference between the ACD group and both the PPROM and PTL groups; the P value for the Student t test comparing the mean GA at delivery of the earliest PTB between the PPROM and PTL groups was .02. All other P values represent the difference between the ACD group and both the PPROM and PTL groups; however, the PPROM and PTL groups were similar.
Drassinower. Type of spontaneous preterm birth and risk of recurrence. Am J Obstet Gynecol 2015.
M ATERIALS
AND
M ETHODS
This retrospective cohort study evaluated women with a current singleton pregnancy and an obstetric history significant for a prior spontaneous preterm birth. Patients were identified from an ultrasound database from a single tertiary center (Columbia University Medical Center, New York, NY). This database includes midtrimester transvaginal ultrasound cervical length measurements because cervical length screening is standard practice at our institution for all women with prior spontaneous preterm birth <37 weeks. Cervical length measurements take place every 2 weeks between 16 weeks onward or more frequently if short cervical length is detected. Patients were included if they received care between 2009 and 2014.
Patients with multiple gestations, major fetal anomalies, abdominal cerclage, physical examinationeindicated cerclage, missing delivery data, a history of iatrogenic preterm birth, or evidence of placental abruption at the time of delivery or by placental pathology were excluded. Additionally, patients were excluded if documentation of the clinical presentation of the prior preterm birth was unclear and thus could not be categorized. This included patients who presented for evaluation of PPROM but were found have cervical dilation because we considered the potential for misclassification among these patients. Approval for this study from the Columbia University Institutional Review Board was obtained. Review of the electronic medical record was performed to obtain individual
Research
demographic, obstetric, and medical information. Documentation of receipt of vaginal or intramuscular progesterone administration was abstracted, as was whether a patient received a history or ultrasound-indicated cerclage. Women were categorized based on whether their prior preterm birth presentation was most compatible with 1 of 3 specific diagnoses: advanced cervical dilation, PPROM, or preterm labor. This designation was based on at least one of the following sources: review of inpatient hospital records for the index preterm birth, obstetric history obtained at a first prenatal visit, and/or a maternal-fetal medicine consultation. Patients were categorized as having a history of advanced cervical dilation if they reported that in the absence of contractions they were found to have a dilated cervix on examination for spotting, increased pelvic pressure, or increased discharge or were otherwise incidentally diagnosed with cervical dilatation. Patients were considered to have a history of PPROM if they reported preterm rupture of membranes in the absence of contractions. Patients were allocated to the preterm labor group if they reported presenting for evaluation for painful contractions or reported painful contractions with concomitant rupture of membranes at the time of presentation to labor and delivery. Whether patients did or did not have additional prior term deliveries was also abstracted. The shortest cervical length measurement between 16 and 24 weeks’ gestation was determined for each patient. Based on our institution’s protocol, cervical length measurements take place every 2 weeks for women with a history of prior preterm birth or more frequently if short cervical length is detected. Each transvaginal cervical length assessment was obtained using the technique published by Iams et al,6 wherein sonographers record the shortest cervical length for each examination that clearly displays the internal and external os with equivalent thickness of the anterior and posterior cervix. Sonographer training at our site was performed in conjunction with a
NOVEMBER 2015 American Journal of Obstetrics & Gynecology
686.e2
Research
Obstetrics
Maternal-Fetal Medicine Units Network study by Grobman et al10 and involved a didactics series and image submission. Additionally, the obstetrical ultrasound guidelines at Columbia University Medical Center require 3 measurements of the cervix, including at least 1 assessment while the patient performs the Valsalva maneuver. The shortest of the 3 cervical length values is reported clinically, and this measurement was recorded into our database for each visit. Sonographic assessments were performed for clinical use; therefore, the gestational age at the initiation of cervical length surveillance as well as the frequency of surveillance was variable between patients. During the study period, ultrasound machines from multiple manufacturers were used to collect clinical data including General Electric (Fairfield, CT), Philips (Andover, MA), Medison (Seoul, South Korea), and Acuson (Mountain View, CA) ultrasound machines. Transvaginal probes (5e9 mHz) were used to obtain cervical length measurement. The primary outcome for this study was preterm delivery <37 weeks. Secondary outcomes included mean gestational age at delivery, preterm delivery <34 weeks, preterm delivery <28 weeks, shortest mean cervical length, and the presence of a short cervix <25 mm prior to 24 weeks. Categorical variables were analyzed with the c2 test, whereas continuous variables were compared using an analysis of variance. A Kruskal-Wallis test was used to compare the median differences for nonparametric data. Pregnancy duration was analyzed with KaplanMeier curves and the log-rank test. All analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC). A value of P < .05 was considered statistically significant. Adjusted analyses were performed using log linear regression and included the following factors: age, race/ ethnicity, the number of prior full-term deliveries, the number of prior preterm deliveries, gestational age of last preterm delivery, and a history of cervical procedures (loop electrosurgical excision procedure, cold-knife cone, and/or
ajog.org
TABLE 2
Outcomes by presentation of prior preterm birth Prior ACD Prior PPROM Prior PTL (n [ 96) (n [ 246) (n [ 180) P value
Outcome
All patients (including patients with history-indicated cerclage) Mean GA at delivery, wks (SD)
34.0 (5.8)
37.2 (3.8) 37.0 (3.5)
< .01
Preterm delivery <37, n (%)
53 (55.2)
67 (27.2)
58 (32.2)
< .01
Preterm delivery <34 wks, n (%)
34 (35.4)
23 (9.4)
21 (11.7)
< .01
Preterm delivery <28 wks, n (%)
17 (17.7)
11 (4.5)
6 (3.3)
< .01
Mean shortest CL prior to 24 wks, mm (SD)
24.0 (13)
35.3 (11)
34.7 (12)
< .01
36 (14.6)
28 (15.6)
< .01
Short cervix prior to 24 wks, n, (%)a 48 (50)
Subgroup analysis excluding patients with history or ultrasound-indicated cerclage Mean GA at delivery, wks (SD)
33.6 (6.0)
37.3 (3.4) 37.3 (3.0)
< .01
Preterm delivery <37 wks, n (%)
21 (60)
60 (27.4)
48 (29.5)
< .01
Preterm delivery <34 wks, n (%)
14 (40)
19 (8.7)
15 (9.2)
< .01
7 (2.8)
3 (1.8)
< .01
Preterm delivery <28 wks, n (%)
6 (17.1)
Mean CL prior to 24 wks, mm (SD) Short cervix prior to 24 wks, n (%)
a
24.6 (13.0)
37.0 (9.9) 36.0 (11.2) < .01
16 (46)
16 (7)
15 (9)
< .01
ACD, advanced cervical dilation; CL, cervical length; GA, gestational age; PPROM, preterm premature rupture of membranes; PTL, preterm labor. a
Short cervix was defined as a transvaginal sonographic cervical length of <25 mm. P values represent the difference between the ACD group and both the PPROM and PTL groups, but the PPROM and PTL groups were similar.
Drassinower. Type of spontaneous preterm birth and risk of recurrence. Am J Obstet Gynecol 2015.
prior dilation and evacuation procedure). We chose a log linear model over a logistic regression model because the former permits estimating the true relative risks instead of odds ratios. Because the design of this study is a retrospective cohort study, we estimated the relative risks directly using a log linear model.11
R ESULTS Of 595 cases potentially eligible for inclusion, 44 patients were excluded because the etiology of the prior preterm birth could not be established, 18 patients were excluded because of missing delivery data, 6 patients were excluded because of a history of suspected placental abruption, and 5 patients were excluded because of major fetal anomalies, resulting in 522 patients remaining in the analysis. Of patients included in the analysis, 96 (18.4%) had a history of advanced cervical dilation, 246 (47.1%) had a history of PPROM, and
686.e3 American Journal of Obstetrics & Gynecology NOVEMBER 2015
180 (34.5%) had a history of preterm labor. Demographic characteristics and risk factors for preterm delivery are summarized in Table 1. Race and insurance status differed significantly between the groups. Additionally, patients with a history of preterm labor were significantly more likely to have had a full-term birth, and patients with history of advanced cervical dilation were more likely to have had multiple prior preterm births and more likely to be treated with a history-indicated cerclage. The mean gestational age of the prior preterm delivery was significantly earlier in the prior ACD group compared with the prior PPROM and PTL groups. Use of 17-alpha-hydroxyprogesterone caproate was similar among all three groups. Table 2 demonstrates primary and secondary outcomes. Women with a history of ACD were significantly more likely to have recurrent preterm
Obstetrics
ajog.org
Entire cohort
Patients with cerclage excluded
displays a Kaplan-Meier curve that demonstrates that time to delivery (survival) was significantly decreased for women with prior ACD.
Unadjusted relative risk (95% CI)
Unadjusted relative risk (95% CI)
C OMMENT
TABLE 3
Unadjusted relative risks for primary and secondary outcomes
Outcome Delivery <37 wks Prior PPROM
Referent
Referent
Prior ACD
2.0 (1.5e2.7)
2.2 (1.5e3.1)a
Prior PTL
1.2 (0.9e1.6)
1.1 (0.8e1.5)
a
Prior PPROM
Referent
Referent
Prior ACD
3.8 (2.4e6.1)a
4.6 (2.6e8.3)a
Prior PTL
1.3 (0.7e2.2)
0.6 (0.2e2.0)
Referent
Referent
Delivery <34 wks
Delivery <28 wks Prior PPROM
a
Prior ACD
4.0 (1.9e8.1)
5.3 (1.9e15.0)a
Prior PTL
0.7 (0.3e2.0)
0.6 (0.2e2.2)
Prior PPROM
Referent
Referent
Prior ACD
3.4 (2.4e4.9)a
5.3 (3.1e9.1)a
Prior PTL
1.1 (0.7e1.7)
1.1 (0.6e2.1)
Short cervix
ACD, advanced cervical dilation; PPROM, preterm premature rupture of membranes; PTL, preterm labor. a
Research
Statistically significant difference.
Drassinower. Type of spontaneous preterm birth and risk of recurrence. Am J Obstet Gynecol 2015.
delivery <37 weeks compared with women with a history of PPROM or PTL (55.2% vs 27.2% vs 32.2%, respectively; P < .01). The mean gestational age at delivery for women with prior ACD was 34.0 weeks compared with 37.2 for women with a history of PPROM and 37.0 weeks for women with a history of preterm labor (P < .01). Delivery at <34 weeks and 28 weeks was also significantly higher in the prior ACD group. The mean cervical length prior to 24 weeks was significantly shorter in patients with a history preterm delivery because of advanced cervical dilation when compared with those with a history of preterm birth because of PPROM or preterm labor; these patients were also more likely to have a short cervix <25 mm prior to 24 weeks. A subgroup analysis was performed excluding patients treated with cerclage because this treatment may have affected the obstetric outcome (Table 2). For all of the
primary and secondary outcomes, results were similar to the initial analysis. In an unadjusted analysis, patients with a history of ACD had a relative risk for preterm birth <37, <34, and <28 weeks of 2.0, 3.8, and 4.0 respectively, with prior PPROM as the referent (Table 3). This elevated risk remained statistically significant in a sensitivity analysis eliminating patients with cervical cerclage. In a multivariate analysis including such factors as the number of prior preterm births, the gestational age of prior preterm birth, and demographic factors, ACD continued to be associated with significantly elevated risk for recurrent preterm birth at <37 weeks, <34 weeks, and <28 weeks (Table 4). In contrast, the most important predictors of short cervical length were African American and Hispanic race, and ACD was not associated with significantly increased risk for short cervix <25 mm at <24 weeks. The Figure
In this cohort, the overall rate of recurrent preterm delivery was 34%, consistent with previously published data.2,5 Patients with a history of advanced cervical dilation were at an increased risk for recurrent preterm birth compared with those with a history of PPROM or PTL, a relationship that persisted in multivariable analysis adjusting for several demographic and obstetric risk factors. A subgroup analysis excluding patients with a history-indicated or ultrasoundindicated cerclage revealed similar risk between preterm birth risk and prior ACD. Although it is recognized that different clinical preterm birth presentations may represent different syndromes and pathogenic pathways,7,8 the impact of the clinical presentation of the prior spontaneous PTB on the risk of recurrence has not been previously addressed in the literature. Furthermore, studies on interventions for prevention of recurrent spontaneous preterm birth have generally not differentiated patients based on the initial presentation in a prior pregnancy.5 This analysis had several findings consistent with previously reported data.2 It demonstrated that the gestational age of the prior preterm delivery, the number of prior preterm deliveries, and black race were significant risk factors for recurrent preterm birth. This analysis also found that short cervix occurred in a higher proportion of patients with a history of advanced cervical dilation, although in an adjusted analysis, this association was not significant when controlling for race and the number of prior preterm births. We offer 2 possible explanations for this finding. Patients with a history of advanced cervical dilation were more likely to get a cerclage. It is possible that cerclage prevents cervical shortening by interfering with the natural progression of cervical insufficiency. Another possible explanation is that patients with
NOVEMBER 2015 American Journal of Obstetrics & Gynecology
686.e4
Research
Obstetrics
ajog.org
TABLE 4
Log-linear model for primary and secondary outcomes Outcome
Preterm delivery <37 wks
Preterm delivery <34 wks
Preterm delivery <28 wks
Short cervix prior to 24 wks
Referent
Referent
Referent
Referent
Obstetric history Prior PPROM
a
2.77 (1.58e4.84)
a
2.33 (1.02e5.35)
a
1.11 (0.69e1.74)
Prior ACD
1.81 (1.32e2.48)
Prior PTL
1.19 (0.89e1.59)
1.21 (0.70e2.11)
0.71 (0.27e1.92)
0.40 (0.20e0.80)a
Prior full-term birthb
0.88 (0.76e1.03)
0.99 (0.80e1.23)
1.08 (0.79e1.48)
0.99 (0.70e1.40)
White
Referent
Referent
Referent
Referent
African American
1.04 (0.67e1.63)
0.88 (0.45e1.72)
1.50 (0.52e4.34)
2.49 (1.09e5.68)a
Hispanic
1.15 (0.76e1.73)
1.02 (0.52e2.0)
0.93 (0.29e3.0)
2.74 (1.30e5.76)a
Government-insured clinic
0.83 (0.63e1.10)
0.69 (0.42e1.15)
1.02 (0.50e2.08)
0.96 (0.50e1.84)
Cervical procedure
0.86 (0.59e1.27)
0.82 (0.46e1.48)
0.55 (0.22e1.38)
1.18 (0.52e1.84)
History-indicated cerclage
0.99 (0.71e1.39)
0.98 (0.56e1.70)
1.01 (0.47e2.19)
0.62 (0.30e1.31)
Race
Log-linear regression model was also adjusted for the following continuous variables: number of prior preterm births and gestational age of prior preterm delivery. ACD, advanced cervical dilation; PPROM, preterm premature rupture of membranes; PTL, preterm labor. a
Statistically significant difference; b Prior full-term birth indicates the number of prior full-term births (continuous variable).
Drassinower. Type of spontaneous preterm birth and risk of recurrence. Am J Obstet Gynecol 2015.
a history of ACD delivered at earlier gestational age or may have had ACD in the index pregnancy and therefore stopped receiving cervical length measurements. Cervical cerclage is an intervention that has been shown to be effective in preventing preterm delivery in select patient populations.12-14 However, there are currently no studies specifically addressing the efficacy of cerclage in patients with a history of advanced cervical dilation. Despite this lack of evidence, many providers offer a historyindicated cerclage to patients suspected to have had prior ACD (based on a history of preterm delivery because of advanced cervical dilation), an approach supported by the American Congress of Obstetricians and Gynecologists.9 An alternate strategy is to perform serial cervical length surveillance with the addition of an ultrasound-indicated cerclage in the presence of cervical shortening <25 mm.15 Our data show that patients with a history of advanced cervical dilation are at significant risk of recurrent PTB mediated by cervical shortening,
particularly in the setting of risk factors such as African American race and multiple prior preterm births. We suggest that it is therefore reasonable to consider a history-indicated cerclage for women with a history of advanced cervical dilation who may be at particularly high risk for preterm birth. Furthermore, future research is warranted to determine the relative benefits of prophylactic versus ultrasound-indicated cerclage for extremely high-risk women. There are several factors to consider in interpreting the validity of this study’s findings. All patients included in this study had a history of prior preterm birth, and the vast majority of patients received progesterone to reduce the risk of recurrent preterm birth. Although the analysis of the modulating effects of this intervention is beyond the scope of this analysis because of the ubiquity of progesterone use in preterm birth prevention strategies, this will likely be a limitation of subsequent preterm birth research as well and makes our results generalizable. Whereas our analysis was not powered to determine the effect of cerclage on outcomes, a sensitivity
686.e5 American Journal of Obstetrics & Gynecology NOVEMBER 2015
analysis excluding patients with cerclage yielded similar results, suggestive that our findings are applicable to populations with and without this intervention. Other limitations in this study are inherent to its retrospective nature. Cervical length measurements were performed for clinical purposes, and therefore, there is heterogeneity in the number and frequency of cervical length measurements performed in each patient. Additionally, patients with ACD who disproportionately delivered at earlier gestational ages no longer contributed cervical length measurements, and women remaining pregnant could provide a skewed representation of cervical length throughout pregnancy; this makes our findings potentially biased toward underestimating the prevalence of a short cervix in women with a history of ACD. However, given the large proportion of women who developed a short cervix with a history ACD determination of a relationship was still possible. Another limitation of this study is that the etiology of the prior preterm birth
Obstetrics
ajog.org
FIGURE
Association between prior preterm delivery and subsequent pregnancy
Research
in chorioamnionitis, PPROM, or PTL. This potential for misclassification of diagnosis could lead to biased results. Although we attempted to minimize the risk by careful review of medical records and excluded cases if the etiology of the prior preterm birth could not be reliably determined (for example, by excluding patients who presented for evaluation of PPROM but were found to be dilated on examination), given the complexity of diagnosis we cannot exclude it completely. The difficulty in differentiating the clinical presentation that preceded the prior preterm delivery may be a reason that its impact on the risk of recurrent preterm birth has not been elucidated further in research literature. In conclusion, our study showed that the clinical presentation that led to a prior preterm birth as assessed by patient history is a highly significant predictor of recurrent preterm delivery. We propose that future investigation and riskreducing interventions should consider these important factors in optimizing research strategies and patient care. -
REFERENCES
Kaplan-Meier curve showing the association between a prior preterm delivery because of advanced cervical dilatation and a shorter gestation in the subsequent pregnancy compared with PPROM and preterm labor. Log rank rest, P ¼ .001. ACD, advanced cervical dilation; PPROM, preterm premature rupture of membranes; PTL, preterm labor. Drassinower. Type of spontaneous preterm birth and risk of recurrence. Am J Obstet Gynecol 2015.
was, in part, self-reported. In some cases, the prior preterm birth occurred at our institution in which case the medical records for that delivery were available for review and the etiology was established with a high degree of certainty. In other cases, our classification on the etiology of the prior preterm birth was based on a review of the initial obstetrical visits, which in some situations relied on a patient’s report of her history. Although this is a limitation of our study,
it also makes our results generalizable because often providers must base their approach to a current pregnancy on the patient’s report of a previous pregnancy when review of medical records is not possible. Finally, we recognize that diagnosing cervical insufficiency is challenging because advanced dilation is often asymptomatic and may go unrecognized for an indefinite period of time, leading to exposed membranes that may result
1. Lockwood CJ, Kuczynski E. Risk stratification and pathological mechanisms in preterm delivery. Paediatr Perinat Epidemiol 2001;15(Suppl) 2:78-89. 2. Spong CY. Prediction and prevention of recurrent spontaneous preterm birth. Obstetr Gynecol 2007;110:405-15. 3. Mercer BM, Goldenberg RL, Moawad AH, et al. The preterm prediction study: effect of gestational age and cause of preterm birth on subsequent obstetric outcome. National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network. Am J Obstet Gynecol 1999;181:1216-21. 4. Fonseca EB, Celik E, Parra M, Singh M, Nicolaides KH; Fetal Medicine Foundation Second Trimester Screening Group. Progesterone and the risk of preterm birth among women with a short cervix. N Engl J Med 2007;357:462-9. 5. Meis PJ, Klebanoff M, Thom E, et al. Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate. N Engl J Med 2003;348:2379-85. 6. Iams JD, Goldenberg RL, Meis PJ, 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.
NOVEMBER 2015 American Journal of Obstetrics & Gynecology
686.e6
Research
Obstetrics
7. Creasy RK, Resnik R, Iams JD. Creasy and Resnik’s maternal-fetal medicine: principles and practice, 6th ed. Philadelphia, PA: Saunders/ Elsevier; 2009. 8. Iams JD, Cebrik D, Lynch C, Behrendt N, Das A. The rate of cervical change and the phenotype of spontaneous preterm birth. Am J Obstet Gynecol 2011;205:130.e1-6. 9. American College of Obstetricsw and Gynecologists. Cerclage for the management of cervical insufficiency. ACOG Practice bulletin no. 142. Obstet Gynecol 2014;123:372-9. 10. Grobman WA, Thom EA, Spong CY, et al. 17-alpha-hydroxyprogesterone caproate to
ajog.org prevent prematurity in nulliparas with cervical length less than 30 mm. Am J Obstet Gynecol 2012;207:390.e1-8. 11. Spiegelman D, Hertzmark E. Easy SAS calculations for risk or prevalence ratios and differences. Am J Epidemiol 2005;162: 199-200. 12. Berghella V, Rafael TJ, Szychowski JM, Rust OA, Owen J. Cerclage for short cervix on ultrasonography in women with singleton gestations and previous preterm birth: a metaanalysis. Obstet Gynecol 2011;117:663-71. 13. Owen J, Hankins G, Iams JD, et al. Multicenter randomized trial of cerclage for
686.e7 American Journal of Obstetrics & Gynecology NOVEMBER 2015
preterm birth prevention in high-risk women with shortened midtrimester cervical length. Am J Obstet Gynecol 2009;201:375.e1-8. 14. Final report of the Medical Research Council/Royal College of Obstetricians and Gynaecologists multicentre randomised trial of cervical cerclage. MRC/RCOG Working Party on Cervical Cerclage. Br J Obstet Gynaecol 1993;100:516-23. 15. Berghella V, Mackeen AD. Cervical length screening with ultrasound-indicated cerclage compared with history-indicated cerclage for prevention of preterm birth: a meta-analysis. Obstet Gynecol 2011;118:148-55.