Objective
The objective of the study was to compare the efficacy and outcomes of 2 vs 1 stitch at the time of cervical cerclage placement for preterm birth prevention.
Study Design
This was a retrospective cohort study of women with singleton gestation undergoing history- or ultrasound-indicated transvaginal cervical cerclage prior to 24 weeks . The primary outcome was delivery at less than 37 weeks. The secondary outcomes included gestational age at delivery at less than 35, less than 34, less than 32, less than 28, and less than 24 weeks, perioperative details at the time of cerclage placement and removal, and maternal and neonatal outcomes. Comparison was made between patients with 2 vs 1 stitch at the time of cerclage placement. History- and ultrasound-indicated cerclages were analyzed separately.
Results
Four hundred forty-four patients met inclusion criteria, 237 being history indicated (2 stitches, n = 86, 1 stitch, n = 151), and 207 ultrasound indicated (2 stitches, n = 117, 1 stitch, n = 90). Gestational age at delivery at less than 37 weeks was not significantly different between the 2 groups for both history- and ultrasound-indicated cerclage, even after adjusting for demographic differences and suture type (39% vs 35%; adjusted odds ratio, 1.38; 95% confidence interval, 0.64–3.01; and 44% vs 49%; adjusted odds ratio, 0.66; 95% confidence interval, 0.27–1.61, respectively).
Conclusion
Two stitches at the time of cerclage do not appear to improve pregnancy outcome either in the history- or the ultrasound-indicated procedures, compared with 1 stitch.
Cervical cerclage is currently used in about 0.4% of pregnancies in the United States for preterm birth prevention. Cervical cerclage placement indications include a history of cervical insufficiency (recurrent second-trimester losses/preterm births), cervical length less than 25 mm on transvaginal ultrasound, or dilated cervix on physical examination. The cerclage is typically placed between 12-23 weeks.
Different surgical techniques for cerclage have been reported. The current methods used are mostly modifications of the original Shirodkar and McDonald techniques. No technique has been shown to be superior in reported studies. The most common current technique is the McDonald procedure, in which a purse-string suture of monofilament suture (such as polypropylene) or polyester fiber tape (such as Mersilene) is placed in the cervix “as high as possible to approximate the level of the internal os.” Some practitioners have chosen to place an additional stitch at the time of initial cerclage procedure.
Proposed mechanisms for which a second suture might improve the efficacy of cerclage include a greater cervical height and an additional support to the posterior cervix. The question as to whether the additional stitch improves the efficacy of the procedure and desired outcome remains incompletely investigated. Limited data show no benefit on placing 2 stitches instead of 1 at the time of initial cerclage placement, but studies had no controls, had a small sample size, and/or the analyses grouped together different indications for cerclage.
Our study aimed to compare outcomes of 2 vs 1 stitch at the time of transvaginal cervical history-indicated or ultrasound-indicated cerclage placement for preterm birth prevention.
Materials and Methods
We conducted a retrospective cohort study of women with a singleton gestation who underwent transvaginal cervical cerclage and delivered at Thomas Jefferson University (TJUH) and Albert Einstein Medical Center (AEMC; both in Philadelphia, PA, between January 1994 and June 2011. The study was approved by the institutional review board at each institution. Patients who underwent cerclage placement were identified using an existing database at TJUH and a billing code system at AEMC. Women with a singleton gestation who underwent history- or ultrasound-indicated transvaginal cerclage placement prior to 24 weeks were included for study.
Patients included in the history-indicated cerclage group had cerclage placed based on their poor obstetrical history and/or risk factors as determined by their obstetrical provider, which in general were a history of multiple prior preterm births and/or second-trimester losses . Patients were allocated to the ultrasound-indicated cerclage group when the indication for the cervical cerclage was a transvaginal cervical length less than 25 mm before 24 weeks. Starting in 2003, singleton gestations with prior spontaneous preterm birth received 17-hydroxy progesterone caproate starting at 16-20 weeks.
The patient population of both hospitals that participated in the study is similar, with both hospitals serving the inner-city women of the city of Philadelphia. The number of sutures placed and the suture material used was chosen by the operating surgeon. At TJUH, cervical cerclages were usually performed using the McDonald technique, with 1 stitch of Mersilene 5 mm tape (Ethicon, Inc, Somerville, NJ), placed in a purse string fashion. At AEMC, the McDonald technique using a nonabsorbable braided polyester suture (silky II Polydek; Deknatel, Cambridge, MA) was performed. This technique was frequently modified at AEMC by the placement of 2 stitches at the time of initial cerclage. After the first stitch was placed and tied, a second stitch, usually of the same suture, was placed in a similar fashion proximal to the first suture, closer to the internal os.
Women who underwent a transabdominal cerclage, a surgical technique different from McDonald and Shirodkar, physical examination-indicated cerclage, multiple gestations, cerclage placement after 24 weeks, ultrasound-indicated cerclage with cervical length larger than 25 mm, and no medical records available to extract the data were excluded from this analysis. Patients who underwent voluntary termination of pregnancy or medically indicated delivery prior to 35 weeks’ gestation for preeclampsia, intrauterine fetal demise, and/or nonreassuring fetal status were also excluded from the analysis.
All the maternal medical records were reviewed. Patient characteristics, operative details, removal of cerclage data, and maternal and neonatal outcomes were collected and analyzed.
The types of complications at the time cerclage placement were defined as pregnancy loss less than 20 weeks’ gestation; cervical laceration (requiring the use of hemostatic measures including suture, ferric subsulfate solution [Monsel’s solution], and/or silver nitrate); repeat cerclage (need of second procedure during the same pregnancy); excess bleeding (requiring the use of hemostatic measurement different from pressure alone); fistula formation; intraoperative rupture of membranes; and cervical stenosis. The types of complications at the time of cerclage removal were defined as none, retained suture (unable to completely remove stitch at initial attempt), cervical laceration, significant pain (the use of nonplanned regional block, intravenous analgesia/anesthesia), or excess bleeding.
The primary outcome was gestational age at delivery at less than 37 weeks in women with 2 stitches compared with women with 1 stitch at the time of transvaginal cervical cerclage.
The secondary outcomes included gestational age at delivery at less than 35, less than 34, less than 32, less than 28, and less than 24 weeks; perioperative details at the time of cerclage placement and removal; and maternal and neonatal outcomes. Perioperative details analyzed included gestational age (GA) at placement (weeks); type of cerclage (McDonald vs Shirodkar); the cervical length at placement if ultrasound indicated (millimeters); cervical dilation at placement (centimeters); type of suture (Mersilene 5 mm tape; Ethicon, or a polyester nonabsorbable braided suture including Ethibond; Ethicon; Tevdek; Deknatel; Silky II Polydek; Deknatel); operative complications at the time of cerclage placement (no, yes); and type complication at the time cerclage as described.
Comparison was made between patients who underwent 2 vs 1 stitch at the time of transvaginal cerclage placement. Patients who underwent history-indicated cerclage were analyzed separately from those who underwent ultrasound-indicated cerclage. The 2 stitches technique was considered the experimental group, whereas the 1 stitch group was considered the control group.
Data were analyzed using SAS statistical software version 9.3 (SAS Institute Inc, Cary, NC). A Student t test and Fisher exact test were used as appropriate. Categorical variables are summarized by frequencies and percentages with P values from a Fisher exact test. Continuous variables are summarized by mean ± SD, and P values are from Student t tests. Nonnormally distributed variables are summarized by median and interquartile range with P values from a Mann-Whitney-Wilcoxon test. A P < .05 was considered significant. Multivariable logistic regression analysis was applied to the primary outcome and other cutoffs of gestational age at delivery.
Results
Study population
We identified 726 patients undergoing cervical cerclage during the study period. Of those, 327 (45%) were done at TJUH and 399 (55%) at AEMC. After reviewing all medical records, 444 patients were included for analysis and 282 patients were excluded as illustrated in the Figure . The types of sutures used for the cerclage included Mersilene 5 mm tape; Ethicon; Prolene; Ethicon; or a nonabsorbable braided polyester suture (Ethibond; Ethicon; Tevdek; Deknatel; or Silky II Polydek; Deknatel).
History-indicated cerclage
Of 237 women who underwent history-indicated cerclage, 86 (36.3%) had 2 stitches and 151 (63.7%) had 1 stitch placed. African American was the predominant race in both groups. There were more Hispanics in the 2 stitch group (18% vs 5%) and more white females in the 1 stitch group (25% vs 9%). Women who had 1 stitch placed were slightly older and had significantly more second-trimester losses (14-24 weeks) and more cone/loop electrosurgical excision procedure (LEEP) procedures. Progesterone use (either 17α-hydroxyprogesterone or vaginal progesterone) was higher in the group of women with 2 stitches ( Table 1 ).
| Variable | 2 stitches (n = 86) | 1 stitch (n = 151) | P value | ||
|---|---|---|---|---|---|
| n | % | n | % | ||
| Age a | 29.0 ± 6.0 | 30.6 ± 5.6 | .048 | ||
| Race | < .001 | ||||
| African American | 60 | 71 | 102 | 68 | |
| White | 8 b | 9 b | 38 b | 25 b | |
| Hispanic | 15 b | 18 b | 7 b | 5 b | |
| Others (Asians, others) | 2 | 2 | 4 | 3 | |
| Gravity a | 4.6 ± 1.9 | 4.9 ± 2.2 | .236 | ||
| Parity a | 0.7 ± 1.0 | 0.6 ± 0.9 | .808 | ||
| BMI, kg/m 2 a | 31.3 ± 8.0 | 30.7 ± 8.1 | .582 | ||
| Smoking | 11 | 13 | 21 | 14 | .846 |
| Prior preterm birth | |||||
| 14-24 wks | 55 b | 64 b | 117 b | 77 b | .034 b |
| 24-36 6/7 wks | 53 | 62 | 77 | 51 | .136 |
| GA earliest preterm birth a | 24.5 ± 4.9 | 23.3 ± 4.3 | .078 | ||
| Prior preterm births (20-36 wk), n a | 1.3 ± 1.0 | 1.4 ± 1.1 | .632 | ||
| Second-trimester losses (14-24 wk), n a | 1.0 ± 1.1 | 1.2 ± 0.9 | .214 | ||
| Prior cone/LEEP | 3 b | 3 b | 19 b | 13 b | .020 b |
| More than 1 D&C | 24 | 28 | 49 | 32 | .559 |
| Uterine anomalies | 1 | 1 | 0 | 0 | .363 |
| DES exposure | 1 | 1 | 8 | 5 | .161 |
| Previous cerclage | 58 | 67 | 97 | 64 | .671 |
| Use of progesterone | 30 b | 35 b | 21 b | 14 b | < .001 b |
The mean gestational age at cerclage placement was 15.0 ± 3.1 for the 2 stitches group and 13.7 ± 1.6 weeks for the 1 stitch group. The type of suture differed among the 2 groups. Polyester nonabsorbable braided suture was the suture of choice for women when 2 stitches were placed, and Mersilene 5 mm tape (Ethicon) was the preferred suture for women who underwent cerclage with 1 stitch (81% and 75%, respectively). There was no difference in the incidence of procedure-related complications either at the time of placement or the removal of the stitch(es). The mean gestational age at removal was 34.0 ± 5.4 weeks for the 2 stitches group and 33.7 ± 6.1 weeks in the 1 stitch group ( Table 2 ).
| Variable | 2 stitches (n = 86) | 1 stitch (n = 151) | P value | ||
|---|---|---|---|---|---|
| n | % | n | % | ||
| GA placement cerclage, wks a | 15.0 ± 3.1 | 13.7 ± 1.6 | < .001 | ||
| Type cerclage | .300 | ||||
| McDonald | 86 | 100 | 147 | 97 | |
| Shirodkar | 0 | 0 | 4 | 3 | |
| Cervical dilation at placement, cm a | 0.2 ± 0.4 | 0.1 ± 0.3 | .426 | ||
| Type suture | < .001 | ||||
| Mersilene 5 mm tape | 9 | 10 | 108 | 75 | |
| Polyester nonabsorbable braided suture b | 70 | 81 | 26 | 18 | |
| Prolene | 1 | 1 | 10 | 7 | |
| Combination | 6 | 7 | 0 | 0 | |
| Complications placement | .655 | ||||
| Yes | 10 | 12 | 14 | 9 | |
| No | 76 | 88 | 137 | 91 | |
| Type complication placement | |||||
| Pregnancy loss <20 wks | 5 | 6 | 10 | 7 | > .999 |
| Cervical laceration | 1 | 1 | 1 | 1 | > .999 |
| Repeat cerclage | 4 | 5 | 4 | 3 | .466 |
| Others c | 1 | 1 | 1 | 1 | > .999 |
| GA removal a | 34.0 ± 5.4 | 33.7 ± 6.1 | .742 | ||
| Complication removal | |||||
| None | 78 | 91 | 136 | 90 | > .999 |
| Retained suture | 1 | 1 | 8 | 5 | .166 |
| Cervical laceration | 1 | 1 | 3 | 2 | > .999 |
| Other (significant pain, bleeding) | 0 | 0 | 4 | 3 | .300 |
| Cerclage placement to removal, d a | 131.0 ± 42.7 | 137.4 ± 45.2 | .287 | ||
| Cerclage removal to delivery, d a | 12.3 ± 13.1 | 10.4 ± 15.9 | .347 | ||
b Polyester nonabsorbable braided suture (Ethibond; Ethicon, Inc, Somerville, NJ. Tevdek, Deknatel, Cambridge, MA. Silky II Polydek; Deknatel);
c Others included excess bleeding, vesicovaginal fistula, stenosis, and intraoperative rupture of membranes.
The mean days from cerclage placement to removal was not significantly different between the 2 groups (131.0 days in 2 stitches vs 137.4 days in 1 stitch, P = .287) as well as the mean days from cerclage removal to delivery (12.3 vs 10.4 days, P = .347). Most cerclage removals occurred in a planned fashion (72% vs 61%, P = NS), in a hospital setting (68% vs 58%, NS), without the use of analgesia (67% vs 72%, respectively, P = NS) in the 2 and in the 1 stitch groups, respectively.
The primary outcome, gestational age at delivery less than 37 weeks, was not significantly different between the 2 groups, even after adjusting for demographic differences and suture type (35% vs 39%; P = .914; adjusted odds ratio [aOR], 1.38; 95% confidence interval [CI], 0.64–3.01). Neither was a difference at less than 35, less than 34, less than 32, less than 28, and less than 24 weeks’ gestation ( Table 3 ).
| Variable | 2 stitches (n = 86) | 1 stitch (n = 151) | OR | 95% CI | aOR | 95% CI | ||
|---|---|---|---|---|---|---|---|---|
| n | % | n | % | |||||
| GA delivery <37 wks | 30 | 35 | 59 | 39 | 0.84 | 0.48–1.45 | 1.38 | 0.64–3.01 |
| GA delivery <35 wks | 19 | 22 | 35 | 23 | 0.94 | 0.50–1.77 | 1.64 | 0.64–4.20 |
| GA delivery <34 wks | 15 | 17 | 33 | 22 | 0.76 | 0.38–1.49 | 1.51 | 0.56–4.06 |
| GA delivery <32 wks | 10 | 12 | 28 | 19 | 0.58 | 0.27–1.26 | 0.99 | 0.32–3.06 |
| GA delivery <28 wks | 9 | 10 | 22 | 15 | 0.69 | 0.30–1.57 | 1.18 | 0.35–3.98 |
| GA delivery <24 wks | 7 | 8 | 15 | 10 | 0.80 | 0.31–2.06 | 1.27 | 0.33–4.80 |
Maternal and neonatal outcomes were analyzed. Of note, there was a higher frequency of cesarean deliveries in the patients who had 2 stitches at the time of initial cerclage placement (33 vs 50%, P = .015). However, there was no significant difference noted regarding the indication for cesarean delivery or the history of prior cesarean delivery. Neonatal Apgar scores at 1 and 5 minutes were significantly different between the 2 groups. Neonatal birthweight, arterial cord gas pH, and neonatal intensive care unit (NICU) admission was similar among the groups ( Table 4 ).
| Variable | 2 stitches (n = 86) | 1 stitch (n = 151) | P value | ||
|---|---|---|---|---|---|
| n | % | n | % | ||
| Mode delivery | .015 | ||||
| SVD/assisted vaginal | 43 | 50 | 101 | 67 | |
| Cesarean section | 43 | 50 | 50 | 33 | |
| Prior cesarean section | 21 | 24 | 40 | 26 | .760 |
| PPROM | 24 | 28 | 33 | 22 | .343 |
| Cervical laceration at time of delivery | 1 | 1 | 4 | 3 | .656 |
| Indication for delivery | |||||
| Spontaneous labor/PROM | 57 | 66 | 107 | 71 | .468 |
| Chorioamnionitis | 7 | 8 | 10 | 7 | .794 |
| Scheduled delivery (c/s-IOL) | 16 | 19 | 28 | 19 | > .999 |
| Other a | 13 | 15 | 13 | 9 | .135 |
| Indication for cesarean section | |||||
| Arrest of labor | 11 | 26 | 12 | 28 | > .999 |
| Fetal indication b | 15 | 35 | 9 | 18 | .105 |
| Maternal indication c | 1 | 2 | 0 | 0 | .936 |
| Scheduled cesarean section d | 16 | 37 | 29 | 58 | .072 |
| Neonatal | |||||
| Birthweight e | 2900.0 ± 907 | 2797.2 ± 1051 | .441 | ||
| Apgar f | |||||
| 1 min | 8 (8–9) | 8 (7–8) | < .001 | ||
| 5 min | 9 (9–9) | 9 (8–9) | .012 | ||
| Arterial cord gas pH e | 7.25 ± 0.06 | 7.24 ± 0.09 | .635 | ||
| NICU admission | 16 | 19 | 27 | 18 | > .999 |
a Other (fetal status, oligohydramnios, abruption, preeclampsia);
b Fetal indication (nonreassuring fetal status, cord prolapse, fetal growth restriction);
c Maternal indication (preeclampsia);
d Scheduled cesarean section (elective, repeat, macrosomia, malpresentation, myomectomy, placenta previa);
Ultrasound-indicated cerclage
Of the 207 women who underwent ultrasound-indicated cerclage, 117 (56.5%) had 2 stitches and 90 (43.5%) had 1 stitch. African American was again the predominant race in both groups. There were more Hispanics in the 2 stitches group (10% vs 1%) and more white females in the 1 stitch group (15% vs 5%). Parity, number of prior preterm births, and diethylstilbestrol (DES) exposure was significantly different between the groups. The mean parity was 0.8 ± 1.2 for the 2 stitches group and 0.4 ± 0.8 for the 1 stitch group ( P = .006). Women who received 1 stitch had more prior preterm births than those who received 2 stitches (mean 0.8 ± 0.8 vs 0.6 ± 0.9, P = .037). DES exposure was a risk factor in 5 women in the 1 stitch group, whereas none was identified in the 2 stitches group ( P = .015) ( Table 5 ).
| Variable | 2 stitches (n = 117) | 1 stitch (n = 90) | P value | ||
|---|---|---|---|---|---|
| n | % | n | % | ||
| Age a | 27.3 ± 6.7 | 28.7 ± 5.9 | .126 | ||
| Race a | .002 | ||||
| African American | 96 | 84 | 71 | 81 | |
| White | 6 | 5 | 13 | 15 | |
| Hispanic | 12 | 10 | 1 | 1 | |
| Others (Asian, others) | 1 | 1 | 3 | 3 | |
| Gravity a | 4.0 ± 2.3 | 3.9 ± 1.9 | .904 | ||
| Parity a | 0.8 ± 1.2 | 0.4 ± 0.8 | .006 | ||
| BMI, kg/m 2 a | 31.1 ± 8.5 | 31.9 ± 9.2 | .529 | ||
| Smoking | 17 | 15 | 19 | 21 | .268 |
| Prior preterm birth | |||||
| 14-24 wks | 38 | 32 | 40 | 44 | .084 |
| 24-36 6/7 wks | 37 | 32 | 36 | 40 | .241 |
| GA earliest preterm birth | 26.4 ± 5.6 | 24.5 ± 4.9 | .067 | ||
| Prior preterm births (20-36 wk), n a | 0.6 ± 0.9 | 0.8 ± 0.8 | .037 | ||
| Second-trimester losses 14-24, n a | 0.4 ± 0.6 | 0.5 ± 0.7 | .064 | ||
| Prior cone/LEEP | 11 | 9 | 9 | 10 | > .999 |
| More than 1 D&C | 38 | 32 | 32 | 36 | .659 |
| Uterine anomalies | 0 | 0 | 1 | 1 | .435 |
| DES exposure | 0 | 0 | 5 | 6 | .015 |
| Previous cerclage | 13 | 11 | 11 | 12 | .830 |
| Use progesterone | 35 | 30 | 23 | 26 | .534 |
The mean gestational age at cerclage placement was 20 weeks for both groups. The cervical length at placement was also similar between both groups (13 and 15 mm). The type of suture differed among the technique used. Polyester nonabsorbable braided suture was the preferred suture for women who underwent cerclage with 2 stitches, and Mersilene 5 mm tape (Ethicon, Inc) was the suture of choice for women when 1 stitch was placed (94% vs 68%). There was no significant difference in the incidence of procedure related complications either at the time of placement or removal of the stitch(es) and the mean gestational age at removal of the suture(s).
There were more pregnancy losses at less than 20 weeks in the 1 stitch group (n = 6 vs n = 2); however, it did not reach statistical significance ( P = .08) ( Table 6 ). The mean days from cerclage placement to removal was not significantly different between the 2 groups (96.5 in 2 stitches vs 91.6 in 1 stitch, P = .373) as well as the mean days from cerclage removal to delivery (11.3 vs 9.7, P = .279).
| Variable | 2 stitches (n = 117) | 1 stitch (n = 90) | P value | ||
|---|---|---|---|---|---|
| n | % | n | % | ||
| GA placement cerclage, wks a | 20.0 ± 2.6 | 19.6 ± 2.5 | .235 | ||
| Type cerclage | > .999 | ||||
| McDonald | 117 | 100 | 89 | 99 | |
| Shirodkar | 0 | 0 | 1 | 1 | |
| CL placement, mm a | 13.4 ± 7.3 | 14.6 ± 7.2 | .235 | ||
| Cervical dilation at placement, cm | 0.6 ± 0.6 | 0.7 ± 0.8 | .184 | ||
| Type suture | < .001 | ||||
| Mersilene 5 mm tape | 2 | 2 | 61 | 68 | |
| Prolene | 3 | 3 | 5 | 6 | |
| Polyester nonabsorbable braided suture b | 110 | 94 | 24 | 27 | |
| Combination | 2 | 2 | 0 | 0 | |
| Complications | > .999 | ||||
| Yes | 10 | 9 | 8 | 9 | |
| No | 107 | 91 | 82 | 91 | |
| Type complications | .072 | ||||
| Pregnancy loss <20 wks | 2 | 2 | 6 | 7 | .080 |
| Cervical laceration | 0 | 0 | 0 | 0 | — |
| Repeat cerclage | 8 | 7 | 2 | 2 | .192 |
| Others c | 0 | 0 | 0 | 0 | — |
| GA removal a | 33.6 ± 5.0 | 32.3 ± 6.4 | .113 | ||
| Complications removal | |||||
| None | 105 | 90 | 83 | 92 | .632 |
| Retained suture | 9 | 8 | 5 | 6 | .589 |
| Cervical laceration | 0 | 0 | 0 | 0 | — |
| Other (significant pain, bleeding) | 2 | 2 | 2 | 2 | > .999 |
| Cerclage placement to removal, d a | 96.5 ± 33.5 | 91.6 ± 42.5 | .373 | ||
| Cerclage removal to delivery, d a | 11.3 ± 11.0 | 9.7 ± 10.7 | .279 | ||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
