Transvaginal cervical cerclage

INTRODUCTION — Cervical cerclage refers to a variety of surgical procedures in which synthetic suture or tape is used to reinforce the cervix. By mechanically increasing its tensile strength, the occurrence of adverse perinatal events associated with cervical insufficiency can be reduced. These adverse perinatal events, which are often interrelated, can include:

●Prolapse of the fetal membranes into the vagina

●Intraamniotic infection

●Preterm prelabor rupture of the fetal membranes (PPROM)

●Preterm labor and birth

●Fetal loss

However, the efficacy of cerclage for preventing these adverse events compared with no intervention or other interventions is controversial.

Most cerclages are placed via a transvaginal approach. The transabdominal approach is more invasive but allows higher placement at the cervicoisthmic portion of the uterus, while transvaginal cerclages are generally placed distal to the internal os. The transabdominal approach is used in patients unable to undergo a transvaginal procedure or those who failed to deliver a healthy newborn after at least one previous prophylactic transvaginal cerclage.

This topic will discuss the procedure for transvaginal cervical cerclage. General issues related to cervical insufficiency and procedures for transabdominal cervical cerclage are reviewed separately. (See "Cervical insufficiency" and "Transabdominal and laparoscopic cervicoisthmic cerclage".)

GOAL — Cerclage is performed to reduce pregnancy loss/preterm birth in patients with cervical insufficiency, which may be based on their past obstetric history, premature cervical shortening on transvaginal ultrasound examination, or a prematurely dilated and/or effaced cervix on a digital or speculum examination. (See "Cervical insufficiency" and "Short cervix before 24 weeks: Screening and management in singleton pregnancies", section on 'Clinical approach'.)

CONTRAINDICATIONS — The major absolute contraindications to cerclage are clinical scenarios where the procedure is unlikely to reduce the risk of preterm birth or improve fetal outcome:

●Fetal anomaly incompatible with life

●Intrauterine infection

●Active preterm labor

●Preterm prelabor rupture of membranes (PPROM)

●Fetal demise

●Active uterine bleeding (eg, placental abruption); however, placenta previa is not an absolute contraindication to cerclage placement. (See "Placenta previa: Management", section on 'Cerclage'.)

The chance of an adverse outcome is increased in patients with widely dilated cervixes and/or prolapsed membranes; therefore, shared decision-making is particularly important in these cases. There is no dilation at which cerclage should not be attempted; however, the prognosis worsens when cervical dilation exceeds 4 cm [1]. The presence of fetal membranes prolapsing through the external cervical os is a relative contraindication to the procedure because up to 65 percent of patients will experience iatrogenic rupture of the membranes in this setting [2-5]. Exposure of the membranes to the vaginal flora also likely increases the risk of infection. Nevertheless, cerclage has been placed with successful newborn outcome in patients with widely dilated cervixes, 0 mm ultrasound-measured cervical length, and/or prolapsed membranes [6]. (See 'Replace prolapsed membranes, if present' below.)

Multiple gestation is not a contraindication. Cerclage may be indicated in those with a past history of cervical insufficiency and possibly in some patients with a short cervix, whereas routine cerclage is not useful for prolonging pregnancy. (See "Twin pregnancy: Management of pregnancy complications", section on 'Patient selection for cerclage' and "Triplet pregnancy", section on 'Unproven interventions'.)

UPPER AND LOWER GESTATIONAL AGE THRESHOLDS FOR CERCLAGE PLACEMENT — Most cerclages are placed between 12 and 24 weeks of gestation. There is no consensus about the precise lower and upper limits of gestational age for performing the procedure.

●Pregnancies less than 12 weeks – The procedure generally is not performed before 12 weeks of gestation because results of aneuploidy screening may not be available. If aneuploidy screening is not performed, waiting until the end of the first trimester allows most miscarriages related to aneuploidy to occur. Even if it is known that the fetus is euploid, waiting until the end of the first trimester permits sonographic evaluation for major fetal anomalies [7].

●Pregnancies 24 to 28 weeks – Cerclage placement at 24 to 28 weeks of gestation, a period characterized by high neonatal morbidity and mortality if delivery occurs, is controversial. Most clinicians avoid placing a cerclage after fetal viability has been reached (generally regarded as approximately 24 weeks of gestation) since the procedure may cause accidental rupture of the fetal membranes leading to extremely preterm delivery, with its attendant high risk of neonatal morbidity and mortality. However, each case between 24 and 28 weeks must be individualized, weighing the risks of the procedure against the likely outcome with expectant management.

●Pregnancies greater than 28 weeks – A cerclage generally is not placed after 28 weeks because births at this gestational age are likely to have a reasonably good outcome.

PROCEDURE — The cerclage procedure is largely based on data from case series, observational studies, and expert opinion; there are few small and no large randomized trials evaluating any aspect of the procedure.

Preoperative assessment

Fetal assessment — Before scheduling the cerclage, the clinician should:

●Confirm fetal cardiac activity and gestational age

●Obtain a fetal anatomic survey for identifiable structural anomalies that could affect the patient's decision to continue the pregnancy

●Offer aneuploidy screening if not already performed

Evaluation for infection

●Cervicovaginal infection – We do not routinely screen for cervicovaginal infections as there is insufficient evidence that preprocedure screening for sexually transmitted infection improves cerclage outcome [8]. If the patient is symptomatic or at high risk of acquiring a sexually transmitted infection and has no documentation of recent negative test results, the author tests for gonorrhea and chlamydia. If antibiotic therapy is indicated for positive test results, treatment is completed prior to cerclage placement, if possible. Evaluation and treatment of sexually transmitted infection according to standard guidelines is good obstetric practice. (See "Screening for sexually transmitted infections", section on 'Chlamydia and gonorrhea' and "Prenatal care: Initial assessment", section on 'Laboratory tests'.)

●Subclinical intraamniotic infection – Intraamniotic infection is a contraindication to the procedure since the cerclage is not likely to be effective [9-11] and prolonging the pregnancy places the mother and fetus at risk of sepsis [10,12]. However, the prevalence of infection is not consistent across patient populations, diagnosis of infection requires an invasive procedure, and randomized trials to determine the utility of amniocentesis before cerclage have not been performed; therefore, amniocentesis for diagnosis of subclinical infection is performed selectively, not routinely.

•History-indicated cerclage – Amniocentesis is not performed before history-indicated cerclage given the very low prevalence of subclinical intraamniotic infection in early pregnancy and the risks for procedure-related pregnancy loss and fetal harm after early amniocentesis [13].

•Ultrasound-indicated cerclage – Most clinicians do not perform amniocentesis before ultrasound-indicated cerclage because the prevalence of subclinical infection is relatively low (1 to 2 percent) in patients with a closed cervix and the clinical significance of Ureaplasma and Mycoplasma species, which are the most common organisms detected, is unclear [13,14].

•Physical examination-indicated cerclage – The best approach before physical examination-indicated cerclage (also called rescue or emergency cerclage) is more controversial. The author performs amniocentesis when he suspects intraamniotic infection even though the patient is afebrile. If the Gram stain and biomarkers of intraamniotic infection (eg, glucose, leukocyte count, lactate dehydrogenase, interleukin-6) are not indicative of infection, then he proceeds with cerclage. The author is generally unwilling to wait 48 hours to obtain microbiologic culture results on amniotic fluid before performing a physical examination-indicated procedure.

The incidence of intraamniotic infection ranges from 10 to 50 percent when the cervix is dilated ≥2 cm on physical examination and/or membranes are prolapsed into the vagina [13]. The incidence is increased, but lower (<5 to 10 percent), in patients with ultrasound findings suggestive of inflammation (membrane edema, separation of membranes from the decidua, or debris [sludge] in the amniotic fluid (image 1)) or membranes visible and exposed at the external os, but cervical dilation <2 cm and no prolapse into the vagina.

Cerclage is not performed when there is laboratory evidence of intraamniotic infection. Although treatment of intraamniotic infection without delivery has been reported, it is investigational. Diagnosis and treatment of intraamniotic infection are reviewed in more detail separately. (See "Intraamniotic infection (clinical chorioamnionitis)", section on 'Diagnosis'.)

●Urinary tract infection – There is no consensus regarding preprocedure screening for asymptomatic bacteriuria and insufficient data showing that preprocedure screening improves pregnancy outcome [8].

Observation period before unplanned procedures — The author observes candidates for ultrasound- or physical examination-indicated cerclage for up to 24 hours before proceeding with surgery to help exclude preterm prelabor rupture of membranes (PPROM), preterm labor, occult abruption, and intraamniotic infection. Cerclage is unlikely to be effective in these settings and may increase maternal morbidity.

●PPROM is excluded using standard methods. (See "Preterm prelabor rupture of membranes: Clinical manifestations and diagnosis", section on 'Diagnostic evaluation and diagnosis'.)

When the cervix is widely dilated, a small to moderate amount of fluid that contains both a mucous and a watery component may collect in the posterior vaginal fornix and likely represents transudation across intact membranes. It is important to distinguish this fluid from amniotic fluid using standard methods. (See "Preterm prelabor rupture of membranes: Clinical manifestations and diagnosis", section on 'Diagnostic evaluation and diagnosis'.)

●Mild, irregular contractions noted on a uterine contraction monitor can be a normal finding and usually resolve over time, but regular contractions, especially those of increasing frequency and intensity, may indicate idiopathic preterm labor, subclinical intrauterine infection, or occult abruption. Placement of a cerclage in the setting of contractions can lead to tearing of the cervical tissue, which may compromise future pregnancies.

Prophylactic pharmacotherapy

Prophylactic antibiotics and tocolytics — Prophylactic antibiotics have been considered to reduce the risk for intraamniotic infection, and are commonly used before physical examination-indicated procedures. Use of prophylactic indomethacin, a prostaglandin synthetase inhibitor, has also been considered because cerclage placement has been associated with transient elevation in prostaglandin levels and prostaglandins induce uterine contractions [15]; thus, indomethacin is commonly used before physical examination-indicated procedures. Use of these interventions is less common before ultrasound-indicated procedures. The American College of Obstetricians and Gynecologists and others have not made a strong recommendation for or against use of prophylactic antibiotics or tocolytics to improve the efficacy of cerclage, regardless of timing or indication, because of very limited available evidence [16-20].

Our approach is as follows:

●History-indicated cerclage – Antibiotic prophylaxis is not administered as there appears to be minimal risk that a history-indicated procedure will promote intraamniotic infection [21,22]. Similarly, tocolytics are not administered as these patients do not have clinically measurable uterine irritability and virtually no data are available on the use of tocolytics in the first trimester.

●Ultrasound-indicated cerclage – The author does not administer antibiotic or indomethacin prophylaxis, but will administer postprocedure indomethacin for 48 hours to patients who have contractions/uterine irritability after the procedure. A retrospective cohort study in patients undergoing ultrasound-indicated cerclage reported similar preterm birth rates whether or not prophylactic indomethacin was administered (spontaneous preterm birth <35 weeks: 20 out of 51 [39 percent] with indomethacin versus 17 out of 50 [34 percent] without indomethacin; risk ratio [RR] 1.15, 95% CI 0.69-1.93) [23]. However, given limited data regarding antibiotic and/or indomethacin prophylaxis, practice patterns vary and may include use of one or both of these drugs.

●Physical examination-indicated cerclage – Antibiotic prophylaxis and indomethacin are routinely administered perioperatively to prolong latency. In a randomized trial of 53 patients undergoing physical examination-indicated cerclage, use of both prophylactic antibiotics and indomethacin resulted in a nonstatistical reduction in preterm birth <24, <28, and <36 weeks compared with no treatment, but significantly increased the proportion of patients with prolongation of latency over 28 days (92 versus 63 percent) [24].

We use the following regimen because it is the only one that has been evaluated in a randomized trial [24]; however, other regimens may be as or more effective (eg, cefazolin 2 g [3 g for patients ≥100 kg] or use of another broad-spectrum antibiotic, indomethacin 50 or 100 mg loading dose followed by 25 or 50 mg every 6 hours for 48 hours).

•Cefazolin 1 g (2 g for patients ≥100 kg) intravenously preoperatively and at 8 and 16 hours postoperatively

•Indomethacin 50 mg orally preoperatively and at 8 and 16 hours postoperatively

Preoperative administration of indomethacin may provide uterine relaxation and reduce membrane prolapse, thus facilitating cerclage placement [25]. Potential toxic fetal effects (eg, oligohydramnios, premature closure of the ductus arteriosus) from in utero exposure to indomethacin typically occur with exposures >48 hours and with increasing gestational age after 20 weeks, with the greatest risk after 30 weeks. These risks are discussed separately. (See "Inhibition of acute preterm labor", section on 'Fetal side effects'.)

Other pharmacotherapy

●Group B Streptococcus – In patients colonized with group B Streptococcus, antibiotic prophylaxis for obstetric procedures such as cerclage is not recommended, although this has not been studied directly. (See "Prevention of early-onset group B streptococcal disease in neonates", section on 'Patients undergoing obstetric procedures'.)

●Progesterone – The author does not initiate use of progesterone supplementation pre- or postoperatively, but continues progesterone supplementation postoperatively in patients who are already on the drug. These data are reviewed separately. (See "Cervical insufficiency", section on 'Cerclage placement and use of progesterone supplementation'.)

Anesthesia — Neuraxial anesthesia is preferred over general anesthesia for its overall safety, but either method is acceptable. (See "Anesthesia for nondelivery obstetric procedures", section on 'Choice of anesthetic technique'.)

If additional uterine relaxation is needed during the procedure, such as when the membranes are prolapsed, it can be achieved with intravenous nitroglycerin.

Preparation

●Ensure that the bladder is empty before beginning the procedure to enhance exposure; however, placement of a catheter for urinary drainage during the procedure is unnecessary.

●Prepare the vagina and cervix with an antiseptic or saline solution, avoiding contact with the fetal membranes if exposed.

●The availability of an assistant to help expose the operative field is useful and generally essential for physical examination-indicated procedures.

Replace prolapsed membranes, if present — Prolapsed fetal membranes substantially increase the risk for iatrogenic PPROM, which occurs in up to 65 percent of cases [2-5]. If cerclage placement is attempted in this setting, the prolapsed membranes must be replaced in the uterine cavity before applying the cerclage. The optimum technique has not been established by randomized trials. Our approach follows:

●Place the patient in steep Trendelenburg position to allow gravity to retract the membranes.

●If unsuccessful, less invasive options include:

•Administer a uterine relaxant (eg, nitroglycerin).

•Backfill the bladder in 250 mL increments through a bladder catheter [26], although a full bladder tends to reduce exposure of the operative field and pull the cervix deeper into the pelvis.

•Place ring forceps or stay sutures of 00 silk around the circumference of the external os and then gently pull and shake the cervix to help ease the membranes back into the uterus (figure 1).

•Push the membranes back with a smooth surfaced device, such as a gloved finger, sponge-filled condom, or bladder catheter balloon (a specialized balloon device may be available for this purpose [27]). A 30 mL bladder catheter can be used to hold the membranes in the uterus while the cerclage is placed; it is deflated and removed just before the knot is secured. However, such techniques may be associated with an increased risk of membrane rupture.

●If less invasive options are unsuccessful, transabdominal amniocentesis with amnioreduction under ultrasound guidance reduces amniotic fluid volume and pressure in the prolapsed sac, allowing it to retract back into the uterine cavity [28,29]. Approximately 150 to 250 mL is typically removed.

Technical goals — The primary technical goal of cerclage placement is to reinforce the cervix at the level of the internal os; lengthening the cervix is a secondary benefit [30-34]. In one study, after cerclage placement, longer upper cervical length (the length of closed cervix between the cerclage and internal os), but not total cervical length, correlated with birth after 28 and after 32 weeks of gestation [35]. Although another study reported that cervical height (ie, the length of cervix between the cerclage and external os) was not an important factor in determining subsequent pregnancy outcome [36], an expert review concluded that achieving a cerclage height >20 mm reduced preterm birth compared with shorter cervical height [37].

Choosing a McDonald versus Shirodkar procedure — The two most common transvaginal techniques for cerclage were described by Shirodkar [38] and McDonald [39]; modifications of both procedures have also been described. The Shirodkar cerclage is placed as close as possible to the level of the internal os after surgically reflecting the bladder anteriorly and the rectum posteriorly, whereas the McDonald cerclage is a purse-string suture that does not involve any dissection (thus, theoretically, it cannot be placed as close to the internal cervical os as the Shirodkar). In contrast to the McDonald cerclage, the Shirodkar suture does not pass through the cervical stroma.

The author's preference is to perform the Shirodkar procedure because it can be placed closer to the internal os, does not pass through the cervical stroma, and he finds it technically easier to place in patients with advanced cervical effacement. Others prefer the McDonald procedure because they find it easier to perform and remove.

The body of data shows no significant differences in pregnancy outcome between the two procedures [20,37,40-51]. However, no randomized trial has compared the McDonald and Shirodkar procedures directly. One study observed McDonald and Shirodkar procedures had similar obstetric outcomes in patients undergoing their first cerclage, but higher birth weight when Shirodkar rather than McDonald cerclage was performed for the second procedure (3020 and 2470 grams, respectively) [9].

Technique — Techniques for the Shirodkar and the McDonald cerclage are described below. (See 'Shirodkar cerclage' below and 'McDonald cerclage' below.)

●Needle type – The optimum type of needle has not been evaluated in randomized trials. The choice should be based on general surgical principles and the operator's experience and preference.

●Choice of suture – In a multicenter randomized trial comparing monofilament versus braided sutures for performing history- or ultrasound- indicated transvaginal cerclage, the frequency of pregnancy loss was similar in both groups (80 in 1003 [8.0 percent] versus 75 in 999 [7.6 percent]; adjusted RR 1.05, 95% CI 0.79-1.40) [52]. The type of cerclage was at the discretion of the surgeon; 17 percent of patients in each group underwent bladder dissection, suggesting a Shirodkar procedure was performed. Pregnancy loss was defined as miscarriage and perinatal mortality, including any stillbirth or neonatal death in the first week of life.

Use of a monofilament suture resulted in lower rates of clinical chorioamnionitis (2.7 versus 6 percent; adjusted RR 0.45, 95% CI 0.29-0.71) and maternal sepsis (3.9 versus 6.8 percent; adjusted RR 0.58, 95% CI 0.40-0.82), but no statistically significant differences in the rate of suspected or confirmed neonatal sepsis, mean gestational age at delivery, or rate of preterm birth <28, 32, or 37 weeks. It is unclear why the increased risk for chorioamnionitis in the braided suture group did not translate into an increased risk for pregnancy loss and preterm birth in that group, as these complications are consequences of chorioamnionitis. The authors hypothesized that the finding might have been related to bias among the outcome assessors.

●Knot position – The optimum knot position (anterior versus posterior) has not been evaluated in randomized trials. The choice should be based on general surgical principles and the operator's experience and preference.

Shirodkar cerclage — This procedure is more complicated than the McDonald cerclage because it requires incisions and dissection of the paracervical area. The author's approach is described below.

●The cervix is pulled toward the surgeon with one or two ring forceps while an assistant retracts the vaginal sidewalls.

●1 to 2 mL of sterile saline is injected into the submucosa to raise a wheal before the incision is made to facilitate dissecting tissue planes.

●A scalpel with a number 10 blade or an electrocautery needle is used to make a 1 to 3 cm vertical or transverse incision on the posterior cervix at the junction of the rugated vaginal epithelium and the smooth cervix (figure 2). The author prefers a 2 to 3 cm transverse incision. Making the first incision posteriorly prevents the operative field from becoming obscured by bleeding from the anterior incision, which can occur even with use of electrocautery.

●After the posterior incision is made, a transverse incision is made anteriorly.

●The rectum is bluntly dissected off the posterior cervix and the bladder is bluntly dissected off of the anterior cervix using a finger, sponge on a stick, or peanut sponge on a long clamp. The dissection should be carried back far enough to allow the surgeon to palpate the insertion of the uterosacral and cardinal ligaments onto the cervix at the level of the internal os. Electrocautery can be used to control small bleeders.

●Long curved Allis clamps, or similar tissue forceps clamps (eg, Teale Vulsellum), are used to grasp and approximate the lateral edges of the anterior and posterior aspects of the transverse incisions and some paracervical tissue.

●Two atraumatic (blunted) needles premounted with a single 5 mm polyester braided tape (Mersilene tape) are used for the cerclage, but polyester braided or polypropylene nonbraided monofilament suture can be used instead [52,53]. Bending the needle to reduce the curvature is sometimes helpful for guiding the needle to the desired position. The tip of one needle is introduced anteriorly at the lateral edge of the incision at the level of the internal os (or as close as possible) and threaded submucosally adjacent to the cervical stroma (and medial to the cervical branches of the uterine vessels) to emerge at the lateral edge of the posterior incision at the level of the internal os. If not at the internal os, the cerclage should be at least 2 cm cephalad to the external os, as feasible [54]. Intraoperative ultrasound can be helpful for judging the site of the suture relative to the internal os, maternal bladder, and rectum [55]. (See 'Number of cerclages' below.)

●The procedure is then repeated on the opposite side and the two ends are tied tightly using four to seven square knots. It is the surgeon's preference as to whether the stitch is placed so that the knot is tied anteriorly or posteriorly. The author usually finds it easier to perform the cerclage with anterior placement of the knot, although there are rare instances of an anterior knot causing bladder discomfort, and even eroding into the bladder. It is also easier to remove the cerclage in the office if the knot is anterior to the cervix.

●The Mersilene tape is cut 2 to 3 cm in length and then tagged with 2-0 silk that is left long. If a suture is used, the ends are left longer.

●The epithelium may be reapproximated with a fine chromic catgut suture, although this is not necessary if good hemostasis is achieved. It is not necessary to bury the ends of the knot under the epithelium or anchor the tape to the cervix. Avoiding burying and anchoring the cerclage facilitates removal prior to birth. Alternatively, if cesarean birth is planned, the Shirodkar cerclage can be left in-situ indefinitely postpartum for use in a future pregnancy. It is advantageous to completely bury the knot under the vaginal epithelium in these cases to minimize vaginal discharge.

McDonald cerclage — The procedure is begun by grasping the anterior and posterior lips of the cervix with one or two ring forceps. We insert a curved needle loaded with large caliber nonabsorbable synthetic suture (at least number 1 or 2 braided or monofilament) at 12 o'clock, at the junction of the rugated vaginal epithelium and the smooth cervix just distal to the vesicocervical reflection and at least 2 cm above the external os, as feasible [54]. Alternatively, the needle can be inserted posteriorly at the cervicovaginal reflection at approximately 6 o'clock. As with the Shirodkar cerclage, there is no evidence that knot position (anterior versus posterior) [56] or choice of suture (monofilament versus braided) [52] affects pregnancy loss rate.

Four to six deep bites of a purse-string suture are taken circumferentially around the entire cervix as high (close to the internal os) as safely possible, avoiding the bladder, rectum, and uterine vessels (at 3 and 9 o'clock). Approximately 1 cm of space is left between the exit of one deep bite and the entry of the next deep bite. Each deep bite should extend at least midway into the cervical stroma to reduce the risk that the suture will pull out over time, but should not enter the endocervical canal (figure 3). The pass at 6 o'clock is particularly important because this is the most common site for pull-through [39]. The two ends of the suture are then tied securely and cut, leaving the ends long enough to grasp with a clamp when it is time to remove it. Intraoperative ultrasound can be helpful for judging the site of the suture relative to the internal os, maternal bladder, and rectum [55].

Number of cerclages — A single cerclage is usually adequate, if well placed. In some cases, an inadequate initial cerclage is used for traction, and then a second cerclage is placed in a more optimal position closer to the internal os. A second cerclage may be needed to achieve adequate closure of the cervix when the procedure is performed on a widely dilated cervix with prolapsed membranes. If two cerclages are placed, they are generally removed at the same time.

Although some clinicians routinely place a second cerclage, this practice did not improve outcome in three retrospective studies [57-59]. In addition, a randomized trial found that placing a second stitch at the external os to keep the mucus plug in place (termed cervical occlusion) did not increase gestational age at delivery or decrease neonatal intensive care unit days or neonatal mortality [60].

POSTOPERATIVE CARE AND FOLLOW-UP

Counseling and care

●Hospital discharge – Cerclage is typically an outpatient surgery procedure. Most patients can be discharged after recovery from the anesthetic and when able to ambulate and void; however, a longer period of in-hospital observation may be indicated for some patients who undergo physical examination-indicated cerclage because of their increased risk for complications. (See 'Complications' below.)

●Documentation – The fetal heart rate and normal amniotic fluid volume should be documented prior to discharge.

●Analgesia – Acetaminophen alone provides adequate analgesia for most patients.

●What to expect – Patients are told to expect some spotting, cramps, and dysuria (due to minor muscle injury from the vaginal wall retractors) which will abate within a few days. Those who have undergone a Shirodkar procedure may note passage of the fine chromic catgut in two to three weeks as the stitches dissolve; they should be forewarned that this does not represent loss of the cerclage itself.

Patients who have undergone cerclage placement have an increased frequency of uterine contractions [42], but the presence of uterine irritability is not predictive of an increased risk for preterm birth.

●When to call the clinician – Patients should report any leakage of fluid from the vagina so that they can be evaluated for membrane rupture and any vaginal bleeding other than spotting.

●Activity – Although there is no evidence that coitus adversely affects perinatal outcome, the author asks patients to maintain pelvic rest for at least one week after a history-indicated procedure and to use condoms thereafter (semen contains prostaglandins). Patients who have had an ultrasound- or physical examination-indicated cerclage are managed more conservatively; the author typically asks them to limit physical activity and coitus until a favorable gestational age is reached, usually 32 to 34 weeks of gestation, although there is no high-quality evidence that decreasing these activities improves outcome.

Follow-up — Patients followed as outpatients are seen on a regular basis with frequent (weekly or biweekly) visits for cervical examination. The author finds ultrasound assessment of cervical length useful for identifying those patients at highest risk for preterm birth [61-65]. Proximal cervical shortening after 23 weeks may affect subsequent management including advising a reduction in physical activity, increased frequency of prenatal visits, and increased concern about the potential for preterm birth, which might involve prompt administration of antenatal corticosteroids for fetal lung maturation and, if birth is imminent, magnesium sulfate for cerebral palsy prophylaxis. (See "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery" and "Neuroprotective effects of in utero exposure to magnesium sulfate".)

The author does not routinely follow patients with serial fetal fibronectin assays, as the value of testing asymptomatic patients is unproven. If performed, the test should not be done until at least four weeks postoperatively and a positive result may be less reliable in this setting [66,67].

CERCLAGE REMOVAL

●Timing – The cerclage is removed at 36+0 to 37+0 weeks of gestation in anticipation of labor or immediately upon onset of preterm labor to avoid cervical laceration and/or uterine rupture. The absolute risk of laceration is unclear given the small number of reports [68]. Whether to remove the cerclage in the setting of preterm prelabor rupture of membranes (PPROM) is controversial. (See 'Removal of cerclage after PPROM' below.)

In contrast to a McDonald cerclage, a Shirodkar cerclage does not have to be removed if cesarean birth is anticipated and future pregnancies are planned. However, there is a theoretical risk of reduced fertility from inflammation/infection of the cervix due to the foreign body and a risk of erosion into adjacent tissue.

●Site of removal – A McDonald cerclage usually can be cut and removed in the office without analgesia. A Shirodkar cerclage often requires a return to the operating room for removal, either because the knot is buried under the vaginal epithelium or the Mersilene tape has been infiltrated by cervical granulation tissue.

●Follow-up – Patients are generally sent home after cerclage removal to await the onset of labor, which generally occurs within two weeks; only an approximate 10 percent deliver spontaneously within 48 hours of planned cerclage removal [69].

Removal of cerclage after PPROM — Whether to remove the cerclage if PPROM occurs is a matter of debate. One concern is that removal will lead to earlier birth; however, retention of the foreign body may increase the risk of infection, leading to preterm birth and maternal and newborn morbidity. Several observational studies have addressed the management of such patients, with inconsistent results [43,70-78]. Gestational age at PPROM appears to be the most important determinant of neonatal outcome [74]. The only randomized trial of cerclage removal versus retention after PPROM found no statistically significant differences between groups in any pregnancy outcome, but the trial was terminated early after only 56 of the proposed 142 patients had been recruited, and was underpowered for all of the outcome measures [79].

Based on the available limited data and personal clinical experience, the author removes the cerclage in patients with PPROM when:

●There is any evidence of chorioamnionitis, or

●The pregnancy is at least 32 weeks of gestation. Before 32 weeks, in the absence of clinically apparent infection, the cerclage is left in place as there is greater concern about the possible increased risk of neonatal morbidity and mortality from preterm birth with cerclage removal than the possible increased risk of ascending infection with the cerclage left in place.

COMPLICATIONS — The frequency of complications is higher with increasing gestational age and cervical dilation, but not higher with McDonald versus Shirodkar cerclage. In a study including 38 history-indicated cerclages, 96 ultrasound-indicated cerclages, and 145 physical examination-indicated cerclages, the incidence of perioperative complications differed significantly: 7.9, 10.4, and 27.6 percent, respectively [80]. Only physical examination-indicated procedures were associated with serious complications (prelabor rupture or membranes within two weeks of the procedure in 27 cases [18.6 percent]), particularly when membranes prolapsed into the vagina.

●Membrane rupture — After a history-indicated cerclage, the median frequency of preterm prelabor rupture of the fetal membranes (PPROM) has been reported to be approximately 2 percent [2]. Rupture of membranes intraoperatively or in the immediate postoperative period is a concern with ultrasound-indicated cerclage and a major concern of physical examination-indicated cerclage, especially with advanced cervical dilation and/or prolapsed fetal membranes where this complication has been reported in up to 65 percent of procedures [2-5]. In the absence of perioperative rupture, the risk for prelabor rupture of membranes at <34 weeks of gestation may not differ by cerclage indication and was about 10 percent in one study [81].

●Intraamniotic infection — The median frequencies of intraamniotic infection after history-indicated and nonhistory-indicated cerclage are 2 and 25 percent, respectively [2]. Maternal sepsis has been reported after history-indicated, ultrasound-indicated, and physical examination-indicated cerclage [82].

●Suture migration — Suture migration has been reported in 3 to 13 percent of cases [83]. This often occurs late in gestation and thus is of little clinical consequence. When migration occurs early in pregnancy (before 24 weeks), the clinician must decide whether to repeat the procedure. It is likely that the same factors that led to failure of the first cerclage will affect a second procedure. In fact, placement of a second (reinforcing) cerclage may worsen the outcome when cervical shortening has occurred in the presence of a history-indicated cerclage [84], and is not generally recommended [21]. Hospitalization and/or a course of antenatal glucocorticoids are options if a preterm potentially viable birth appears likely. (See "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery".)

●Other — Cervical dystocia and cervical trauma in labor have been reported in fewer than 5 percent of patients; however, uterine rupture as a consequence is rare [9,85]. Excessive bleeding and fistula formation are also rare.

PREGNANCY OUTCOME — Cervical cerclage is the conventional treatment for cervical insufficiency, despite the paucity of data from randomized trials proving its efficacy. Most case series report a live birth rate of 70 to 90 percent after history-indicated cerclage, as compared with 10 to 30 percent prior to the procedure [86]. However, using patients as their own controls (ie, pregnancy success rate is compared before and after cerclage) is subject to bias since changes in the patient and her management other than cerclage may have accounted for the higher rate of success in the subsequent pregnancy. In fact, trials in which patients were randomly assigned to undergo cerclage or no cerclage report much higher live birth rates in the untreated group than observed in historic controls [87,88].

The timing of cerclage also affects outcome. There is universal agreement that a physical examination-indicated cerclage performed in the presence of advanced cervical change and prolapsed membranes has the worst outcome. It should be appreciated that there is significant risk that an emergency procedure may convert a previable birth to a very low birth weight preterm birth (24 to 27 weeks) with the potential for serious long-term neurodevelopmental disability. Thus, outcomes at the limit of viability and the potential options should be clarified with the patient; discussion with the neonatology service cane be helpful. (See "Periviable birth (limit of viability)".)

●In a 2017 meta-analysis of randomized trials of cerclage versus no cerclage in singleton pregnancies at high risk of pregnancy loss based on history and/or ultrasound finding of short cervix and/or physical examination, placement of a cerclage [89]:

•Reduced preterm birth <34 weeks (18 versus 24 percent; risk ratio [RR] 0.77, 95% CI 0.66-0.89; nine trials, 2415 pregnancies; high-quality evidence)

•Probably reduced perinatal loss (7.5 versus 9.2 percent; RR 0.82, 95% CI 0.65-1.04; 10 trials, 2927 pregnancies, moderate-quality evidence)

There were too few trials for each of the indications to evaluate whether there were differences in efficacy by indication.

Pregnancy outcome with each of the three types of cerclage is reviewed in more detail separately (see "Cervical insufficiency", section on 'Obstetric history-based cervical insufficiency' and "Cervical insufficiency", section on 'Ultrasound-based cervical insufficiency' and "Cervical insufficiency", section on 'Physical examination-based cervical insufficiency') Pregnancy outcome with cerclage in twin pregnancies is also reviewed separately. (See "Twin pregnancy: Management of pregnancy complications", section on 'Patient selection for cerclage'.)

SUMMARY AND RECOMMENDATIONS

●Goal – Cervical cerclage refers to a variety of surgical procedures in which synthetic suture or tape, or wire, is used to reinforce the cervix and thereby prevent pregnancy loss/early preterm delivery. (See 'Introduction' above and 'Goal' above.)

●Timing and contraindications – Most cerclages are placed between 12 and 24 weeks of gestation. The major contraindications are clinical scenarios where the procedure is unlikely to reduce the risk of preterm birth or improve fetal outcome, such as fetal anomaly incompatible with life, intrauterine infection, active preterm labor, preterm prelabor rupture of membranes (PPROM), fetal demise, or active uterine bleeding (eg, placental abruption). (See 'Upper and lower gestational age thresholds for cerclage placement' above and 'Contraindications' above.)

●Preoperative assessment

•An ultrasound examination is performed proximate to the procedure to determine if there are identifiable fetal structural anomalies that might affect the patient's decision to continue the pregnancy and to confirm gestational age and viability. Aneuploidy screening is offered, if not already performed. (See 'Fetal assessment' above.)

•Before ultrasound- and physical examination-indicated cerclage, the patient is monitored for up to 24 hours before proceeding with surgery to help exclude the presence of PPROM, occult abruption, infection, and preterm labor. (See 'Observation period before unplanned procedures' above.)

●Pharmacotherapy (See 'Prophylactic antibiotics and tocolytics' above.)

•History-indicated cerclage – Prophylactic antibiotics or tocolytics are not administered since infection and uterine contractions/irritability are uncommon.

•Ultrasound-indicated cerclage – We do not administer prophylactic antibiotics or tocolytics since infection and uterine contractions/irritability are uncommon, but acknowledge that the risk is slightly higher than in history-indicated procedures and that practice patterns vary in this setting.

•Physical examination-indicated cerclage – We suggest antibiotic and indomethacin prophylaxis to prolong latency (Grade 2C). These patients are at high risk of developing infection and postprocedure uterine contractions/irritability. We use cefazolin and indomethacin.

●Procedure

•The choice of procedure should be based on general surgical principles and the operator's experience and preference. The body of data shows no significant differences in outcome or complications between McDonald (figure 3) and Shirodkar (figure 2) procedures. The frequency of complications is higher with increasing gestational age and cervical dilation. (See 'Choosing a McDonald versus Shirodkar procedure' above and 'Complications' above.)

•A single cerclage is usually adequate. (See 'Number of cerclages' above.)

•If suture migration subsequently occurs early in pregnancy (before 24 weeks), we suggest not placing another cerclage (Grade 2C). (See 'Complications' above.)

•Rupture of membranes intraoperatively or in the immediate postoperative period occurs primarily in patients with advanced cervical dilation and/or prolapsed fetal membranes (reported in up to 65 percent of physical examination-indicated cerclages). (See 'Complications' above.)

●Removal – The cerclage is removed electively at 36+0 to 37+0 weeks of gestation or immediately upon onset of preterm labor. (See 'Cerclage removal' above.)

●Prelabor rupture of membranes – In otherwise asymptomatic patients ≥32 weeks of gestation with prelabor rupture of membranes, we suggest removing the cerclage rather than expectant management (Grade 2C). Before this gestational age, we feel the possible increased risk of preterm birth with cerclage removal is a greater concern than the possible increased risk of ascending infection if the cerclage is left in place. (See 'Removal of cerclage after PPROM' above.)

●Effectiveness – In singleton pregnancies at high risk of pregnancy loss based on history and/or ultrasound finding of short cervix and/or physical examination, placement of a cerclage reduced preterm birth <34 weeks by 23 percent and reduced perinatal loss by 18 percent. However, outcomes vary by the clinical setting at the time of placement. (See 'Pregnancy outcome' above.)