Unfortunately, there is no clear consensus for the diagnosis of CI despite its relatively common occurrence. Recurrent, relatively painless, second-trimester fetal losses or preterm births in the absence of contractions or vaginal bleeding remains the gold standard for diagnosis. Therefore, it is crucial that a careful and thorough obstetric and gynecologic history is taken at the onset of prenatal care. Although this presentation is classic, it is clearly recognized that cervical competence is a continuous variable and cannot simply be categorized as “competent” or “incompetent.” For example, contractions may be present as a late sign of CI after prolonged exposure of the membranes to the vaginal flora. Furthermore, with the increasing use of transvaginal ultrasound, cervical shortening and dilation of the internal os can be recognized prior to the onset of symptoms and even without digital examination of the cervix.

Transvaginal sonography of the cervix can be easily performed as early as 14–16 weeks of gestation, when the lower segment of the uterus is developed well enough to allow reproducible measurements of cervical length and architecture of the internal os. Ultrasound cervical length screening at 18–22 weeks’ gestation has been proposed. A length of less than 25 mm is generally considered shortened and suggestive of CI, although the probability of preterm delivery at a given cervical length varies according to gestational age. Also important is any dynamic change noted after Valsalva or mild fundal pressure, as well as differences in cervical appearance or length between consecutive measurements. Other changes which have been described include dilation and funneling of the internal os.

In nonpregnant patients, several tests including physical examination and ultrasound or radiographic studies are available to facilitate the diagnosis of CI. On physical examination, the easy passage of a number 8 Hagar dilator or a number 15 Pratt dilator is essentially diagnostic of CI. Hysterosalpingography showing dilation of the internal os to greater than 6 mm is also diagnostic of CI. These tests are both inconvenient and of limited diagnostic utility as they often yield equivocal results in patients with an unclear history.

Although there are several postulated causes of CI, it is believed that ascending intrauterine infection with inflammation and cervical trauma are the most common causes. Cervical trauma most commonly results from surgical interventions including conization, loop electrosurgical excision procedures, repetitive or second-trimester therapeutic terminations, and obstetric injuries. More than one first-trimester termination or a single second-trimester termination increases the incidence of CI. Obstetric injuries include compression necrosis of the cervix due to a prolonged second stage of labor, and spontaneous as well as iatrogenic lacerations of the cervix such as Duhrssen’s incisions performed during vaginal delivery. Further obstetric injury may include extension of the uterine incision into the cervix at the time of cesarean section.

Congenital defects including mullerian anomalies, exposure to diethylstilbestrol, maternal deficiencies in elastin or collagen polymorphisms are less common causes of CI. Rarely, acquired anatomic defects such as large polyps or cervical myomata may be associated with CI. Premature signaling of “cervical ripening,” molecular signals from fetal, trophoblast or maternal sources, is increasingly studied and may explain inconsistent CI in consecutive pregnancies in the same mother.

Once a diagnosis of CI has been established or the need for intervention is identified, treatment has traditionally been by surgical correction using an encircling or cerclage suture. The most commonly used technique is the McDonald cerclage, which was described in 1957. With this, a purse-string suture of four or five bites is placed around the cervix. The material most commonly used is 5 mm Mersilene tape. Mersilene provides better tensile strength and is less likely to pull through the cervix in later gestation. Prolene and nylon are both more easily passed through the cervical tissue but are also more likely to pull through the tissues, given their smaller calibers. The suture is placed below the level of the internal os and must be placed deep into the substance of the cervix to prevent lacerations. The lateral blood supply should be left outside the purse-string suture. The knot is then placed anteriorly and one end left long enough to facilitate removal at 36–37 weeks’ gestation. The McDonald technique differs from the modified Shirodkar cerclage described below in that more suture is left exposed within the vagina.