REVIEW ARTICLE

From Concept to Practice: The Recent

History of Preterm Delivery Prevention.
Part I: Cervical Competence
Alex C. Vidaeff, M.D., M.P.H.,1 and Susan M. Ramin, M.D.1

ABSTRACT

The prevention of preterm delivery (PTD) is a major desiderate in contemporary

obstetrics and a societal necessity. The means to achieve this goal remain elusive, and the
research efforts have been punctuated by several ineffective intervention proposals. More
recently, new areas of proposed preventive strategy have arisen, focusing on cervical
competence, subclinical infection, and hormonal effects. This review, based on a compre-
hensive and unbiased review of the available literature, will address the rationale and
current status of these new concepts of significant clinical interest. We will describe the
microcosm of thought and research dedicated to the prevention of PTD during the last 10
years, and the arduous efforts to establish a linkage between predictive observations and
therapeutic hypotheses. In its first part, the review will discuss the recently emerging view
that regardless of the preterm labor etiology, a common pathway is eventually reached,
reflected in cervical changes. The new concept of functional cervical insufficiency is
addressed from the perspective of cervical assessment by ultrasound. Although the existing
research has not accurately quantified yet the usefulness of ultrasound cervical examination,
the technique has become commonplace. Several recent studies have demonstrated a
continuum of risk between shorter cervix on ultrasound and higher rate of PTD, leading to
the hypothetical argument that women with short cervix on ultrasound might benefit from
cervical cerclage. Observational and randomized clinical trials of cerclage as a modality of
pregnancy prolongation have provided conflicting results. The relative paucity of data and
the conflicting nature of the available evidence dictate caution whenever decisions for
cervical monitoring or intervention are made.

KEYWORDS: Preterm delivery, prediction, prevention, cervical insufficiency, cervical

length, cerclage

T he prevention of preterm delivery (PTD) is a
major desiderate in contemporary obstetrics and a societal
necessity. In 2002, 12% of deliveries in the United States
were premature.1 During the last 10 years, the incidence
of PTDs in Europe and the United States has remained
unchanged or has been increasing, frustrating patients
and clinicians alike. The means to achieve prevention
remain elusive, and the research efforts have been punc-
tuated by several ineffective intervention proposals. En-
hanced prenatal care, including frequent office visits and

American Journal of Perinatology, Volume 23, Number 1, 2006. Address for correspondence and reprint requests: Alex C. Vidaeff, M.D., Associate
Professor, Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology, and Reproductive Sciences, University of Texas–Houston
Medical School, 6431 Fannin, Suite 3.604, Houston, TX 77030. 1Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology, and
Reproductive Sciences, University of Texas–Houston Medical School, Houston, Texas. Copyright # 2006 by Thieme Medical Publishers, Inc., 333
Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662. DOI 10.1055/s-2005-923437. Published online December 21, 2005. 0735-
1631,p;2006,23,01,003,014,ftx,en;ajp51930x.
3
4 AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 23, NUMBER 1
frequent cervical examinations, patient education,
cessation of work, bed rest, home uterine monitoring,
prophylactic tocolytics, nutritional supplements, and
social support, have all been advanced, at one time or
another, as prophylactic interventions, but eventually
could not be associated with any consistent evidence of
benefit.
More recently, there have been at least three new
areas of proposed preventive strategies focusing on
cervical competence, subclinical infection, and hormonal
effects. This article, published in two parts, reviews the
rationale and current status of these new concepts of
significant clinical interest. We describe the microcosm
of thought and research dedicated to the prevention of
PTD during the last 10 years, and the arduous efforts to
establish a linkage between predictive observations and
therapeutic hypotheses. From Concept to Practice is the
scientific process connecting hypothesis-generating re-
sults to hypothesis-testing studies.
The public health principles of primary, secon-
dary, and tertiary health care are applicable to the study
of preterm labor and delivery. Because many factors,
known and unknown, may be responsible for initiation
of preterm labor, identifying them and providing pri-
mary preventive strategies (prevention of risk) in the
entire pregnant population is difficult, if not impossible.
Tertiary care, which represents treatment provided to
symptomatic individuals, is notoriously ineffective in
preterm labor. At present, the use of tocolytics, cortico-
steroids, antibiotics, and bed rest, while it may reduce
morbidity, does not prevent PTD.
As opposed to primary and tertiary care for
preterm labor, the secondary level of care has obvious
intuitive appeal. Its goal is to identify women at risk and
to administer care that minimizes their risk. However,
the number of risk factors for preterm labor, and their
relative importance varies among pregnancies, even in
the same woman. Screening strategies based on early
identification of risk factors in otherwise asymptomatic
gravidas have proved unsuccessful. Further complicating
the secondary care attempt is the fact that more than half
of preterm deliveries occur in women who have no
apparent risk factors. To date, no such screening and
subsequent intervention strategy has yet been found to
be consistently effective. Given the complexity of PTD
pathophysiology, it is unlikely that a single test and
intervention will ever be effective.
THE NEW CERVICAL INSUFFICIENCY
POSTULATE
Cervical insufficiency (formerly known as cervical incom-
petence), characterized by early cervical changes in the
absence of uterine contractions, has been traditionally
viewed as the expression of an anatomically defective
cervix, congenital or acquired. A more recently emerging
2006
view is that regardless of the preterm labor etiology, a
common pathway is eventually reached, reflected in
cervical changes. The common pathway may be an
inflammatory response with upregulation of cytokines,
prostaglandins, matrix metalloproteinases, and possibly
other local factors, all of them resulting in premature
cervical ripening. It has been observed that painless
cervical changes are invariably, sooner or later, associated
with premature myometrial contractions if premature
delivery occurs. In contrast to the traditional view, with
focus on anatomical and mechanical factors, the new
concept alludes more to a functional cervical insuffi-
ciency.2–4 The mechanical etiology is not excluded, but
seems to be limited to a restricted population subset,
representing only 10 to 25% of second-trimester losses.5
For this subset, the idea of cerclage was advanced in the
early 1900s.6 Under the new cervical insufficiency postu-
late, the final common pathway may be influenced
theoretically, independently or interdependently, by mul-
tiple interventions including not only cerclage, but also
antibiotics, anti-inflammatory drugs, or progesterone.
Another facet of the new concept is the perception
of cervical competence as an extended biological contin-
uum, with progressing degrees of insufficiency, replacing
the traditional dichotomous view of cervical competence
as being present or absent. Only the ultrasound assess-
ment of the cervix has made possible this change in
paradigm, and cervical length measurement has emerged
as the surrogate marker for cervical competence.
Figure 1 Schematic illustration of a sagittal view of the cervix
indicating the presence of funneling and the measurement of the
residual cervical length.

Figure 2 Transvaginal ultrasound image of a normal cervix: length 3.8 cm, and no internal os changes.
Cervical assessment by ultrasound provides infor-
mation unobtainable by digital examination. The cervix
has an upper and lower portion, with the vesicouterine
fold at the lowermost edge of the bladder demarcating
them. Consequently, up to half of the cervix is usually
located in the abdomen rather than vagina. When
cervical changes occur, the upper cervix changes first,
undetected by digital examination. When cervical efface-
ment is appreciable in the vagina, the upper cervix has
already disappeared, significantly shortening the endo-
cervical canal. It has also been observed that cervical
shortening precedes the internal os changes (dilatation
and funneling; Fig. 1). With the opening of the internal
os, the risk of PTD increases. Although it has been
established that shortening of the cervix and the internal
os changes are significant predictors of PTD, the same
findings remain as questionable discriminators, given
their low specificity. Many gravidas have more advanced
changes, such as cervical dilatation up to 2 cm, detectable
even on digital examination, and many of them do not
deliver prematurely.
Figure 3 Transvaginal ultrasound image of a normal cervix (length 3.0 cm), with a posterior lower segment contraction generating the
false impression of funneling.
RECENT HISTORY OF PTD PREVENTION/VIDAEFF, RAMIN 5
Although the existing research has not yet
accurately quantified the usefulness of ultrasound
cervical examination, the technique has become com-
monplace.
THE TECHNIQUE OF ULTRASOUND
CERVICAL ASSESSMENT
The use of a transvaginal probe for cervical assessment is
considered more accurate than transabdominal scanning,
given the close proximity to the cervix and less cervical
distortion from transducer pressure or a full bladder. The
probe is inserted into the anterior vaginal fornix until the
isthmic portion of the uterus and the internal os are
identified. The probe is rotated to obtain a sagittal (long
axis) view of the cervix, and then slowly withdrawn to
relieve any pressure, until the image begins to blur. At
this moment the probe is advanced slightly. An ideal
image shows the external os as a small echodense notch,
the internal os as a small V-shaped notch, and a faint
echodense line between them (Fig. 2). The measurement
6 AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 23, NUMBER 1
is from notch to notch. The distal notch is easier to
identify if ultrasonic gel is applied on the cervix by prior
digital examination or with the probe.
The measurements should be performed with an
empty bladder. An overdistended bladder or too much
pressure with the probe can artificially elongate the
cervix. A full bladder can also obscure the internal os
dilatation.7 Contractions of the lower uterus can also
lead to false elongation and false funneling (pseudofun-
neling; Fig. 3). It is recommended that at least three
cervical measurements be performed, 5 minutes apart, to
identify temporary dynamic changes, and the shortest
adequate measurement be reported.
In cases with premature rupture of membranes, or
placenta previa with active bleeding, transperineal ultra-
sound can be used instead. There is a high degree of
agreement between transperineal and transvaginal meas-
urements, although, given the distance from the cervix,
the images tend to be less clear with transperineal
scanning.
CERVICAL LENGTH AS A BIOLOGICAL
PARAMETER
The average length of the nonpregnant cervix by trans-
vaginal ultrasound in premenopausal women has been
reported to be between 3.1 and 3.8 cm.8,9 In the
pregnant uterus at 11 weeks, Zalar10 found a longer
median cervical length than in the nonpregnant state
(4.6 cm). It appears that the isthmic uterine portion, by
not distending with the rest of the corpus, adds to the
cervical length in the first trimester. The same author
considered that women with a cervical length in the first
trimester > 4.5 cm may be at a lower risk for PTD.
Regarding the significance of a shorter cervix so early in
pregnancy there are no clear data, and cervical measure-
ments before 13 to 16 weeks may not be helpful.
After the first trimester, others have found a mean
cervical length of 4.2 cm at 14 weeks, 3.7 to 3.8 cm at
15 to 24 weeks, 3.6 cm at 26 weeks, 3.5 cm at 27 to
28 weeks, and 3.3 cm at 29 to 30 weeks11,12 (Table 1).
Table 1 Mean Cervical Length in the General
Population
Pregnancy Status Cervical Length (cm)
Nonpregnant premenopausal 3.1–3.8
Pregnant (wk)
11 4.6
14 4.2
15–24 3.7
26 3.6
27–28 3.5
29–30 3.3
Based on data from Jackson et al, Zalar,10 Guzman et al,11 and
8 a cervix of  2.5 cm at 22 to 24 weeks gestation had a
Okitsu et al.12
2006
The average cervical length reported in the literature was
not influenced by parity.13,14 The normal cervical length,
as any other biological parameter, is normally distributed
(along a bell-shaped curve), with the mean (the 50th
percentile) at 20 to 28 weeks gestation of approximately
3.5 cm (Fig. 4).
CERVICAL LENGTH AND PTD
In a landmark study in 1996, Iams et al15 demonstrated
that the risk of PTD is inversely proportional to the
length of the cervix on transvaginal sonography between
24 and 28 weeks gestation in an unselected U.S. pop-
ulation. The same observation was verified subsequently
in earlier pregnancies, twin,16 and triplet pregnancies.17
With a cervical length
3.5 cm, the risk of PTD was
low, not higher than 10%, even with a history of prior
PTD < 26 weeks. The risk increases significantly with
a cervical length of  2.5 cm to 32% risk for delivery
< 32 weeks gestation. When the cervical length is
< 1.5 cm the same risk is 48%. One has to realize
though, that even with a cervical length < 1.5 to 2.5 cm
the majority of women will not deliver preterm,
demonstrating that cervical length alone is not adequate
to predict cervical competence.
A better predictive power is achieved when
cervical length measurement is combined with fetal
fibronectin (fFN) testing (Table 2). A cervical length
< 2.5 cm at 22 to 24 weeks gestation and a positive
fFN test translates into a 64% risk of PTD, 59% risk
of delivery < 35 weeks, and 35% risk of delivery
< 32 weeks.18,19 A cervical length  1.5 cm with positive
fFN predicts delivery < 32 weeks in 90% of cases, and
< 28 weeks in 60% of cases.20 Women with cervical
length < 2.5 cm, positive fFN, and a prior PTD have a
64% risk of delivery < 35 weeks, and 50% risk of delivery
< 32 weeks.18,19 In the same cohort dataset, only 2.4% of
patients with a prior PTD but with negative fFN and
normal cervical length delivered < 30 weeks.21 Even in
a randomized trial, only 3.4% of women with a prior
PTD before 32 weeks gestation but with cervical
length > 2.5 cm between 18 and 20 weeks delivered
prematurely.22 According to the American College of
Obstetricians and Gynecologists (ACOG), the benefit
of such testing is more to identify those high-
risk patients by historic criteria who actually are at
minimal risk for PTD during the index pregnancy, for
reassurance and avoidance of unnecessary surveillance
or interventions.23
Owen et al24 have reported in women with a prior
PTD and cervical length < 2.5 cm at 16 to 24 weeks a
relative risk (RR) of PTD < 35 weeks of 4.5. In the study
by Iams et al15 on an unselected population, women with
RR of 6.1 for PTD < 35 weeks compared with women
 RECENT HISTORY OF PTD PREVENTION/VIDAEFF, RAMIN 7

Figure 4 The normal distribution of cervical length in the population at 20 to 28 weeks gestation; percentiles and their corresponding
measurement in centimeters (m, mean;
, standard deviation).

with cervical length > 4 cm. The same cervical measure-
ment at 28 weeks is associated with a RR of 9.5. If the
cervix is  1.3 cm at 22 to 24 weeks, the RR is 13.9; at
28 weeks, the RR is 24.9.15 An increasing RR with
gestational age speaks in support of the postulate of
cervical competence as a continuum. The augmentation
in RR is not an expression of increasing test accuracy
with advancing gestational age, but rather the effect of
nearing the event under study (PTD). There is no
improved diagnostic accuracy with sequential testing
over that associated with a single baseline measurement
properly timed.25 Nearing the event under study (PTD
< 35 weeks) also increases the specificity of the cervical
assessment, although the sensitivity decreases from 69%
at 16 to 18 weeks, to 40% at 22 to 24 weeks.26 Such a low
sensitivity still compares favorably with the approxi-
mately 20% sensitivity of digital examination.27 It
is interesting to note that in an unselected Finnish
population of 3694 women, the sensitivity of cervical
ultrasound evaluation at 18 to 22 weeks gestation
for PTD < 35 weeks was only 19%. The cervical length
nomogram in the Finnish women was also slightly

Table 2 Risk of Preterm Delivery (PTD) in Asymptomatic Women Based on Sonographic

Cervical Length, Fetal Fibronectin (fFN) Testing, History of PTD, or Combinations of Such
Predictors
Ultrasound Cervical
Length (cm) fFN History of PTD Risk PTD < 32 Weeks Risk PTD < 35 Weeks

> 2.5  þ 2.4% 3.4%

< 2.5   32%
< 2.5 þ  35% 59%
< 2.5 þ þ 50% 64%
< 1.5  48%
< 1.5 þ 90%
15,18 19,21
Based on data from Iams et al, Goldenberg et al, Heath et al,20 and Althuisius et al.22
8 AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 23, NUMBER 1
Figure 5 Transvaginal ultrasound image of a cervix displaying normal length (3.3 cm), but incipient funneling at the internal os.
different, with the 10th percentile represented by 3.2 cm
instead of 2.5 cm in the United States.28 Nonetheless, the
same observation was made of a continuum of risk
between shorter cervix and higher rate of PTD.
OTHER CERVICAL ULTRASOUND
PREDICTORS OF PTD
The interdependence, if any, between cervical length and
cervical funneling or dilatation continues to be debated.
Funneling (or beaking) implies an internal os dilatation
(maximal funneling width) of at least 5 mm (Fig. 5). The
predictive value of funneling for PTD was suggested in
singleton and twin pregnancies,29 and a systematic review
found funneling to be a risk factor for PTD independ-
ently of residual cervical length.30 However, this assump-
tion was not verified in large cross-sectional and cohort
studies, including the original study by Iams et al.15
It is still unclear if the so-called cervical stress test
(applying transfundal pressure and watching for funnel-
ing) at 16 to 24 weeks gestation as described by Guzman
et al31 can be accepted as a predictor for PTD.
The only other finding on second-trimester cer-
vical ultrasound, besides cervical length and funneling,
that has been independently associated with recurrent
PTD was a canal dilatation of 2 to 4 mm through the
entire length of the canal.32 This finding remains to be
verified by other investigators.
LIMITATIONS OF CERVICAL
SONOGRAPHY
Cervical sonography is limited by the timing of the
examination. Cervical measurements before 16 weeks
may not be helpful. Although, based on a cohort study
of 154 women, Naim et al33 concluded that a cervical
length < 3 cm before 16 weeks is strongly associated
2006
with preterm birth (risk of 23 versus 12% when the
cervical length was greater), the difference was not
statistically significant.
Likewise, the information obtained by cervical
sonography after 28 to 31 weeks is limited. Shortening
of the cervix begins approximately 10 weeks before
delivery in any gestational age group, beginning as early
as 16 weeks gestation.12 The same cervical shortening
during 5 to 10 weeks occurs in pregnancies resulting in
term deliveries.34 Thus, ultrasound assessment of the
cervix after 28 weeks is irrelevant. After 28 to 31 weeks,
the cervical length is also more variable between indi-
viduals and the predictive value of cervical measurement
is lost.35
Although the role of cervical sonography is over-
emphasized at present, digital examination may be more
relevant at times. Once the cervix is dilated 2 cm, the
sagittal imaging of the cervix on ultrasound is no longer
accurate in appreciating dilatation.7 At this point, digital
examination is more useful, and once the cervical dila-
tation can be detected by digital examination, there is
no further role for sonographic cervical assessment.
There are insufficient data to establish a cutoff
value for cervical length in multiple gestations. Accord-
ing to Guzman et al, with a cervical length > 2.0 cm
between 21 and 24 weeks gestation, 94% of twin preg-
nancies will deliver beyond 32 weeks.36 But in a large
French multicenter study of twin pregnancies, there was
no clear cutoff value for predicting PTD < 32 weeks
based on ultrasound at 22 weeks.37 At 27 weeks ges-
tation, a 2.5-cm cutoff value performed better, but
because neither tocolysis nor cerclage had been shown
effective in the prevention of PTD, the authors could not
recommend the routine use of cervical ultrasound
measurement in twin pregnancies.
The most significant (although theoretical) limi-
tation of cervical ultrasound assessment is the fact that it

remains unclear whether cervical changes are associated
with PTD regardless of etiological pathway.
CAN CERCLAGE PREVENT OR DELAY
PRETERM BIRTH?
In both singleton and multiple pregnancies, the place-
ment of a cerclage increases cervical length in patients
with cervical shortening preoperatively.38 Given that the
literature supports an inverse relationship between cer-
vical length and risk of PTD, it is appealing to believe
that cerclage may translate into pregnancy prolongation.
Nevertheless, the precise action of cerclage, its timing,
and applicability remain speculative. Is the mechanical
support of the cervix able to reverse or delay a condition
that has many underlying causes? Observational and
randomized clinical trials so far have provided conflicting
results.
Even in women with a clinical history typical for
cervical insufficiency, there are limited data to support
prophylactic cerclage. Only four randomized trials have
been conducted in women with clinical or historical risk
factors for cervical insufficiency. Twin pregnancy is a
clinical risk factor for PTD. In a randomized trial of 50
twin pregnancies, placement of an elective cerclage was
not associated with any benefit.39 In a French multi-
center trial, 506 women with moderate historical or
clinical risk for PTD were randomly assigned to pro-
phylactic cerclage or no cerclage.40 The rate of PTD was
higher among women assigned to cerclage (6.7 versus
5.5%), although the difference was not statistically sig-
nificant. In this study, high-risk or low-risk women were
excluded. Conversely, a British group decided to ran-
domly assign 194 women at high risk for PTD to
cerclage or no cerclage.41 Again, there was no difference
between the groups. In the largest randomized trial of
elective cerclage conducted in 12 countries, 1292 grav-
idas of different risk level were included.42 Among
gravidas with unclear historical or clinical factors of
risk, there was a lower rate of PTD < 37 weeks gestation
in women with cerclage versus those without cerclage,
although the difference was not statistically significant
(26 versus 31%; p ¼ 0.07). Only when the analysis was
limited to deliveries < 33 weeks did the difference reach
statistical significance (13 versus 17%; p ¼ 0.03); how-
ever, there was no reduction in perinatal mortality with
cerclage. In a stratified secondary analysis, only a history
of at least three spontaneous preterm deliveries or
second-trimester losses was associated with a significant
and beneficial reduction in PTD rate (15 versus 32%;
p ¼ 0.015). The authors estimated that 25 women would
receive a cerclage to prevent one birth < 33 weeks, and
advised that the procedure of elective cerclage be limited
to those women with at least three second-trimester
pregnancy losses or preterm deliveries.42 They also
estimated that as many as 96% of cerclages are unneces-
RECENT HISTORY OF PTD PREVENTION/VIDAEFF, RAMIN 9
sary, indicating the overuse consequent to the inaccuracy
of historical risk assessment. It has been reported that
70% of patients with even recurrent mid-trimester loss
will be delivered at term in their subsequent pregnancy
with no treatment.43 The historical diagnosis of cervical
insufficiency remains elusive, allowing for speculation
regarding indication or success rates. The inclusion of
women without a significant component of cervical
insufficiency in studies underestimates any demonstrable
benefit.
In all the above-described studies, the assessment
of risk for singleton pregnancies was subjective, based on
individual physician judgment, and therefore nonrepro-
ducible. Because of the conflicting results of these
studies, and the evidence that the majority of women
selected on historical risk factors actually do not need
cervical cerclage, interest has emerged for further select-
ing the candidates based on transvaginal cervicography
findings. In the 1990s cervical ultrasound was adopted as
an objective and reproducible anatomical assessment
tool.44 Guzman et al45 have reported that cervical sono-
graphic surveillance with cerclage only for cervical length
< 2 cm, between 15 and 24 weeks, would avoid cerclage
in 65% of cases otherwise eligible for elective cerclage by
traditional criteria. A decision based on ultrasound
demonstrable changes can be viewed as a departure
from the elective cerclage in favor of the therapeutic
cerclage, without any loss in efficacy.45,46 The effect of
such therapeutic cerclages has been studied in non-
randomized and randomized trials.
In one retrospective study from the United King-
dom, Shirodkar cerclage placement in 22 of 43 asymp-
tomatic women with cervical length  1.5 cm at 23 weeks
was associated with a 10-fold reduction in the rate of
PTD < 32 weeks (5 versus 50%; p ¼ 0.001).47 But in the
United States, a retrospective analysis of McDonald
cerclage placement in 39 of 63 women with funneling
or cervical length < 2.5 cm did not demonstrate any
improvement in the rate of PTD < 35 weeks (OR, 1.1;
95% CI, 0.3 to 4.6).48 In still another retrospective
report, placement of a cerclage in 43 of 85 unselected
women with cervical length  3.0 cm appeared to pro-
long the pregnancy by a mean of 2 weeks.49
The results extracted from historical cohorts are
also conflicting. Novy et al50 reported on 20 of 35
women with early cervical changes (cervical length
< 3.0 cm and funneling), not in labor, who received a
cerclage (mainly of Shirodkar type). Cerclage prolonged
the pregnancy by a mean of 6 weeks (p ¼ 0.003). In the
same report, among an additional 35 women with
advanced cervical dilatation (2 to 5 cm) and visible
membranes, those who received cerclage (n ¼ 19) deliv-
ered on average 4 weeks later (mean gestational age, 30.2
versus 25.6 weeks in those without cerclage; p ¼ 0.008).
Conversely, Hassan et al51 reported no benefit in 25 of
80 unselected women with cervical length  1.5 cm
10 AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 23, NUMBER 1
between 14 and 24 weeks gestation, who received a
McDonald cerclage. The procedure did not reduce the
rate of PTD < 34 weeks, but was associated with a
significantly increased rate of preterm premature rupture
of membranes occurring within a mean of 22 hours from
the intervention (65 versus 36%; p < 0.05). The mean
gestational age at delivery in the cerclage group was
29 weeks, compared with 33 weeks for women without
cerclage.
Placement of a cerclage in women with twin pre-
gnancies and a mid-trimester cervical length  2.5 cm
failed to improve outcomes in a nonrandomized pro-
spective study.52 In a case-control study of triplet preg-
nancies with cervical shortening on ultrasound, Poggi
et al38 found a higher rate of PTD < 32 weeks in the
cerclage group, although the difference was not statisti-
cally significant.
Owing to the heterogeneity in study population
and study design, no composite interpretation of these
observational data is possible. Unfortunately, more
recent randomized clinical trials also provided conflict-
ing results. The first randomized trial published by Rust
et al53 in 2000, in which 61 patients with sonographic
evidence of internal os dilatation between 16 and
24 weeks gestation were randomly assigned to cerclage
or no cerclage, no improvement in perinatal outcomes
was detected in association with cerclage placement. The
same investigators in the United States later reported on
138 women with funneling or cervical length < 2.5 cm at
16 to 24 weeks gestation.3 The study included 12%
multiple pregnancies, but historical risk factors did not
play a role in the selection of study subjects. The women
included in the study were randomly assigned to cerclage
or no cerclage, and all subjects were placed on modified
bed rest at home. Before random assignment, all subjects
underwent amniocentesis to rule out intra-amniotic
infection. One to 3 days before random assignment,
and one day after randomization, all subjects received
antibiotics and indomethacin. The study results indi-
cated no difference between the groups in the rate of
PTD, neonatal morbidity, or neonatal mortality.
Investigators from the Netherlands conducted a
randomized study of cervical cerclage (Cervical Incom-
petence Prevention Randomized Cerclage Trial).22
Women with historical risk factors were first randomly
assigned to prophylactic cerclage versus expectant man-
agement. Prophylactic cerclage did not improve the rate
of PTD < 34 weeks. The women in the expectant
management group continued to be followed with serial
ultrasounds, becoming eligible for a second random-
ization when the cervical length was < 2.5 cm. Con-
sequently, 35 women were randomly assigned to cerclage
or no cerclage. The 19 women in the cerclage group also
received 6 days of broad-spectrum antibiotics, perioper-
ative indomethacin, and modified bed rest at home. The
remaining women in the no cerclage group also received
2006
antibiotics and bed rest at home, but no indomethacin.
The rate of PTD < 34 weeks was 0% in the cerclage
group versus 44% in the no cerclage group.54 These
findings are in sharp contrast with those elicited in the
U.S. randomized trial,3 but unfortunately, comparisons
between these two studies are difficult because of re-
markable differences among populations and study
methods. As an example, the women in the Dutch study
appear to have been at a higher risk of PTD than the
women in the North American study, and they remained
eligible for cerclage up to 27 weeks, not 24 weeks as in
the American study. Although the outcome results in the
Dutch study are significant, given the small sample size,
random error cannot be ruled out categorically.
The largest randomized trial of cervical cerclage
in women with short cervix was published recently by the
Fetal Medicine Foundation in the United Kingdom.55
The study was conducted at 12 hospitals in Europe,
Africa, and South America, selecting the patients from a
cohort of 47,123 women undergoing routine transvagi-
nal ultrasound at 22 to 24 weeks gestation. The 253
women who were found to have a cervical length of
< 1.5 cm and agreed to participate in the study were
randomly assigned to either expectant management or
Shirodkar cerclage. The women assigned to cerclage
received one dose of intravenous erythromycin intra-
operatively, but no other interventions were employed,
except for corticosteroids for fetal maturation. Of the
women in the expectant management group, 26% deliv-
ered < 33 weeks versus 22% in the cerclage group. The
difference was statistically insignificant (RR, 0.84; 95%
CI, 0.54 to 1.31). The results speak against routine
cervical examinations by ultrasound, but even in a sub-
group analysis of women with a history of PTD at 16 to
32 weeks, no outcome improvement could be demon-
strated with cerclage.
The cutoff of 1.5 cm was selected as entry crite-
rion because it more accurately predicts a negative out-
come than does a higher cutoff value. Still, Groom
et al,56 based on a prospective study that correlated
outcome with the preoperative cervical length and pres-
ence of visible membranes at the time of cerclage,
recommended against waiting until the cervical length
is  1.5 cm in high-risk women. Once the cervix is
< 1.5 cm, almost one third of women will have visible
fetal membranes, whereas fetal membranes were not
visible when the cervix was > 1.5 cm. The presence of
visible fetal membranes significantly and independently
worsens the prognosis. The gestational age at delivery
was 23 weeks in the group with dilatation of internal
cervical os and visible membranes, compared with
37 weeks in the group with no visible membranes at
the time of cerclage. It has been speculated that the
exposure of fetal membranes directly to the vaginal
ecosystem enhances the inflammatory process leading
to preterm labor. Guzman et al57 reported that women

with cervix < 2.0 cm have a significantly higher fre-
quency of acute inflammatory lesions in the placenta.
Equally conflicting conclusions were advanced by
two recent meta-analyses of randomized trials, although
both meta-analyses were based on almost the same set of
data. Bachman et al58 consider that cerclage is beneficial
in reducing the risk of PTD < 34 weeks, whereas
Drakeley et al,59 on a more cautious note, warn that
the benefits of cerclage have not been proven yet. A
common deficiency of these two meta-analyses is that
the data were pooled from studies with different ascer-
tainment criteria; in some studies, the patients were
included based on ultrasound findings, whereas in others
the eligibility criteria were purely historical. Still another
meta-analysis published by Belej-Rak et al60 included
only studies that used cervical changes on ultrasound as
entry criteria. Only six such studies were available for
systematic review: two randomized, and four observa-
tional. The ascertainment bias may have been decreased
in this meta-analysis, but the inclusion of observational
studies increased the susceptibility to selection bias. The
authors concluded that the available evidence does not
support the placement of cerclage for a sonographically
detected short cervix. Although an increase in birth
weight was noted with cerclage, no benefit was demon-
strable with regard to interval to delivery, gestational age
at delivery, rate of PTD, or neonatal mortality and
morbidity.
The most challenging presentation of cervical
insufficiency, membranes visible at the os or prolapsed
beyond a dilated external os, was the subject of another
randomized Dutch trial. The efficacy of emergency
cerclage plus antibiotics, indomethacin, and bed rest
was assessed versus bed rest alone plus antibiotics for
1 week in 23 women with gestational ages up to
27 weeks.61 A 4-week difference in mean gestational age
at delivery was found in favor of the cerclage group
(29.9 versus 25.9 weeks), although the difference did not
reach statistical significance, probably because of the
limited sample size. PTD < 34 weeks was significantly
lower in the cerclage group (seven of 13) compared with all
10 deliveries in the no cerclage group (p ¼ 0.02). Neonatal
morbidity was significantly lower in the cerclage group,
but no difference was seen in neonatal survival.
Given the conflicting results, and the paucity of
data from randomized studies on therapeutic cerclage,
the management remains speculative, and the decision to
proceed with cerclage should be made with caution; the
patient should be involved in the decision making.
ACOG considers that the value of abnormal cervical
findings on ultrasound is questionable, especially in the
absence of proven treatments affecting outcome.23 The
Society of Obstetricians and Gynecologists of Canada
does not support outline screening assessment of mid-
trimester cervical length.62 Based on available evidence,
it is unlikely that a woman with short cervix found
RECENT HISTORY OF PTD PREVENTION/VIDAEFF, RAMIN 11
incidentally on ultrasound and with no history of adverse
pregnancy outcome would benefit from cerclage. Under
the circumstances, the recommendation of the American
College of Radiology that the cervix and lower uterine
segment be imaged as part of every obstetric ultrasound
examination in the second trimester is unwarranted, and
possibly harmful if leading to unnecessary treatment.63
On the other hand, women with history of PTD
and short cervix on ultrasound may benefit from cerc-
lage. According to ACOG, elective cerclage can be
considered in a woman with history of at least three
unexplained mid-trimester pregnancy losses or preterm
deliveries.64 Women with less categorical historical risk
factors for cervical insufficiency may be assessed by
ultrasound between 16 and 20 weeks, and a cerclage
may be placed before 24 weeks if the cervical length
< 2.5 cm. Once initiated, the cervical changes can
progress quickly; when indicated, monitoring should
be performed at 1- to 2-week intervals. There is no
evidence that placement of a cerclage in multiple preg-
nancies is beneficial, and the available evidence does not
suggest that women exposed to diethylstilbestrol in utero
should be managed any differently from nonexposed
women. ACOG does not recommend cervical assess-
ment in women with no risk factors for cervical insuffi-
ciency.64
The value of postcerclage cervical assessment is
unclear. Postcerclage cervical length was found to have
predictive value for PTD < 36 weeks in a study of 25
singleton and four twin gestations,65 but these results
could not be replicated by others.66
The benefits, risks, and cost effectiveness of
screening women with vaginal ultrasound and perform-
ing a cerclage continue to be studied, and it is hoped that
more valuable data will become available. A National
Institutes of Health multicenter randomized clinical trial
of McDonald cerclage is underway in the United States,
targeting women with a previous PTD < 32 weeks and
cervical length < 2.5 cm between 16 and 22 weeks.
REFERENCES
1. Martin JA, Hamilton BE, Sutton PD, et al. Births: final data
for 2002. Natl Vital Stat Rep 2003;52:1–113
2. Romero R, Mazor M, Munoz H, et al. The preterm labor
syndrome. Ann NY Acad Sci 1994;734:414–429
3. Rust OA, Atlas RO, Reed J, et al. Revisiting the short cervix
detected by transvaginal ultrasound in the second trimester:
why cerclage therapy may not help. Am J Obstet Gynecol
2001;185:1098–1105
4. Novy MJ. Cervical cerclage [letter]. Am J Obstet Gynecol
2002;187:1426–1427
5. Mann EC, McLaren WD, Hoyt OB. The physiology and
clinical significance of the uterine isthmus. Am J Obstet
Gynecol 1961;81:209–222
6. Herman GE. Note on Emmet’s operation as a preventive of
abortion. Br J Obstet Gynaecol 1902;2:256–257
12 AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 23, NUMBER 1
7. Yost NP, Bloom SL, Twickler DM. Pitfalls in the ultrasonic
cervical length measurement for predicting preterm birth.
Obstet Gynecol 1999;93:510–516
8. Jackson GM, Ludmir J, Bader TJ. The accuracy of digital
examination and ultrasound in the evaluation of cervical
length. Obstet Gynecol 1992;79:214–218
9. Eppel W, Schurz B, Frigo P, et al. Vaginal ultrasound in the
puerperium: cervix and lower uterine segment on the first post
partum day and uterus six weeks post partum. Z Geburtshilfe
Perinatol 1992;196:217–220
10. Zalar RW Jr. Early cervical length, preterm prelabor and
gestational age at delivery. Is there a relationship? J Reprod
Med 1998;43:1027–1033
11. Guzman ER, Mellon C, Vintzileos AM, et al. Longitudinal
assessment of endocervical canal length between 15 and
24 weeks’ gestation in women at risk for pregnancy loss, or
preterm birth. Obstet Gynecol 1998;92:31–37
12. Okitsu O, Mimura T, Nakayama T, et al. Early prediction of
preterm delivery by transvaginal ultrasonography. Ultrasound
Obstet Gynecol 1992;2:402–409
13. Ayers J, DeGrood P, Compton A, et al. Sonographic
evaluation of cervical length in pregnancy: diagnosis and
management of preterm cervical effacement in patients at risk
for premature delivery. Obstet Gynecol 1988;71:939–944
14. Kushnir O, Vigil DA, Izquierdo L, et al. Vaginal ultrasono-
graphic assessment of cervical length changes during normal
pregnancy. Am J Obstet Gynecol 1990;162:991–993
15. Iams JD, Goldenberg RL, Meis PJ, et al. The length of the
cervix and the risk of spontaneous premature delivery. N Engl
J Med 1996;334:567–572
16. Imseis HM, Albert TA, Iams JD. Identifying twin gestation
at low risk for preterm birth with a transvaginal ultrasono-
graphic cervical measurement at 24 to 26 weeks’ gestation.
Am J Obstet Gynecol 1997;177:1149–1155
17. To MS, Skentou C, Cicero S, et al. Cervical length in triplets:
prediction of spontaneous preterm delivery. Ultrasound
Obstet Gynecol 2000;16:515–518
18. Iams JD, Goldenberg RL, Mercer BM, et al. The preterm
prediction study: recurrence risk of spontaneous preterm
birth. Am J Obstet Gynecol 1998;178:1035–1040
19. Goldenberg RL, Iams JD, Mercer BM, et al. The preterm
prediction study: the value of new vs. standard risk factors in
predicting early and all spontaneous preterm births. Am J
Public Health 1998;88:233–238
20. Heath VCF, Southall TR, Souka AP, et al. Cervical length at
23 weeks of gestation: prediction of spontaneous preterm
delivery. Ultrasound Obstet Gynecol 1998;12:312–317
21. Goldenberg RL, Iams JD, Das A, et al. The preterm
prediction study: sequential cervical length and fetal fibro-
nectin testing for the prediction of spontaneous preterm birth.
Am J Obstet Gynecol 2000;182:636–643
22. Althuisius SM, Dekker GA, van Geijn HP, et al. Cervical
incompetence prevention randomized cerclage trial
(CIPRACT): study design and preliminary results. Am J
Obstet Gynecol 2000;183:823–829
23. American College of Obstetricians and Gynecologists.
Assessment of Risk Factors for Preterm Birth. Practice
bulletin 31. Washington, DC: American College of Obste-
tricians and Gynecologists; 2001
24. Owen J, Yost N, Berghella V, et al. Mid-trimester
endovaginal sonography in women at high risk for sponta-
neous birth. JAMA 2001;286:1340–1348
2006
25. Watson WJ, Stevens D, Welter S, et al. Observations on the
sonographic measurement of cervical length and the risk of
preterm birth. J Matern Fetal Med 1999;8:17–19
26. Iams JD, Newman RB, Thom EA, et al. Frequency of uterine
contractions and the risk of spontaneous preterm birth. N
Engl J Med 2002;346:250–255
27. Berghella V, Tolosa IE, Kuhlman K, et al. Cervical
ultrasonography compared with manual examination as a
predictor of preterm delivery. Am J Obstet Gynecol 1997;177:
723–730
28. Taipale P, Hiilesmaa V. Sonographic measurement of uterine
cervix at 18–22 weeks gestation and the risk of preterm
delivery. Obstet Gynecol 1998;92:902–907
29. Yang JH, Kuhlman K, Daly S, et al. Prediction of preterm
birth by second trimester cervical sonography in twin
pregnancies. Ultrasound Obstet Gynecol 2000;15:288–291
30. Leitich H, Brunbauer M, Kaider A, et al. Cervical length and
dilatation of the internal cervical os detected by vaginal
ultrasonography as markers for preterm delivery: a systematic
review. Am J Obstet Gynecol 1999;181:1465–1472
31. Guzman ER, Vintzileos AM, McLean DA, et al. The natural
history of a positive response to transfundal pressure in
women at risk for cervical incompetence. Am J Obstet
Gynecol 1997;176:634–638
32. Yost NP, Owen J, Berghella V, et al. Second-trimester
cervical sonography: features other than cervical length to
predict spontaneous preterm birth. Obstet Gynecol 2004;103:
457–462
33. Naim A, Haberman S, Burgess T, et al. Changes in cervical
length and the risk of preterm labor. Am J Obstet Gynecol
2002;186:887–889
34. Zilianti M, Azuaga A, Calderon F, et al. Monitoring the
effacement of the uterine cervix by transperineal sonography:
a new perspective. J Ultrasound Med 1995;14:719–724
35. Goldenberg RL, Iams JD, Miodovnik M, et al. The preterm
prediction study: risk factors in twin gestations. Am J Obstet
Gynecol 1996;175:1047–1053
36. Guzman ER, Walters C, O’Reilly-Groon C, et al. Use of
cervical ultrasonography in prediction of spontaneous preterm
birth in twin gestations. Am J Obstet Gynecol 2000;183:
1108–1113
37. Vayssière C, Favre R, Audibert F, et al. Cervical length and
funneling at 22 and 27 weeks to predict spontaneous birth
before 32 weeks in twin pregnancies: a French prospective
multicenter study. Am J Obstet Gynecol 2002;187:1596–
1604
38. Poggi SH, Ghidini A, Landy HJ, et al. Role of transvaginal
sonographic cervical length in predicting spontaneous preterm
delivery in triplet pregnancies with therapeutic cerclage.
J Reprod Med 2003;48:785–788
39. Dor J, Shalev J, Mashiach S, Blankstein J, Serr DM. Elective
cervical suture of twin pregnancies diagnosed ultrasonically in
the first trimester following induced ovulation. Gynecol
Obstet Invest 1982;13:55–60
40. Lazar P, Gueguen S, Dreyfus J, et al. Multicenter and
controlled trial of cervical cerclage in women at moderate risk
of preterm delivery. Br J Obstet Gynaecol 1984;91:731–735
41. Rush RW, Isaacs S, McPherson K, et al. A randomized
controlled trial of cervical cerclage in women at high risk for
preterm delivery. Br J Obstet Gynaecol 1984;91:724–730
42. MacNaughton MC, Chalmers IG, Dubowitz V, et al. Final
report of the Medical Research Council/Royal College of

Obstetrics and Gynaecology multicentre randomised trial of
cervical cerclage. Br J Obstet Gynaecol 1993;100:516–523
43. Harger JH, Archer DF, Marchese SG, et al. Etiology of
recurrent pregnancy losses and outcome of subsequent
pregnancies. Obstet Gynecol 1983;62:574–581
44. Burger M, Weber-Rossler T, Willmann M. Measurement of
the pregnant cervix by transvaginal sonography: an interob-
server study and new standards to improve the interobserver
variability. Ultrasound Obstet Gynecol 1997;9:188–193
45. Guzman ER, Forster JK, Vintzileos AM, et al. Pregnancy
outcomes in women treated with elective versus ultrasound-
indicated cervical cerclage. Ultrasound Obstet Gynecol 1998;
12:323–327
46. Kurup M, Goldkrand J. Cervical incompetence: elective,
emergent, or urgent cerclage. Am J Obstet Gynecol 1999;181:
240–246
47. Heath VCF, Souka AP, Erasmus I, et al. Cervical length at
23 weeks of gestation: the value of Shirodkar suture for the
short cervix. Ultrasound Obstet Gynecol 1998;12:318–322
48. Berghella V, Daly SF, Tolosa JE, et al. Prediction of preterm
delivery with transvaginal ultrasonography of the cervix in
patients with high-risk pregnancies: does cerclage prevent
prematurity? Am J Obstet Gynecol 1999;181:809–815
49. Hibbard JU, Snow J, Moawad AH. Short cervical length by
ultrasound and cerclage. J Perinatol 2000;20:161–165
50. Novy MJ, Gupta A, Wothe DD, et al. Cervical cerclage in the
second trimester of pregnancy: a historical cohort study. Am J
Obstet Gynecol 2001;184:1447–1456
51. Hassan SS, Romero R, Maymon E, et al. Does cerclage
prevent preterm delivery in patients with a short cervix?. Am J
Obstet Gynecol 2001;184:1325–1331
52. Newman RB, Krombach RS, Myers MC, et al. Effect of
cerclage on obstetrical outcome in twin gestations with a
shortened cervical length. Am J Obstet Gynecol 2002;186:
634–640
53. Rust OA, Atlas RO, Jones KJ, et al. A randomized trial of
cerclage versus no cerclage among patients with ultrasono-
graphically detected second-trimester preterm dilatation of
the internal os. Am J Obstet Gynecol 2000;183:830–835
54. Althuisius SM, Dekker GA, Hummel P, et al. Final results of
the cervical incompetence prevention randomized cerclage
trial (CIPRACT): therapeutic cerclage with bed rest versus
bed rest alone. Am J Obstet Gynecol 2001;185:1106–1112
RECENT HISTORY OF PTD PREVENTION/VIDAEFF, RAMIN 13
55. To MS, Alfirevic Z, Heath VCF, et al. Cervical cerclage
for prevention of preterm delivery in women with short
cervix: randomised controlled trial. Lancet 2004;363:1849–
1853
56. Groom KM, Shennan AH, Bennett PR. Ultrasound-
indicated cervical cerclage: outcome depends on preoperative
cervical length and presence of visible membranes at time of
cerclage. Am J Obstet Gynecol 2002;187:445–449
57. Guzman ER, Shen-Schwartz S, Benito C, et al. The
relationship between placental histology and cervical ultra-
sonography in women at risk for pregnancy loss and
spontaneous preterm birth. Am J Obstet Gynecol 1999;181:
793–797
58. Bachmann LM, Coomarasamy H, Honest H, et al. Elective
cervical cerclage for prevention of preterm birth: a systematic
review. Acta Obstet Gynecol Scand 2003;82:398–404
59. Drakeley AJ, Roberts D, Alfirevic Z. Cervical cerclage for
prevention of preterm delivery: meta-analysis of randomized
trials. Obstet Gynecol 2003;102:621–627
60. Belej-Rak T, Okun N, Windrim R, et al. Effectiveness of
cervical cerclage for a sonographically shortened cervix: a
systematic review and meta-analysis. Am J Obstet Gynecol
2003;189:1679–1687
61. Althuisius SM, Dekker GA, Hummel P, et al. Cervical
incompetence prevention randomized cerclage trial: emer-
gency cerclage with bed rest versus bed rest alone. Am J
Obstet Gynecol 2003;189:907–910
62. Van den Hof M, Crane J. Ultrasound cervical assessment in
predicting preterm birth. J Soc Obstet Gynaecol Can 2001;23:
418–421
63. Iams JD. Prediction and early detection of preterm birth.
Obstet Gynecol 2003;101:402–412
64. American College of Obstetricians and Gynecologists.
Cervical Insufficiency. Practice bulletin 48. Washington,
DC: American College of Obstetricians and Gynecologists;
2003
65. Guzman E, Houlihan C, Vintzileos A, et al. The significance
of transvaginal ultrasonographic evaluation of the cervix in
women treated with emergency cerclage. Am J Obstet
Gynecol 1996;175:471–476
66. Dijkstra K, Funai E, O’Neill L, et al. Change in cervical
length after cerclage as a predictor of preterm delivery. Obstet
Gynecol 2000;96:346–350