Ultrasound Obstet Gynecol 2015; 45: 358–362

Published online 29 January 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/uog.14742

How to . . . .
Practical advice on imaging-based techniques and investigations with
accompanying slides and videoclips online

How to measure cervical length

K. O. KAGAN* and J. SONEK†‡
*Department of Obstetrics and Gynaecology, University of Tübingen, Tübingen, Germany; †Fetal Medicine Foundation USA, Dayton, OH,
USA; ‡Division of Maternal Fetal Medicine, Wright State University, Dayton, OH, USA

BACKGROUND It has been shown that only 15% of symptomatic

patients presenting with preterm contractions will actually
Cervical-length measurement using transvaginal sono- deliver within the next 7 days8 . Cervical length can be used
graphy (TVS) is an essential part of assessing the risk to distinguish between ‘true’ and ‘false’ labor. Sotiriadis
of preterm delivery. At mid-gestation, it provides a et al.9 summarized six studies, including about 1800
useful method with which to predict the likelihood women, that assessed the risk for preterm delivery within
of subsequent preterm birth in asymptomatic women. 1 week, based on cervical length at the time of admission.
In women who present with threatened spontaneous With a cut-off of 15 mm, the sensitivity was about 60%
preterm labor, TVS measurement of cervical length for a false-positive rate of about 10%. The respective
can help to distinguish between ‘true’ and ‘false’ positive and negative likelihood ratios were 5.7 and 0.5.
spontaneous preterm labor. Additionally, there is some With cut-offs of 20 and 25 mm, the sensitivities increased
evidence that measurement of the cervix at the 11+0 to 75% and 80%, but at the expense of false-positive
to 13+6-week scan can help to establish the risk of rates of 20% and 30%, respectively.
preterm birth1,2 . Recent work10 has shown that cervical length at 11–13
To et al.3 reported on cervical-length measurement weeks’ gestation, in combination with maternal history,
between 22 and 24 weeks’ gestation in 39 000 women can detect about half of the pregnancies that result in
with a singleton pregnancy. The cervical length was preterm delivery before 34 weeks, for a false-positive
found to be distributed normally, with a mean length rate of 10%. However, the difference between the
of 36 mm. In about 1% of the women, the length was median cervical length of the normal group and that
15 mm or less. This cut-off is generally used to define the of the preterm-birth group was only 5 mm (32.5 mm
high-risk group in interventional studies4 . In most studies vs 27.5 mm)10 , which raises doubts as to whether this
focusing on asymptomatic twin pregnancies at 20 to 24 technique can be used for general screening.
weeks, a cut-off of 25 mm is applied5 . Celik et al.6 used
cervical-length measurements obtained between 20 and
24 weeks’ gestation, along with maternal history, in more PRACTICAL POINTS
than 58 000 women to create computed risk models for
preterm delivery. They compared patients who delivered There are essentially four methods that can be used
before 28 weeks, between 28 and 30 weeks, between 31 to evaluate the uterine cervix: digital examination,
and 33 weeks, and between 34 and 36 weeks’ gestation. transabdominal ultrasound, transperineal ultrasound
For a 10% false-positive rate, the sensitivities were 81%, (TPS) and TVS. It is the digital examination that provides
59%, 53% and 29%, respectively. In a Health Technology the most comprehensive evaluation of the cervix, assessing
Assessment report, Honest et al.7 summarized the results dilatation, position, consistency and length. However, this
of five studies that used cervical-length measurements examination suffers from being subjective. It is limited
between 20 and 24 weeks, with cut-offs of 20–30 mm, especially in its ability to establish accurately the cervical
to predict preterm birth before 34 weeks’ gestation. The length. It also cannot detect reproducibly any changes
resultant positive likelihood ratios ranged from 2.3 for at the internal cervical os and the upper portion of the
30 mm to 7.6 for 20 mm. cervical canal. Ultrasound, with its ability to penetrate the

Correspondence to: Prof. K. O. Kagan, University of Tuebingen, Calwerstrasse 7, 72076 Tübingen, Germany (e-mail: [email protected])

Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd. HOW TO...
How To . . . 359

identify the internal, as well as the external, cervical

ora. The external os is identified as the point at which the
anterior and posterior lips of the cervix come together.
Identification of the location of the internal os can be more
challenging. In order to do so, the cervical mucosa must
be identified. This is usually hypoechoic, with respect to
the surrounding stroma, though occasionally it can be
slightly hyperechoic. A thin line of demarcation between
the stroma and the cervical mucosa can generally be
identified on ultrasound. The point at which the cervical
mucosa ends is considered to be the internal cervical os.
Of note, the mucosa of the lower uterine segment is much
thinner than is the cervical mucosa and is usually difficult
to identify on ultrasound (Figure 4).

Figure 1 Transvaginal ultrasound image of a normal cervix,

showing the most commonly used method of cervical length Magnification of the ultrasound image needs to be
measurement (double-headed arrow). In a curved cervix, the length appropriate
is underestimated with a straight line. If the cervix is curved and the
straight-line cervical length measurement is short, measurement The image should be magnified sufficiently so that the
obtained in two or more segments provides a more accurate
estimation of length14 . However, in patients with a pathologically
morphology of the cervix is easily identifiable. The cervix
short cervix, the cervical canal will be straight and this adjustment should occupy approximately 50–75% of the image.
in measurement technique is unnecessary.

Pressure from the probe on the cervix should be as little

cervical tissue and display its anatomy, makes an ideal
as possible
modality with which to address both of these issues. Both
TVS and TPS are performed with the patient in a dorsal If excessive pressure is applied on the cervix by the probe,
lithotomy position. In order for the measurement to be the cervix appears artificially to be longer and the presence
accurate and reproducible, the following factors need to of a funnel will be obscured. Generally, in order to identify
be taken into account (Figure 1 and Videoclip S1): the relevant structures, it is helpful to put some pressure
on the cervix initially. However, subsequently this should
Maternal bladder should be essentially empty be reduced to the minimum that is required in order to
view the cervix adequately (Videoclip S2).
The patient having a full bladder can increase artificially
the cervical length. To et al.11 showed that the mean
difference between cervical length measured when the Duration of the examination should be 3–5 min
patient’s bladder is empty and that measured when the
The cervix is not a static structure and the length can
bladder is full is about 4 mm. In addition, a full bladder
vary, for example due to uterine contractions or to
can obscure the presence of cervical funneling by com-
different positions of the patient (Videoclip S3). Therefore,
pressing the two halves of the funnel together (Figure 2).
sufficient time should be allowed for the examination to
detect these changes. It may be helpful to press manually
A longitudinal view of the cervix should be obtained on the uterus or to ask the patient to push downwards
to assess the cervical stability. Some even suggest that
The cervix should be measured along its longitudinal axis,
during the examination the patient stands, with the
which may be different from the patient’s longitudinal
ultrasound probe inside the vagina. Several (at least three)
axis. The cervical canal, in most cases, is a fairly thin
measurements should be obtained during the course of
line. It may have a thin layer of hypoechoic contents. This
the examination and the shortest measurement should be
is true especially in the third trimester. In all likelihood,
it represents accumulation of mucus, and needs to be used for counseling.
differentiated from a thin cervical funnel. This is best
done by delineating the course of the fetal membranes: Calipers should be placed correctly
if they are not prolapsing into the cervical canal and are
located at the level of the internal cervical os, the presence Cervical measurement is obtained by placing calipers at
of a true funnel is unlikely (Figure 3). the external and the internal ora (Figure 1). In 95% of
cases, the difference between two measurements obtained
The cervical canal and surrounding cervical mucosa need by the same sonographer or by two different sonographers
to be identified is about 4 mm12 . If there is funneling, the caliper should
be placed at the apex of the funnel. If the cervix exceeds
In order to avoid inclusion of the isthmus into the 25 mm in length, it will be curved in more than 50% of
cervical-length measurement, care must be taken to cases13 . The standard method of measurement, using a

Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd. Ultrasound Obstet Gynecol 2015; 45: 358–362.
360 How To . . .

27mm 14mm

Figure 2 Transvaginal ultrasound measurement of cervical length in the same patient, with a full bladder (a) and with an empty bladder (b).

Figure 3 Transvaginal ultrasound image of a cervix with Figure 5 Transvaginal ultrasound image of a short cervix with
accumulation of mucus. Amniotic membranes are indicated funneling.
(arrow), showing that funneling is not present.

Figure 4 Transvaginal ultrasound image of a cervix, showing

cervical mucosa as a homogeneous and hypoechoic structure Figure 6 Transvaginal ultrasound image of amniotic fluid sludge
compared to the surrounding stroma. (arrow).

Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd. Ultrasound Obstet Gynecol 2015; 45: 358–362.
How To . . . 361

Figure 7 Transvaginal ultrasound Doppler image of vasa previa (a) and gray-scale image of placenta previa (b).

Figure 8 Transperineal (a) and transabdominal (b) ultrasound images of the cervix in the same patient. The cervical canal is indicated
(arrows).

straight line between the internal and the external ora, will logistic regression analysis that includes both funneling
underestimate the cervical length in these cases. However, and cervical length, funneling has been shown not to
this is of little clinical significance as these patients are be an independent risk factor for spontaneous preterm
at low risk regardless of the exact measurement. In the delivery17 .
high-risk group of patients with a cervical length < 16 mm, • Amniotic fluid sludge can be found as echogenic aggre-
the cervix will always be a straight line. gates close to the internal os or within a funnel. This
In the first trimester, the difference in cervical length appears to be associated with microbial invasion of
between patients who are at risk for preterm birth and the amniotic cavity18,19 . Sludge is an independent risk
those who are not is extremely small; therefore, it would factor for spontaneous preterm delivery, preterm rup-
seem appropriate to correct for cervical curvature by ture of membranes and histological chorioamnionitis
obtaining the measurement either in segments or by in asymptomatic patients at high risk for spontaneous
tracing the canal14 . preterm delivery (Figure 6, Videoclip S4).
• Vasa previa, placenta previa or low-lying placenta can
be diagnosed20 (Figure 7).
Identification of additional significant findings at the
time of the cervical scan Other approaches
• Funneling, defined as protrusion of the amniotic In general, the cervix should be assessed by TVS. In
membranes into the cervical canal, is considered by cases in which this should be avoided, such as in those
some as an additional risk factor for preterm delivery with preterm prelabor rupture of membranes, the cervical
(Figure 5). Various criteria for the diagnosis of true length can be measured by TPS: the transducer is placed
funneling have been published15,16 . However, using a on the perineum and rotated until the complete cervical

Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd. Ultrasound Obstet Gynecol 2015; 45: 358–362.
362 How To . . .

canal and the internal and external ora can be identified. preterm birth in women with threatened preterm labor: a
The placement of the probe is further away from the meta-analysis. Ultrasound Obstet Gynecol 2010; 35: 54–64.
10. Greco E, Gupta R, Syngelaki A, Poon LCY, Nicolaides KH.
cervix than it is on TVS; therefore, there is a reduction in First-Trimester Screening for Spontaneous Preterm Delivery
the detail in which the cervix is seen. In 95% of cases, the with Maternal Characteristics and Cervical Length. Fetal Diagn
difference between TVS and TPS measurements is within Ther 2012; 31: 154–161.
± 5 mm21 . Assessment of cervical length by TAS can be 11. To MS, Skentou C, Cicero S, Nicolaides KH. Cervical
used as an initial evaluation but it should be borne in assessment at the routine 23-weeks’ scan: problems with
transabdominal sonography. Ultrasound Obstet Gynecol 2000;
mind that, especially in cases with a short cervix, this 15: 292–296.
modality tends to overestimate the true cervical length. 12. Heath VC, Southall TR, Souka AP, Novakov A, Nicolaides
Therefore, a proper risk assessment should be based on a KH. Cervical length at 23 weeks of gestation: relation to
TVS or TPS measurement (Figure 8). demographic characteristics and previous obstetric history.
Ultrasound Obstet Gynecol 1998; 12: 304–311.
13. To MS, Skentou C, Chan C, Zagaliki A, Nicolaides KH.
Cervical assessment at the routine 23-week scan: standardizing
REFERENCES techniques. Ultrasound Obstet Gynecol 2001; 17: 217–219.
14. Retzke JD, Sonek JD, Lehmann J, Yazdi B, Kagan KO.
1. Kagan KO, To M, Tsoi E, Nicolaides KH. Preterm birth: the Comparison of three methods of cervical measurement in the
value of sonographic measurement of cervical length. BJOG first trimester: single-line, two-line, and tracing. Prenat Diagn
2006; 113 (Suppl): 52–56. 2013.
2. Sonek J, Shellhaas C. Cervical sonography: a review. Ultrasound 15. Rust OA, Atlas RO, Kimmel S, Roberts WE, Hess LW. Does
Obstet Gynecol 1998; 11: 71–78. the presence of a funnel increase the risk of adverse perinatal
3. To MS, Skentou CA, Royston P, Yu CKH, Nicolaides KH. outcome in a patient with a short cervix? Am J Obstet Gynecol
Prediction of patient-specific risk of early preterm delivery using 2005; 192: 1060–1066.
maternal history and sonographic measurement of cervical 16. Grimes-Dennis J, Berghella V. Cervical length and prediction of
length: a population-based prospective study. Ultrasound preterm delivery. Current Opinion Obstet Gynecol 2007; 19:
Obstet Gynecol 2006; 27: 362–367. 191–195.
4. Fonseca EB, Celik E, Parra M, Singh M, Nicolaides KH, 17. Owen J, Yost N, Berghella V, Thom E, Swain M, Dildy GA,
Fetal Medicine Foundation Second Trimester Screening Group. Miodovnik M, Langer O, Sibai B, McNellis D, National Insti-
Progesterone and the risk of preterm birth among women with tute of Child Health and Human Development, Maternal-Fetal
a short cervix. N Engl J Med 2007; 357: 462–469. Medicine Units Network. Mid-trimester endovaginal sonogra-
5. Conde-Agudelo A, Romero R, Hassan SS, Yeo, L. Transvaginal phy in women at high risk for spontaneous preterm birth. JAMA
sonographic cervical length for the prediction of spontaneous 2001; 286: 1340–1348.
preterm birth in twin pregnancies: a systematic review and 18. Kusanovic JP, Espinoza J, Romero R, Gonçalves LF, Nien
meta-analysis. Am J Obstet Gyneco. 2010; 203: 128.e1–12. JK, Soto E, Khalek N, Camacho N, Hendler I, Mittal P,
6. Celik E, To M, Gajewska K, Smith GCS, Nicolaides KH, On Friel LA, Gotsch F, Erez O, Than NG, Mazaki-Tovi S,
behalf of the Fetal Medicine Foundation Second Trimester Schoen ML, Hassan SS. Clinical significance of the presence
Screening Group. Cervical length and obstetric history predict of amniotic fluid ‘‘sludge’’ in asymptomatic patients at high risk
spontaneous preterm birth: development and validation of a for spontaneous preterm delivery. Ultrasound Obstet Gynecol
model to provide individualized risk assessment. Ultrasound 2007; 30: 706–714.
Obstet Gynecol 2008; 31: 549–554. 19. Espinoza J, Gon alves LF, Romero R, Nien JK, Stites S,
7. Honest H, Forbes CA, Durée KH, Norman G, Duffy SB, Kim YM, Hassan S, Gomez R, Yoon BH, Chaiworapongsa
Tsourapas A, Roberts TE, Barton PM, Jowett SM, Hyde CJ, T, Lee W, Mazor M. The prevalence and clinical significance of
Khan KS. Screening to prevent spontaneous preterm birth: amniotic fluid sludge in patients with preterm labor and intact
systematic reviews of accuracy and effectiveness literature with membranes. Ultrasound Obstet Gynecol 2005; 25: 346–352.
economic modelling. Health Technol Assess 2009; 13: 1–627. 20. Oyelese Y, Smulian JC. Placenta previa, placenta accreta, and
8. Kenyon SL, Taylor DJ, Tarnow-Mordi W. Broad-spectrum vasa previa. Obstet Gynecol 2006; 107: 927–941.
antibiotics for spontaneous preterm labour: the ORACLE II 21. Cicero S, Skentou C, Souka A, To MS, Nicolaides KH.
randomised trial. Lancet 2001; 357(9261): 989–994. Cervical length at 22–24 weeks of gestation: comparison
9. Sotiriadis A, Papatheodorou S, Kavvadias A, Makrydimas G. of transvaginal and transperineal-translabial ultrasonography.
Transvaginal cervical length measurement for prediction of Ultrasound Obstet Gynecol 2001; 17: 335–340.

Slides summarizing practical points, with

accompanying illustrations and videoclips,
are provided as supporting information
online.

Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd. Ultrasound Obstet Gynecol 2015; 45: 358–362.