Research www. AJOG.

org

OBSTETRICS
The rate of cervical change and the phenotype
of spontaneous preterm birth
Jay D. Iams, MD; Deborah Cebrik, MS, MPH; Courtney Lynch, MPH, PhD; Nicholas Behrendt, MD; Anita Das, MPH, PhD

OBJECTIVE: Preterm birth is classified by the presence of uterine con- RESULTS: Of 2521 eligible women, 128 were delivered after preterm labor
tractions and/or amniorrhexis at clinical presentation. This classifica- and 106 after preterm membrane rupture; 89 delivered preterm for a med-
tion does not include prior cervical change. We hypothesized that the ical or obstetrical indication; 2198 delivered at term. The rate of change was
rate of cervical shortening before preterm birth would not differ accord- similar in women who presented with preterm labor (⫺0.96 mm/week) and
ing to clinical presentation. preterm ruptured membranes (⫺0.82 mm/week).
STUDY DESIGN: We analyzed data from a completed study of paired cer- CONCLUSION: Cervical shortening occurs at the same rate before
vical ultrasound measurements to test our hypothesis. Cervical ultrasound spontaneous preterm birth, regardless of presentation.
measurements obtained 4 weeks apart in the second trimester were related
to gestational age and clinical presentation at birth. Key words: cervix, phenotype, preterm birth, ultrasound

Cite this article as: Iams JD, Cebrik D, Lynch C, et al. The rate of cervical change and the phenotype of spontaneous preterm birth. Am J Obstet Gynecol
2011;205:130.e1-6.

V arious systems have been proposed

to classify preterm birth to under-
stand its pathogenesis and provide a
insufficiency vs medical-obstetrical
complications), presumed cause (infec-
tion vs hemorrhage vs uterine stretch),
brane integrity and infection prophy-
laxis for preterm rupture, and restora-
tion of cervical tissue strength for
framework for clinical care. Classifica- and/or whether parturition was initiated women with cervical insufficiency.
tions have been based on gestational age spontaneously or iatrogenically. These strategies have not reduced the
at birth (early vs late), clinical presenta- Gestational age at birth and clinical rate of preterm birth, yet the categories
tion (preterm labor vs preterm mem- presentation are widely used because upon which they are based continue to
brane rupture without labor vs cervical they are most easily identified and most drive basic and epidemiological thinking
often recorded. It has been more difficult about preterm birth, sometimes despite
to identify the causes of preterm birth what is known about the sequence of
From the Division of Maternal-Fetal and to ascertain whether and when par-
Medicine, Department of Obstetrics and events in term and preterm parturition.
turition has begun. Often clinical pre- The parturitional process is currently
Gynecology, The Ohio State University
sentation has been used to infer potential understood as having 3 components: myo-
College of Medicine, Columbus, OH (Drs
Iams and Lynch); AxiStat Inc, San Francisco,
causes and to mark the onset of parturi- metrial contractility, decidual activation,
CA (Ms Cebrik and Dr Das); and the tion as well. Etiologic priority is com-
and cervical ripening. These processes
Division of Maternal-Fetal Medicine, monly assigned to the signs and symp-
overlap and may occur in any sequence but
Department of Obstetrics and Gynecology, toms reported at presentation, but the
in normal parturition at term are thought
University of Colorado School of Medicine, duration of the parturitional process be-
Denver, CO (Dr Behrendt).
to begin with cervical softening and ripen-
fore that time is not well studied.
ing, followed by decidual activation and
Received Feb. 19, 2011; revised April 3, 2011; This deficit in the literature influences
accepted May 5, 2011. research as well as clinical care. Classify- myometrial contractions.1
Reprints not available from the authors. ing preterm birth according to present- The time at which contractions be-
Publicly available data provided by the ing signs and symptoms (eg, preterm la- come regular is typically recorded as the
Biostatistical Coordinating Center for the bor, preterm ruptured membranes, or point when disorganized uterine activity
National Institute of Child Health and Human has coalesced into true labor. The time of
Development, Maternal-Fetal Medicine Units
cervical insufficiency) suggests that each
of these presentations is the initial step in membrane rupture is similarly accepted
Research Network, Rockville, MD.
the parturitional sequence. Thus, dis- as evidence of decidual activation. Both
The views expressed herein are those of the
authors and do not represent the views of the tinctions made between preterm births are easily documented by physical exam-
National Institute of Child Health and Human preceded by preterm labor, preterm ination and standard tests described in
Development, Maternal-Fetal Medicine Units ruptured membranes, or cervical insuf- obstetrical texts, but their actual time of
Research Network. occurrence is uncertain. The onset of
ficiency have led to separate lines of in-
0002-9378/$36.00 cervical ripening has been even more dif-
vestigation and treatment for each: con-
© 2011 Mosby, Inc. All rights reserved.
doi: 10.1016/j.ajog.2011.05.021 traction detection and suppression for ficult to detect and measure. Investiga-
preterm labor, maintenance of mem- tion of this parturitional step has im-

130.e1 American Journal of Obstetrics & Gynecology AUGUST 2011

www.AJOG.org Obstetrics Research

In this secondary analysis, our primary

FIGURE 1
hypotheses were that the rate of change
Selection of study population
(ie, slope) in cervical length in women
presenting with spontaneous preterm la-
bor would not differ from the rate ob-
served in women who delivered after
preterm ruptured membranes and that
both would differ from the slope ob-
served in women with a medical or ob-
stetrical indication for preterm birth or
who delivered at term.
Accurate characterization of the se-
quence of events that precede spontane-
ous preterm birth might define 1 or more
preterm birth phenotypes more accu-
rately than is possible using only the ter-
minal signs and symptoms.

M ATERIALS AND M ETHODS

The Preterm Prediction Study was a pro-
spective multicenter observational study
of risk factors for preterm birth con-
ducted by the Eunice Kennedy Shriver
National Institute of Child Health and
Human Development (NICHD) Mater-
nal-Fetal Medicine Units Research Net-
work between 1992 and 1994.9 The goal
of the study was to identify new markers
of preterm birth. The study design,
methods of analysis, and results have
been previously reported.9-11 The de-
identified data set is now available on re-
Iams. Rate of cervical change and phenotype of spontaneous preterm birth. Am J Obstet Gynecol 2011.
quest from the NICHD. The current
analyses were conducted after the Bio-
proved with the advent of animal models tunity to relate cervical softening and rip- medical Institutional Review Board at
and cervical sonography.2-6 ening (called funneling or shortening the Ohio State University concluded that
Animal and human studies suggest when seen on ultrasound) to the gesta- this analysis of completely de-identified
that cervical preparation for birth begins tional age and clinical circumstances pre- data did not require further review.
soon after conception.3,7 Two phases of ceding preterm birth and to determine the The Preterm Prediction Study en-
cervical change during pregnancy before presence and duration of cervical ripening rolled women with singleton pregnan-
labor have been described: softening and prior to clinical symptoms. An observa- cies who were selected to reflect the par-
ripening.2,8 Softening occurs slowly over tional study in which cervical length mea- ity and race of women receiving prenatal
many weeks and is characterized by an in- surements are masked would be ideal but care at each of 10 participating sites.
crease in compliance with maintenance of difficult because clinicians today may feel Women with pregnancies complicated
tissue competence in a high progesterone/ compelled to react to the identification of by placenta previa, cervical insufficiency,
low estrogen environment.8 Ripening oc- short cervix by recommending an inter- or fetal anomalies were not enrolled. Ex-
curs more rapidly, over weeks or days vention. We therefore accessed clinical and cept for fetal death, major anomalies,
preceding the onset of labor2 and is accom- sonographic data from such a study per- membrane prolapse, regular painful
panied by loss of tissue compliance, de- formed in 1992-1994 to compare the rate contractions, advanced cervical dilation,
creased tensile strength, and reversal of the of cervical shortening in women who de- and oligo- or polyhydramnios, all data
progesterone/estrogen ratio, after which livered at term and preterm, with the latter collected for the study were not revealed
cervical dilation occurs in response to analyzed according to their clinical presen- to care providers or enrollees.
contractions. tation after preterm labor, preterm mem- Subjects were followed up from en-
Observation of the cervix over time with brane rupture, or a medical/obstetrical in- rollment before 24 weeks of gestation
transvaginal sonography offers an oppor- dication for preterm birth. through delivery to gather clinical and

AUGUST 2011 American Journal of Obstetrics & Gynecology 130.e2

Research Obstetrics www.AJOG.org

laboratory data at a series of 4 study visits

scheduled at 2 week intervals. Cervical TABLE 1
sonography was performed by sonogra- Demographic data
phers who were centrally certified by a PTL PPROM Indicated Term
single reviewer (J.D.I.). The cervical Demographic (n ⴝ 128) (n ⴝ 106) (n ⴝ 89) (n ⴝ 2198)
length was measured according to a pro- Age, y
.....................................................................................................................................................................................................................................
tocol described previously11 at the initial Mean (SD) 22.8 (5.38) 24.0 (5.69) 26.7 (6.00) 23.3 (5.50)
visit (visit 1) at 220/7th to 246/7th weeks’ ..............................................................................................................................................................................................................................................
Race, n (%)
gestation and again 4 weeks later, be- .....................................................................................................................................................................................................................................

tween 260/7th and 286/7th weeks’ gestation Black 93 (72.7) 76 (71.7) 62 (69.7) 1336 (60.8)
.....................................................................................................................................................................................................................................
at visit 3. White 35 (27.3) 28 (26.4) 25 (28.1) 816 (37.1)
.....................................................................................................................................................................................................................................
The primary outcome of the Preterm Hispanic 0 (0.0) 2 (1.9) 1 (1.1) 17 (0.8)
Prediction Study was a birth that fol- .....................................................................................................................................................................................................................................
Asian 0 (0.0) 0 (0.0) 0 (0.0) 10 (0.5)
lowed spontaneous preterm labor (PTL) .....................................................................................................................................................................................................................................

or preterm premature ruptured mem- Other 0 (0.0) 0 (0.0) 1 (1.1) 19 (0.9)

.....................................................................................................................................................................................................................................
branes (PPROM) before 350/7th weeks’ Mean (SD) 3.0 (1.72) 3.0 (1.79) 3.4 (2.09) 2.6 (1.60)
..............................................................................................................................................................................................................................................
gestation. Spontaneous preterm births Parity, n (%)
before 32 and 37 weeks were secondary .....................................................................................................................................................................................................................................
Nulliparous 38 (29.7) 37 (34.9) 33 (37.1) 946 (43.0)
outcomes. Labor was defined in the orig- .....................................................................................................................................................................................................................................

inal protocol as progressive cervical dila- Multiparous 90 (70.3) 69 (65.1) 56 (62.9) 1252 (57.0)
..............................................................................................................................................................................................................................................
tion and effacement. Any prior preterm deliveries, n (%) 39 (30.5) 37 (34.9) 33 (37.1) 282 (12.8)
..............................................................................................................................................................................................................................................
Preterm labor was defined as 6 or Prior preterm deliveries
more documented uterine contractions .....................................................................................................................................................................................................................................
n 39 37 33 282
per hour during the admission and 1 or .....................................................................................................................................................................................................................................

more of the following: (1) ruptured Mean (SD) 1.2 (0.57) 1.3 (0.52) 1.6 (0.71) 1.2 (0.51)
..............................................................................................................................................................................................................................................
membranes within 1 hour of onset of Any prior spontaneous preterm 3 (2.3) 3 (2.8) 1 (1.1) 56 (2.6)
contractions; (2) documented cervical deliveries, n (%)
..............................................................................................................................................................................................................................................
change; (3) cervical dilation 2 cm or PPROM, preterm premature rupture of membranes; PTL, preterm labor.
greater internal os; and (4) cervical Iams. Rate of cervical change and phenotype of preterm birth. Am J Obstet Gynecol 2011.
length of 1 cm or less or 50% or greater
effacement, before 37 weeks’ gestation.
Women who delivered before 37 weeks
after presenting with spontaneous pre-
term labor or PPROM were classified as
having a spontaneous preterm birth, in- TABLE 2
cluding those with PPROM in whom la- Change in cervical length between visit 1 (220/7th to
bor was induced. Preterm births that fol- 246/7th weeks) and visit 3 (260/7th to 286/7th weeks)
lowed the induction of labor for other
reasons or cesarean section without la- PTL PPROM Indicated Term
Variable (n ⴝ 128) (n ⴝ 106) (n ⴝ 89) (n ⴝ 2198)
bor that were performed for maternal or
fetal benefit were recorded as indicated Cervical length
slope, mm/wk
preterm births. .....................................................................................................................................................................................................................................

The current analysis was limited to n 128 106 89 2198

.....................................................................................................................................................................................................................................
women who completed both cervical ul- Mean (SD) ⫺0.96 (2.02) ⫺0.82 (1.90) ⫺0.23 (1.71) ⫺0.36 (1.75)
.....................................................................................................................................................................................................................................
trasound examinations (visits 1 and 3) 95% CI ⫺1.31 to ⫺0.60 ⫺1.18 to ⫺0.45 ⫺0.59 to 0.13 ⫺0.43 to ⫺0.28
within the protocol-specified gestational
..............................................................................................................................................................................................................................................
age intervals and for whom delivery out- NS NS
come was available (Figure 1). In this
analysis, births after 370/7th weeks are de- ..............................................................................................................................................................................................................................................
P ⫽ .0228
fined as term, and births before 370/7th (PPROM vs term plus indicated)
weeks are defined as preterm and, as in
..............................................................................................................................................................................................................................................
the original study, categorized as occur- P ⫽ .0005
ring after spontaneous preterm labor, af- (PTL vs term plus indicated)
ter spontaneous preterm PROM, or be- ..............................................................................................................................................................................................................................................
CI, confidence interval; NS, not significant; PPROM, preterm premature rupture of membranes; PTL, preterm labor.
cause of a medical indication for preterm Iams. Rate of cervical change and phenotype of preterm birth. Am J Obstet Gynecol 2011.
birth.

130.e3 American Journal of Obstetrics & Gynecology AUGUST 2011

www.AJOG.org Obstetrics Research

be multiparous, and to have a prior

FIGURE 2
preterm birth.
Change in mean cervical length between 22-24
In all 4 delivery groups, the mean slope
weeks and 260/7th to 286/7th weeks of cervical length was negative, indicat-
ing a decrease in cervical length between
visits 1 and 3. Because the hypothesis of
the analysis was to compare women with
a spontaneous preterm delivery with
those without, women with an indicated
preterm delivery were combined with
women who delivered at term. The F test
from the ANOVA indicated that the
slopes across the 3 groups were signifi-
cantly different (P ⬍ .0001).
Pair-wise comparisons indicated that
the cervical length slope was nearly iden-
tical for women who presented with PTL
(⫺0.96 mm/wk) and those who pre-
sented with PPROM (⫺0.82 mm/wk)
(Table 2 and Figure 2). The slope for
women with a medical or obstetrical in-
dication for preterm birth or who deliv-
ered at term differed significantly from
Iams. Rate of cervical change and phenotype of spontaneous preterm birth. Am J Obstet Gynecol 2011. the slopes for women who delivered pre-
term after PTL (P ⫽ .0005) and after
PPROM (P ⫽ .0228). The mean cervical
S TATISTICAL A NALYSIS range to test the primary hypothesis length measurements in women who de-
Statistical analysis was performed using that the rate of change in cervical livered with PTL or PPROM were
version 9.2 of SAS (SAS Institute, Cary, length in women who delivered pre- shorter at both visit 1 (P ⫽ .01117 and
NC). The number, mean, SD, minimum, term after PTL would not differ from P ⬍ .0001, respectively) and at visit 3
and maximum were determined for all the rate of change observed in women (P ⬍ .0001) than in women who
continuous variables. The 95% confi- who delivered after PPROM and that delivered at term and preterm for a med-
dence interval for the mean (based on both would differ from the slope ob- ical indication (Table 3).
the normal distribution) was deter- served in women with a medical or ob-
mined for mean cervical length at visits 1 stetrical indication for preterm birth or
and 3 and for cervical length slope. Fre- who delivered at term. C OMMENT
quency counts and percentages were de- This analysis supports the hypothesis
termined for all categorical data. For that cervical shortening (softening and
each woman, the slope (or rate of change R ESULTS ripening) progresses at the same rate be-
in millimeters per week) in cervical After excluding women who did not fore spontaneous preterm birth, regard-
length was determined by dividing the undergo 2 cervical ultrasound exami- less of clinical presentation, and suggests
difference in cervical length at visits 1 nations or for whom complete delivery that the process begins before 24 weeks
and 3 by the number of weeks between outcomes were unavailable, there were of gestation. Our analysis does not ad-
visit 1 and visit 3. 2521 women with data available for dress the reasons for the early onset of
A 1-way analysis of variance this analysis (Figure 1). There were 128 softening and ripening, only its timing.
(ANOVA) was used to first test the women who delivered preterm after The advantages of using this data set to
global hypothesis that the mean cervi- PTL, 106 who delivered preterm after address our hypotheses include prospec-
cal length slopes were the same for preterm membrane rupture, 89 who tively defined endpoints and masked col-
women who delivered preterm after delivered preterm for a medical or ob- lection of data by sonographers trained
PTL, after PPROM, with a medical or stetrical indication, and 2198 who de- in a standardized technique.11
obstetrical indication for preterm livered at term. The demographic Our analysis is limited, however, by
birth, and who delivered at term. A P ⫽ characteristics and prior pregnancy the need for a second measurement, re-
.05 level of significance was used for the history of the study population are moving women whose delivery oc-
F test of difference in means. Pairwise provided in Table 1. Women who de- curred between visits 1 and 3, who were
comparisons of means were then per- livered after PTL and PPROM were most likely to have demonstrated a
formed using Tukey’s studentized more likely to be African American, to short cervix with a more rapid decrease

AUGUST 2011 American Journal of Obstetrics & Gynecology 130.e4

Research Obstetrics www.AJOG.org

in length than women still pregnant at

visit 3. We are also significantly limited TABLE 3
by the absence of data collected before Cervical length measurements at visit 1 (220/7th to
visit 1 and after visit 3. Without earlier, 246/7th weeks) and visit 3 (260/7th to 286/7th weeks)
interval, and subsequent cervical length PTL PPROM Indicated Term
measurements, we can only speculate Variable (n ⴝ 128) (n ⴝ 106) (n ⴝ 89) (n ⴝ 2198)
about the causes, timing, frequency, and Visit 1 cervical
shape of the cervical change represented by length, mm
.....................................................................................................................................................................................................................................
the 2 time points available for this study. Mean (SD) 33.6 (9.22) 32.0 (8.38) 34.8 (8.96) 35.7 (7.98)
.....................................................................................................................................................................................................................................
Within these aggregated data, there 95% CI 32.0–35.2 30.4–33.6 32.9–36.7 35.4–36.1
may be several different scenarios. For
..............................................................................................................................................................................................................................................
example, although most women who
P ⬍ .0001
present with preterm amniorrhexis ex- PPROM vs term plus indicated
perience progressive cervical effacement
..............................................................................................................................................................................................................................................
at a rate similar to that shown in Figure 2, P ⫽ .01117
there may be women in whom preterm PTL vs term plus indicated
..............................................................................................................................................................................................................................................
membrane rupture occurs without prior
Visit 3 cervical
cervical shortening. We speculate that in length, mm
these women, the cause of preterm .....................................................................................................................................................................................................................................
Mean (SD) 29.4 (10.29) 28.7 (8.78) 33.8 (8.67) 34.2 (8.20)
rupture (eg, trauma), their history (un- .....................................................................................................................................................................................................................................

explained bleeding), demographic pro- 95% CI 27.6–31.2 27.0–30.4 32.0–36.7 33.9–34.5

file (atypical for spontaneous preterm ..............................................................................................................................................................................................................................................
birth), and their latency interval (longer P ⬍ .0001 PPROM vs term
than most) might define a unique phe- plus indicated
notype. Prospective studies relating pat- ..............................................................................................................................................................................................................................................

terns of cervical shortening to gesta- P ⬍ .0001 PTL vs term

tional age at onset and delivery, clinical plus indicated
..............................................................................................................................................................................................................................................
presentation, and pathologic findings CI, confidence interval; NS, not significant; PPROM, preterm premature rupture of membranes; PTL, preterm labor.
will be needed to identify additional phe- Iams. Rate of cervical change and phenotype of preterm birth. Am J Obstet Gynecol 2011.

notypes of preterm parturition. This ob-

servation does not, however, negate the quired many believers. However, subse- not the result of variable cervical compe-
principal observations of this analysis quent clinical experience13 and research tence or contractions, responds to pro-
that cervical softening and ripening pre- reports14-17 have not been wholly consis- gesterone supplementation, and occurs
cede PTL and PPROM and progress at a tent with that conclusion. at the same rate regardless of clinical pre-
similar rate before both. Cerclage is not a uniformly effective sentation, then preterm cervical soften-
In 1996 the first analysis of these data treatment for short cervix,13-16 appar- ing and ripening should be seen as the
emphasized the rising risk of preterm ently beneficial for some women with a initial clinical manifestation of preterm
birth as cervical length declined at visit prior preterm birth, of no benefit for parturition: short cervix is evidence that
1.11 Within the prevailing phenotype at others, and linked to increased risk of pathological preterm parturition has
that time, in which cervical change was preterm birth in women with twins.14 begun, regardless of its underlying
attributed primarily to uterine activity Meanwhile, supplemental progesterone cause(s).
and because the women enrolled were all has been reported to reduce the risk of
This interpretation has implications for
asymptomatic outpatients, little cre- preterm birth in some women with short
the phenotype of spontaneous preterm
dence was given to the possibility that cervix.17,18
birth and for prenatal care in the second
short cervical length at 22-24 weeks’ ges- The current analysis was initiated be-
tation was the result of the early onset trimester. Preterm birth that is preceded by
cause of these and other reports that
of pathological parturition. Instead, a challenge the preterm labor vs cervical premature cervical change may constitute
continuum of cervical competence was insufficiency paradigm, including those a subset of preterm births that differ from
postulated in which variable cervical in which short cervix may result from those where uterine contractility or decid-
resistance to subsequent contractions intrauterine inflammation and infec- ual activation are the initial steps. Our data
explained the relationship. tion19,20 and by others showing that cer- support the contention that in the progres-
This concept was subsequently en- vical change measured by digital and ul- sion of pathological preterm parturition,
dorsed by the American College of Obste- trasound examinations is superior to cervical shortening often precedes myo-
tricians and Gynecologists.12 The corollary assessment of uterine contractions to metrial contractility and decidual activa-
that a short cervix could be corrected by a predict preterm birth.21 If preterm cervi- tion and suggests that the presence of cer-
properly placed cerclage suture soon ac- cal effacement begins before 24 weeks, is vical shortening at presentation may be a

130.e5 American Journal of Obstetrics & Gynecology AUGUST 2011

www.AJOG.org Obstetrics Research

more appropriate method than clinical thology, and underlying causes of this 13. Berghella V, Daly SF, Tolosa JE, et al. Pre-
presentation to classify preterm birth. phenotype. f diction of preterm delivery with transvaginal ul-
trasonography of the cervix in patients with
The shorter cervical lengths observed high-risk pregnancies: does cerclage prevent
at visit 1 in women who later present ACKNOWLEDGMENTS prematurity? Am J Obstet Gynecol 1999;181:
with PTL and PPROM also suggest that We gratefully acknowledge the assistance of 809-15.
pathological cervical shortening begins the NICHD, the NICHD Maternal Fetal Medicine 14. To MS, Alfirevic Z, Heath VC, et al. Cervical
before 24 weeks. This observation is con- Units (MFMU) Network, the MFMU Preterm cerclage for prevention of preterm delivery in
Prediction Study protocol subcommittee, and women with short cervix: randomised con-
sistent with the increased risk of recur- trolled trial. Lancet 2004;363:1849-53.
the Biostatistical Coordinating Center at
rent preterm birth observed in women 15. Berghella V, Odibo A, To MS, Rust O, Al-
George Washington University in making the
with a previous 16-20 week birth22,23 and database available for this analysis. thuisius SM. Cerclage for short cervix on ultra-
by the high risk of early preterm birth in sonography: meta-analysis of trials using indi-
women with short cervix before 22 vidual patient data. Obstet Gynecol 2005;106:
REFERENCES 181-9.
weeks.24,25 16. Owen J, Hankins G, Iams JD, et al. Multi-
1. Romero R, Lockwood CJ. Pathogenesis of
Most women with a cervical length less spontaneous preterm labor. In: Creasy RK, center randomized trial of cerclage for preterm
than 25 mm at visit 1 in the Preterm Pre- Resnik R, Iams JD, Lockwood CJ, Romero R, birth prevention in high-risk women with short-
diction Study did not deliver preterm.11 eds. Creasy and Resnik’s maternal fetal medi- ened midtrimester cervical length. Am J Obstet
Fewer than 18% had a spontaneous pre- cine: principles and practice. Philadelphia, PA: Gynecol 2009;201:375.e1-8.
Saunders; 2009:521-44. 17. Fonseca EB, Celik E, Parra M, Singh M, Ni-
term birth before 35 weeks’ gestation,
2. Word RA, Li XH, Hnat M, Carrick K. Dynam- colaides KH; Fetal Medicine Foundation Second
and only 27% delivered before 37 weeks’ ics of cervical remodeling during pregnancy and Trimester Screening Group. Progesterone and
gestation.11 Thus, many women appar- parturition: mechanisms and current concepts. the risk of preterm birth among women with a
ently experience this first phase of pre- Semin Reprod Med 2007;25:69-79. short cervix. N Engl J Med 2007;357:462-9.
term parturition but do not progress to 3. Read CP, Word RA, Ruscheinsky MA, Tim- 18. DeFranco EA, O’Brien JM, Adair CD, et al.
mons BC, Mahendroo MS. Cervical remodeling Vaginal progesterone is associated with a de-
decidual activation or myometrial con-
during pregnancy and parturition: molecular crease in risk for early preterm birth and im-
tractility. This is consistent with the rel- characterization of the softening phase in mice. proved neonatal outcome in women with a
atively low sensitivity of cervical length Reproduction 2007;134:327-40. short cervix: a secondary analysis from a ran-
to predict preterm birth observed in the 4. Timmons BC, Fairhurst AM, Mahendroo MS. domized, double-blind, placebo-controlled trial.
original report.9,11 Temporal changes in myeloid cells in the cervix Ultrasound Obstet Gynecol 2007;30:697-705.
during pregnancy and parturition. J Immunol 19. Gomez R, Romero R, Nien JK, et al. A short
We propose that cervical sonography
2009;182:2700-7. cervix in women with preterm labor and intact
could be used to define a common phe- 5. Andersen HF, Nugent CE, Wanty SD, membranes: a risk factor for microbial invasion
notype of spontaneous preterm parturi- Hayashi RH. Prediction of risk for preterm de- of the amniotic cavity. Am J Obstet Gynecol
tion that begins with cervical ripening livery by ultrasonographic measurement of 2005;192:678-89.
and that may but does not always prog- cervical length. Am J Obstet Gynecol 1990; 20. Hassan S, Romero R, Hendler I, et al. A
163:859-67. sonographic short cervix as the only clinical
ress to decidual activation, myometrial
6. Bergelin I, Valentin L. Patterns of normal manifestation of intra-amniotic infection. J Peri-
contractility, and preterm birth. In our change in cervical length and width during preg- nat Med 2006;34:13-9.
clinical experience using cervical sonog- nancy in nulliparous women: a prospective, lon- 21. Iams JD, Newman RB, Thom EA, et al; for
raphy in women at risk of preterm birth, gitudinal ultrasound study. Ultrasound Obstet the National Institute of Child Health and Human
the penultimate clinical presentation is Gynecol 2001;18:217-22. Development Network of Maternal-Fetal Medi-
7. Leppert PC. Anatomy and physiology of cer- cine Units. Frequency of uterine contractions
often characterized primarily by ad-
vical ripening. Clin Obstet Gynecol 1995;38: and the risk of spontaneous preterm delivery.
vanced cervical effacement and dilation, 267-79. N Engl J Med 2002;346:250-5. Erratum in:
accompanied variably by pelvic pressure, 8. Timmons B, Akins M, Mahendroo M. Cervical N Engl J Med 2003;349:513.
increased vaginal discharge, visible or remodeling in pregnancy and parturition. 22. McManemy J, Cooke E, Amon E, et al. Re-
ruptured membranes, and/or contrac- Trends Endocrinol Metab 2010;182:2700-7. currence risk for preterm delivery. Am J Obstet
9. Goldenberg RL, Iams JD, Mercer BM, et al. Gynecol 2007;196:576.e1-6.
tions that are seemingly inadequate to
The preterm prediction study: the value of new 23. Edlow AG, Srinivas SK, Elovitz MA. Second
account for the degree of cervical efface- vs. standard risk factors in predicting early and trimester loss and subsequent pregnancy out-
ment and dilation. all spontaneous preterm births. NICHD MFMU comes: what is the real risk? Am J Obstet Gy-
Women with this phenotype of pre- Network. Am J Public Health 1998;88:233-8. necol 2007;197:581.e1-6.
term cervical ripening seem to perceive 10. Mercer BM, Goldenberg RL, Das A, et al. 24. Hibbard JU, Tart M, Moawad AH. Cervical
The Preterm Prediction Study: a clinical risk as- length at 16-22 weeks’ gestation and risk for pre-
contractions primarily as pelvic pressure
sessment system. Am J Obstet Gynecol 1996; term delivery. Obstet Gynecol 2000;96:972-8.
or increased vaginal discharge, in con- 174:1885-95. 25. Owen J, Yost N, Berghella V, et al; for the
trast to women with an unripe cervix, for 11. Iams JD, Goldenberg RL, Meis PJ, et al. The National Institute of Child Health and Human
whom pain is more often the primary length of the cervix and the risk of spontaneous Development, Maternal-Fetal Medicine Units
complaint.26 Further research relating premature delivery. N Engl J Med 1996;334: Network. Mid-trimester endovaginal sonog-
567-73. raphy in women at high risk for spontaneous
cervical length at presentation to specific
12. Cervical insufficiency. ACOG practice bulle- preterm birth. JAMA 2001;286:1340-8.
presenting symptoms would be useful to tin no. 48. The American College of Obstetri- 26. Olah KS, Gee H. The prevention of preterm
characterize the demography, perinatal cians and Gynecologists. Obstet Gynecol delivery— can we continue to ignore the cervix?
course, neonatal outcomes, placental pa- 2003;102:1091-9 (reaffirmed 2008). Br J Obstet Gynaecol 1992;99:278-80.

AUGUST 2011 American Journal of Obstetrics & Gynecology 130.e6