Cervical Ripening:Biochemical,Molecular, andClinicalConsiderations

HOLGER MAUL, MD, MMSC,* LYNETTE MACKAY, BS,wand ROBERT E. GARFIELD, PhDw*Department of Obstetrics and Gynecology, Universityof Heidelberg, Heidelberg, Germany; and wUniversity ofTexas Medical Branch, 301 University, Department ofObstetrics and Gynecology, Division of ReproductiveSciences, Galveston, Texas

Abstract: The physiologic and pathologic changes ofthe uterine cervix during pregnancy leading tocervical ripening are not well understood thoughare related to the chief pathology and a commonlyperformed intervention in obstetrics: Preterm birthand labor induction. Normal cervical ripening isthought to be controlled by a variety of hormonalchanges occurring during pregnancy leading tosoftening and dilation. Abnormal premature ripen-ing, usually resulting in preterm labor, is thought tobe associated with infection and inflammatoryevents. Despite many studies of the cervix, we stillrely upon relatively crude methods for clinicalevaluation of the cervix. In the past several years,we have developed and evaluated a method tomeasure cervical collagen noninvasively using aninstrument called Collascope. Studies in animals andhumans conducted in a variety of settings indicate

that cervical function can be successfully monitoredusing the Collascope during pregnancy. We suggestthat this technique might be useful to better definemanagement in cases of spontaneous preterm andinduced term cervical ripening.Key words: cervical ripening, preterm labor, colla-scope, light-induced fluorescence, LIF, cervicalincompetence, induction of labor, parturition

In developed countries, obstetrics hasdramatically changed during the pastdecades. Although maternal morbidityand mortality were the key issues in thefirst half of the 20th century, the secondhalf of the 20th century lead to discov-eries that further reduction of maternalmorbidity and mortality was difficultand methodologies to do so were in-effective. Therefore, obstetrics in thesecond half of the 20th century becamefocused on the fetus, fetal complications,

551

Correspondence: Robert E. Garfield, PhD, Professor,Director of the Division of Reproductive Sciences,Department of Obstetrics and Gynecology, TheUniversity of Texas Medical Branch, Galveston, TX77555-1062. E-mail: rgarfiel@utmb.edu

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 49 / NUMBER 3 / SEPTEMBER 2006

CLINICAL OBSTETRICS AND GYNECOLOGYVolume 49, Number 3, 551563r 2006, Lippincott Williams & Wilkins

and reduction of fetal and neonatalmorbidity and mortality. In fact, forthe first time, the fetus itself is nowconsidered to play a significant role inmedicine. However, at present, fetal andneonatal morbidity and mortality seemto have reached a plateau at 5 to 10 per1000 live births in Western Europe andNorth America.From our point of view, the next

factors that must be taken into consid-eration are the processes themselvesleading to maternal and fetal pathology.That is to say that focus should reside onnot only the outcomes but also themethods to assess these outcomes to findfurther clues for improvement of humanwell-being.Therefore, future parturition research

will focus on differing issues includingsafety, patient acceptance, and toleranceof various interventions. Parturition is acomplex process involving different com-partments with changes on the maternalside (myometrium and cervix) and on thefetal side (fetus itself, fetal membranes)leading to delivery of the fetus. Theinterrelationship between these compart-ments is not well understood. Further,the methods to monitor myometrial andcervical changes involved in parturitionare crude and only the end points can bedetected reliably. Active labor (ie, sig-nificant uterine contractions leading tochanges in the shape of the uterus whilecontracting over the incompressible vo-lume containing amniotic fluid, placenta,and fetus), cervical changes (ie, softeningand dilation of the cervix), and changesin fetal membranes occur at differenttimes (see model of these changes,Fig. 1). Early changes, such as increasedelectrical conductance between uterinesmooth muscle cells, or the degradationof cervical collagen leading to softeningof the cervix are not taken into account,unless one relies on only subjective andnot well reproducible methods such asdigital palpation of the cervix. Table 1

summarizes the current methods andtheir accuracy used in uterine monitoringor screening of labor.Preterm labor is the tautology for a

syndrome occurring in a variety ofdifferent clinical patterns responsiblefor approximately 7% to 11% of neo-nates being born prematurely eachyear.1,2 Immaturity and prematurity ac-count for 85% of infant mortality and50% of infant neurologic morbidity.3

The development of effective means toprevent or reduce the occurrence ofpreterm delivery depends on the under-standing of the conditions that initiatelabor. In fact, despite intensive investiga-tion, the rate of preterm births has notdeclined in several decades. The keyissue, in judgment of the effectivenessof preterm labor, is the ability of therigid or soft cervix to withstandthe pressure induced by the uterinecontractions. In addition, pretermcervical ripening may result in cervicalincompetence and may contribute tothe occurrence, enhancement, and main-tenance of uterine contractions bymechanisms only partially understood.On the other hand, with increasing

frequency, labor is induced using drugsthat enhance cervical ripening and

FIGURE 1. Model of uterine cervical andfetal membrane changes during parturition.The uterus, cervix, and fetal membranes gothrough changes as they prepare for laborand ultimately for delivery. These changesmay be driven, at least partially, by indepen-dently timed mechanisms.35

552 Maul et al

support uterine contractions to effectdelivery and prevent complications formothers and newborns that would other-wise be associated with continuation ofpregnancy. Labor induction has becomea common procedure in obstetrics that isaccompanied by a number of potentialcomplications such as uterine overstimu-lation, uterine rupture, and fetal com-promise.

Cervical RipeningAt present in clinical routine there is noreliable, objective, and quantitativemethod to monitor early cervical changesfor a variety of clinical scenarios. Thesewould include detection of the earlyphases of cervical ripening leading tocervical incompetence and eventual pre-term birth, and the assessment of theeffect of cervical ripening agents forlabor induction.Indeed, many biochemical and func-

tional changes occur in cervical connec-tive tissue during pregnancy46 that aresummarized in the term cervical ripen-ing. These changes precede the clinicalsymptoms, which are currently evalu-ated, that is, uterine contractions andcervical dilatation. Cervical ripening is achronic process, which begins within thefirst trimester of pregnancy and progres-sively proceeds until term, and is usuallydescribed as softening, effacement (thin-

ning), and dilation of the cervix. (Thelatter 2 are acute events, which occurbefore delivery). Obviously with this3-step preparation process, which mustoccur in sequence, complex systems mustbe in place to control each forward andirreversible step. Softening must beconsidered a vital process because efface-ment and dilation cannot occur withoutremodeling of the cervix during thesoftening phase. However, none of ourcurrent methods to assess the cervixmeasures this event. Effacement anddilation are often associated with orcontributed to uterine contractions, butit is evident that softening occursindependent of contractions. Further,effacement and dilation are oftenpresent without uterine contractions.The cervix, which is dominated by

fibrous connective tissue, is composed ofan extracellular matrix consisting pre-dominantly of collagen (70% Type I and30% Type III) along with elastin andproteoglycans, and a cellular portionconsisting of smooth muscle, fibroblasts,epithelium, and blood vessels.7 Through-out most of gestation the cervix remainsrigid and closed to secure the products ofconception. A dramatic functional shiftoccurs during parturition as it dilatesthrough a cervical destructive process.The cascade responsible for the processof cervical ripening, and which finallyenables uterine contractions to efface

TABLE 1. Methods and Their Accuracy Used in Uterine Monitoring or Screening of Labor34

Accuracy Invasive

Monitoring of contractions in combination withpelvic examination

Moderate No

Monitoring the state of the cervix Moderate NoSymptomatic self-monitoring Low NoIntrauterine pressure monitoring High YesExternal uterine monitor (tocodynamometer) Erratic NoUltrasound, remote sensing, temperaturemonitoring, blood flow monitors

Mixed No

Endovaginal ultrasonography High negative predictive values NoFetal fibronectin screening test High negative predictive values NoSalivatory estriol test High negative predictive values No

Cervical Ripening 553

and dilate the cervix, is still not fullyunderstood. There is a progressive soft-ening of the cervix beginning earlypregnancy in women, and in rats, duringthe second half of gestation, beginningaround day 13. Studies in cervicalresistance and light-induced fluorescence(LIF) have measured these changes inrats.8,9 Nonpregnant and timed-pregnantrats were examined in our laboratory forcervical softening using a cervimeter. The

uterine cervices were collected at severaltime points during pregnancy and weresubjected to incremental stretching. Thelength-tension curve was derived and theslope of the regression line indicated thedegree of cervical extensibility (Fig. 2). Inour laboratory, LIF studies with theCollascope paralleled these results show-ing a gradual ripening from mid-gesta-tion in rats (Figs. 3, 4 for illustration ofthe Collascope).

FIGURE 2. Changes in the rat cervix innonpregnant animals, during pregnancy andpostpartum. Cervical ripening as measuredwith the cervimeter.8,9 The slope of cervicalextensibility changes is shown.

FIGURE 3. Gradual changes in rat cervix during cervical ripening asmeasured with collascope.8,9 Shown are LIF measurements as ratios ofthe collagen peak versus the reference peak values.

FIGURE 4. Diagram of the components ofthe collascope. Parts include a light source,fiber optic cable, spectrometer, a CCDcamera, and a computer.34

554 Maul et al

Many studies show that hormonesseem to control cervical ripeningalthough the mechanisms and effects oneach step in ripening are not clear.Because antiprogestins induce cervicalripening, this process seems to be con-trolled at least in part by hormones,including progesterone and estrogen,10

relaxin11 and androgens.12 Knock-outmouse models, such as steroid 5a-reduc-tase type 1,13,14 demonstrate failure todeliver owing to failure of cervicalripening. The gene is involved in andro-gen production, and the parturitiondefect can be overcome by treatmentwith 5a-reduced androgens. The enzymeplays an essential role in progesteronecatabolism in the cervix inhibiting theconversion of testosterone to dihydrotes-

tosterone resulting in localized failure ofprogesterone withdrawal. Other mutantmodels, such as null PGF2a, do notundergo luteolysis at term and highprogesterone levels lead to failed parturi-tion. In animal models, the physiologicdecrease in the concentration of proges-terone during the third trimester ofpregnancy initiates a cascade that isanalogous to an inflammatory responsewith influx of polymorphonuclear cells15

and release of matrix-metalloproteinasesinto the cervical stroma, culminating inthe degradation of collagen7 (in humans,a functional decrease of progesteroneconcentration is mostly probably due todecreased sensitivity of the hormonereceptor). Autocrine and paracrine med-iators like cytokines,16 prostaglandins,17

Chart Window

pres

sure

3 (m

mHg)

0

10

20

30

40

em

g3 (u

v)

-2.00e+04

-1.00e+04

0.00e+00

1.00e+04

2.00e+04

10700 10800 10900 11000 11100 11200

Chart Window

pres

sure

5 (m

mHg)

0

10

20

30

em

g5 (u

v)

-2.00e+04

-1.00e+04

0.00e+00

1.00e+04

2.00e+04

8100 8200 8300 8400 8500 8600 8700

2/5/02 11:11:22. 560 AM

2/5/02 11:53:25. 560 AM

Uterine Pressure

Uterine EMG

Uterine Pressure

Uterine EMG

Uterine Recordings from Rats at Term During Labor

Control

BPN

FIGURE 5. Recordings of uterine electromyographic (EMG) bursts andcontractions in laboring control and neurectomized rats. The control rat showspressure and spikes superimposed on the uterine contractions as it pushes. Theneurectomized rat does not push and, therefore, does not have spikes on top of thecontractions (unpublished).

Cervical Ripening 555

platelet activating factor,18 and nitricoxide19 produced by different isoformsof its synthases20 have been shown totake part in physiologic or pathologiccervical ripening thereby forminga complex and poorly understood net-work with short-circuits and parallelpathways, widely influenced by hor-mones and their sensitivities.21 Conver-sely, cytokines (interleukin-1b22), prost-aglandins (eg, prostaglandin E223),platelet activating factor, nitric oxide24

and antiprogestins10 are potent agentsthat have been used for induction ofcervical ripening in animals and hu-mans.18,24 Similarly, lipopolysaccharidesare well known to trigger the hypothe-tical cascade described above. Therefore,

these factors may also be operativein premature cervical ripening associatedwith genital tract infection.24

The nervous system has significantinvolvement in the process of reproduc-tion and could also be involved incervical ripening. Sensory, sympathetic,and parasympathetic fibers are numer-ous in the cervix.25 The sensory compo-nent largely comes from the pelvic nervesL6-S1 dorsal root ganglia and terminatesin the cervix as unmyelinated, small,capsaisin-sensitive sensory neurons.These synthesize neurotransmitters suchas vasoactive neuropeptides calcitoningene-related peptide, substance P, andsecretoneurin, which are locally releasedin the cervix and act through their

pre

ssu

re4

(m

mH

g)

0

10

20

30

em

g4

(m

V)

-100

-50

0

50

100

23700 23750 23800 23850

Chart Window

pre

ssu

re1

(m

mH

g)

-10

0

10

20

30

em

g1

(m

V)

-50

0

50

100

25000 25050 25100 25150 25200

Abdominal Pressure

EMG bursts

Abdominal Pressure

EMG bursts

Control

BPN

Abdominal Recordings from Rats at

Term During Labor

Chart Window

FIGURE 6. Recording of abdominal muscle EMG bursts and abdominal pressurespikes during pushing in laboring control rat (upper traces) and in laboringneurectomized rat (bottom traces). Note that the neurectomized rat does not push(unpublished).

556 Maul et al

receptors to induce inflammatorylikecervical changes with vasodilatation,vascular leakage, and plasma and leuko-cyte extravasations. Estrogen acts via itsreceptors, particularly ER-a, to modu-late neuropeptide synthesis. Our studies(presented at the 2004 scientific meetingof the Society for Gynecologic Investiga-tion, Abstract No. 755), and others26,27

show that pelvic neurectomy in rats doesnot inhibit cervical ripening or initiationof labor, but blocks normal parturitionby elimination of the pushing response(contraction of abdominal muscles) as-sociated with forceful contractions of theuterus during labor. For our telemetricstudies in conscious, unrestrained rats,electrodes were attached to uterine andabdominal muscles, and pressure cathe-ters were introduced into the uterineand abdominal cavities, to record

electromyographic bursts, and pressurechanges from laboring control and neur-ectomized rats (Figs. 5, 6). The sensoryfibers of the pelvic nerve seem to relayresponses from the stretching cervix tothe abdominal muscles of the rat tostimulate these contractions to assist inpushing the pups through the birthcanal.Very little is known about how cervi-

cal ripening can be prevented or inhib-ited. Recently, it has been found thatoral administration of a platelet activat-ing factor receptor antagonist in ratssignificantly increases the duration ofparturition.28 Similarly and in additionwe have found that platelet activatingfactor-antagonist WEB-2170 effectivelyinhibits preterm cervical ripeninginduced by lipopolysaccharides in anin vivo animal model, whereas it was

Model for Cervical Ripening

NO

Vascularpermeability MMPs

GAGremodeling

Infection TermLPS P4 E2

IL -1 IL-8 TN F- (chemotaxis, leukocyte infiltration)

COX2 iNOS

PGE2

ApoptosisCollagenremodeling

++++

++ amplification

Preterm

Relaxin

ECM Degradation

FIGURE 7. Diagram summarizing steps involved in pathologic orphysiologic control of cervical ripening and extracelluar matrixdegradation involving lipopolysaccharides (LPS) or steroid hormones(P4 and E2), cytokines (IL-10, IL-8, TNF-a), prostaglandin, and nitricoxide synthesis enzymes (COX-2 or NOS) leading to changes in vascularpermeability, matrix metalloproteinases (MMPs), gycosoaminoglycans(GAG), collagen remodeling, and apoptosis.

Cervical Ripening 557

not effective in overcoming induction ofthis process by an antiprogestin (RU486).29 A variety of mediators contributeto cervical ripening as summarized inFigure 7. This figure illustrates that thereare at least 2 parallel pathways that leadto extracellular matrix degradation andripening of the cervix involving a patho-logic pathway with infection and aphysiologic pathway involving changesin steroids with both pathways conver-ging on the generation of cytokines,prostaglandins, and nitric oxide resultingin cervical remodeling.

Measurements of CervicalSofteningEvaluation of the first common step,cervical softening, has been hypothesizedto be useful for the early evaluation of

alterations in cervical ripening eitherspontaneously occurring or medicallyinduced. Secondly, it has been suggestedto use such a parameter for the predic-tion of preterm birth.During the past 6 years, our group has

developed and improved a procedure toobjectively detect and quantify earlychanges in the cervical ripening processduring pregnancy. The method is non-invasive, easy to master, rapid, andallows for the quantitative estimation ofcervical ripening. The device, called aCollascope (Fig. 4), was developed in ourlab and uses the light-induced fluores-cence of cross-linked collagen to permitthe attending physician to differentiatebetween the labor versus the nonlaborstate of the cervix earlier than themethods used clinically today (Fig. 8).No other devices or tests currently in usein the clinic can do this.

FIGURE 8. Cervical changes and methods of evaluation.

558 Maul et al

FLUORESCENCE SPECTROSCOPY OFCOLLAGENFluorescence spectroscopy is a widelyused research tool in the biosciences,primarily due to the amount of informa-tion that it can reveal in terms ofmolecular and physical states.30 Fluores-cence spectra offer important details onthe structure and dynamics of macro-molecules and their location at micro-scopic levels. Fluorescence spectroscopyhas been used to examine collagencontent of a variety of tissues includingcancers.31 We have used this methodol-ogy taking advantage of the fluorescentproperties of cross-linked collagen in thecervix.

CERVICAL LIFOne study involving cervical collagenwas conducted to investigate gestationalchanges of cervical LIF, an index forcross-linked collagen, and to estimatewhether LIF correlates with the time-to-delivery interval and is predictive ofdelivery within 24 hours.32 In 29 healthygravidas without signs of labor, LIF wasmeasured about weekly during the lasttrimester and also in 29 patients withsigns of labor. Cervical LIF was ob-tained noninvasively using the colla-scope, which was specifically designedfor this purpose. Spearmans correlation,1-way analysis of variance followed bymultiple pairwise comparisons, and re-ceiver operating characteristics curveanalysis (ROC analysis, Fig. 9) wasperformed. LIF-measurements corre-lated negatively with gestational age(r= 0.508; P

we compared the LIF parameter with theBishop score. Forty-one gravidas atZ37completed weeks of gestation wereenrolled. Twenty-five of the subjectsreceived a half tablet of misoprostol(Cytotec) 50mg intravaginally and 16subjects received dinoprostone (Prepidil)500 mg gel intracervically. The type ofprostaglandin administered was guidedby obstetric indications as determined byphysicians not involved in the researchproject. LIF was measured in a standardfashion and 3 separate readings weretaken from each subject at the samelocation on the cervix. Bishop score andLIF were measured before and 4 hoursafter prostaglandin application by thesame investigator. We found the LIFratio of cervical collagen in all 41patients before prostaglandin applicationto be significantly decreased from0.9820.04 to 0.8850.037 four hoursafter application (P=0.025) (Fig. 11).The decrease in LIF of collagen corre-lated with the initial LIF beforePG-application (r=0.61, P

of 19 patients with an initial highcollagen fluorescence (>1) showed adecrease after prostaglandins in contrastto only 11/22 patients with an initial lowcollagen fluorescence value

gestation and parturition. Prenat NeonatMed. 1998;3:103105.

7. Ludmir J, Sehdev HM. Anatomy andphysiology of the uterine cervix. ClinObstet Gynecol. 2000;43:433439.

8. Glassman W, Byam-Smith M, GarfieldRE. Changes in rat cervical collagenduring gestation and after antiprogester-one treatment as measured in vivowith light-induced autofluorescence.Am J Obstet Gynecol. 1995;173:15501556.

9. Shi L, Shi SQ, Saade GR, et al. Changesin cervical resistance and collagen fluor-escence during gestation in rats.J Perinat Med. 1999;27:188194.

10. Chwalisz K, Garfield RE. Antiproges-tins in the induction of labor. Ann NYAcad Sci. 1994;734:387413.

11. Downing SJ, Sherwood OD. The phy-siological role of relaxin in the pregnantrat. III. The influence of relaxin oncervical extensibility. Endocrinology.1985;116:12151220.

12. Mochizuki M, Maruo T. Effect ofdehydroepiandrosterone sulfate oninterleukin-8 receptor during cervicalripening in late pregnancy. Acta PhysoilHung. 1985;65:267274.

13. Mahendroo MS, Cala KM, Russell DW.5 alpha reduced androgens play a keyrole in murine parturition. Mol Endocri-nol. 1996;10:380392.

14. Mahendroo MS, Porter A, Russell DW,et al. The parturition defect in steroid5a-reductase type 1 knockout mice isdue to impaired cervical ripening.Mol Endocrinol. 1999;13:981992.

15. Junqueira LC, Zugaib M, Montes GS,et al. Morphologic and histochemicalevidence for the occurrence of collage-nolysis and for the role of neutrophilicpolymorphonuclear leukocytes duringcervical dilation. Am J Obstet Gynecol.1980;138:273281.

16. Sennstrom MB, Ekman G, Westergren-Thorsson G, et al. Human cervicalripening, an inflammatory processmediated by cytokines. Mol HumReprod. 2000;6:375381.

17. Calder AA, Embrey MP, Tait T. Ripen-ing of the cervix with extra-amnioticprostaglandin E2 in viscous gel before

induction of labour. Br J ObstetGynaecol. 1977;84:264268.

18. Maul H, Shi L, Marx SG, et al. Localapplication of platelet-activating factorinduces cervical ripening accompaniedby infiltration of polymorphonuclearleukocytes in rats. Am J Obstet Gynecol.2002;187:829833.

19. Shi L, Shi SQ, Saade GR, et al. Studiesof cervical ripening in pregnant rats:effects of various treatments. Mol HumReprod. 2000;6:382389.

20. Tornblom SA, Maul H, Klimaviciute A,et al. mRNA expression and localizationof bNOS, eNOS, and iNOS in humancervix at preterm and term labour.Reprod Biol Endocrinol. 2005;10:33.

21. Marx SG, Wentz MJ, Mackay LB, et al.Effects of progesterone on iNOS, COX-2, and collagen expression of the cervix.J Histochem Cytochem. 2006. In press.

22. Chwalisz K, Benson M, Scholz P,et al. Cervical ripening with the cyto-kines interleukin 8, interleukin 1 betaand tumour necrosis factor alpha inguinea-pigs. Hum Reprod. 1994;9:21732181.

23. Thomson AJ, Lunan CB, Cameron AD,et al. Nitric oxide donors induce ripeningof the human uterine cervix: a rando-mised controlled trial. Br J ObstetGynaecol. 1997;104:10541057.

24. Kniss DA, Iams JD. Regulation ofparturition update. Endocrine and para-crine effectors of term and pretermlabor. Clin Perinatol. 1998;25:819836.

25. Mowa CN, Papka RE. The role ofsensory neurons in cervical ripening:effects of Estrogen and neuropeptides.J Histochem Cytochem. 2004;52:12491258.

26. Higuchi T, Uchide K, Honda K, et al.Pelvic Neurectomy abolishes the fetus-expulsion reflex and induces dystocia inthe rat. Exp Neurol. 1987;96:443455.

27. Martnez-Gomez M, Cruz Y, Pacheco P,et al. The sensory but not muscularpelvic nerve branch is necessary forparturition in the rat. Physiol Behavior.1998;63:929932.

28. Zhu YP, Hoffman DR, Hwang SB, et al.Prolongation of parturition in thepregnant rat following treatment with a

562 Maul et al

platelet activating factor receptorantagonist. Biol Reprod. 1991;44:3942.

29. Maul H, Shi L, Marx SG, et al. Platelet-activating factor antagonist WEB-2170inhibits lipopolysaccharide-induced, butnot antiprogestin-induced, pretermcervical ripening in timed-pregnantrats. Am J Obstet Gynecol. 2003;189:963967.

30. Undenfriend S. Fluorescence Assay inBiology and Medicine.Vol. I. New York:Academic Press; 1962.

31. Ramanujam N, Mitchell MF, Mahade-van A, et al. Fluorescence spectroscopy:a diagnostic tool for cervical intrepithe-lial neoplasia (CIN). Gynecol Oncol.1994;52:3138.

32. Maul H, Olson G, Fittkow CT, et al.Cervical light-induced fluorescence in

humans decreases throughout gestationand before delivery: preliminary obser-vations. Am J Obstet Gynecol. 2003;188:537541.

33. Fittkow CT, Maul H, Olson G, et al.Light-induced fluorescence of the humancervix decreases after prostaglandin ap-plication for induction of labor at term.Eur J Obstet Gynecol Reprod Biol. 2005;123:6266.

34. Garfield RE, Maul H, Maner WL, et al.Uterine electromyography and light-induced fluorescence in the managementof term and preterm labor. J SocInvestig. 2002;9:265275.

35. Garfield RE, Chwalisz K, Shi L, et al.Instrumentation for the diagnosis ofterm and preterm labour. J PerinatMed. 1998;26:413436.

Cervical Ripening 563