OBSTETRICS

OBSTETRICS

Diagnosis of Premature Rupture of Membranes: Inspiration From the Past and Insights for the Future
Amira El-Messidi, MD, FRCSC, RDMS,1 Alan Cameron, MD, FRCOG, FRCP (Glas)2
1 2

Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Royal Victoria Hospital, McGill University, Montreal QC Fetal Medicine, Department of Obstetrics and Gynaecology, The Queen Mothers Hospital, Glasgow, Scotland

Abstract
Objective: To review the diagnostic methods described to confirm premature rupture of membranes during pregnancy, and to assess their effectiveness in establishing the diagnosis. Data Sources and Extraction: The medical literature was searched to identify all relevant studies and reviews on methods for diagnosis of membrane rupture published in English up to January 31, 2009. Medline and the Cochrane databases were searched, and reference lists in identified articles were also examined. Articles not available through journals online editions were retrieved by manual search. Study Selection: We identified 71 original studies and reviews on diagnostic methods of chorioamniotic membrane rupture published in English. These articles were reviewed and results were summarized based on the diagnostic test assessed. Conclusion: Recognition of the importance of and difficulties in confirmation of rupture of the chorioamniotic membranes pervades past and present obstetric publications. The subjectivity and poor sensitivity of early diagnostic techniques for confirmation of ruptured membranes sparked technical advancements using biochemical markers. None of these biochemical tests have gained popularity, although novel techniques involving placental markers such as placental alpha microglobulin-1 may provide a future solution to the problem of diagnosing chorioamniotic membrane rupture.

en anglais. Ces articles ont t analyss et les rsultats ont t rsums en fonction du test diagnostic valu. Conclusion : La reconnaissance de limportance de la confirmation de la rupture des membranes chorioamniotiques et de ses difficults est omniprsente dans les publications passs et prsentes du domaine de lobsttrique. La subjectivit, la faible sensibilit et les premires techniques diagnostiques visant la confirmation de la rupture des membranes ont donn lieu des perces techniques faisant appel des marqueurs biochimiques. Aucun de ces tests biochimiques ne sest dmarqu, et ce, malgr le fait que des techniques novatrices faisant appel des marqueurs placentaires (comme lalpha-microglobuline-1 placentaire) puissent fournir une future solution au problme que pose le diagnostic de la rupture des membranes chorioamniotique. J Obstet Gynaecol Can 2010;32(6):561569

INTRODUCTION

remature rupture of membranes may occur at term or immediately preceding labour, or it may be an unexpected complication during the preterm period, when it is referred to as preterm premature rupture of membranes.

Rsum
Objectif : Analyser les mthodes diagnostiques dcrites pour confirmer la rupture prmature des membranes pendant la grossesse, ainsi quafin den valuer lefficacit pour ce qui est de ltablissement du diagnostic. Sources et extraction des donnes : Des recherches ont t menes dans la littrature mdicale en vue didentifier toutes les tudes et analyses pertinentes portant sur les mthodes de procder au diagnostic de la rupture des membranes qui ont t publies en anglais jusquau 31 janvier 2009. Des recherches ont aussi t menes dans Medline et les bases de donnes Cochrane, et les listes de rfrences des articles identifis ont galement t examines. Les articles ntant pas disponibles par lintermdiaire des ditions lectroniques des revues spcialises ont t rcuprs par recherche manuelle. Slection des tudes : Nous avons identifi 71 tudes et analyses originales portant sur les mthodes permettant de diagnostiquer la rupture des membranes chorioamniotiques qui ont t publies

Chorioamniotic membrane rupture may have several underlying causes, although in many cases PROM and PPROM will not have recognized etiologies. The pathophysiology leading to PROM at term has been shown to be different from the pathophysiology leading to PPROM. At term, weakening of the membranes may result from physiologic changes combined with shearing forces induced by contractions.14 Generalized weakness of the membranes has been more difficult to identify with prematurely ruptured membranes.5 PPROM may result from a focal deficit rather than a generalized weakness of the membranes.6 Term PROM complicates approximately 8% of pregnancies.7 Among these, approximately 50% of affected women will begin labour spontaneously within 12 hours, 70% within 24 hours, 85% within 48 hours, and 95% within 72 hours.79 Fetal morbidities associated with term PROM include
JUNE JOGC JUIN 2010 l 561

OBSTETRICS

ascending infection and in utero cord compression.7 Maternal risks of term PROM include chorioamnionitis and postpartum febrile morbidity.7,9 Preterm PROM, a complication of 2% to 20% of all deliveries,10 is a known important contributor to maternal and perinatal morbidity and perinatal mortality. Latency in PPROM, defined as the interval between PROM and birth,7 is known to be inversely related to gestational age at rupture, and is also related to a multitude of other factors, including number of fetuses,11 severity of oligohydramnios,12 myometrial thickness,13 and the existence of maternal or obstetrical complications. The major cause of perinatal morbidity and mortality associated with PPROM is prematurity.7 Morbidities related to prematurity include respiratory distress syndrome, necrotizing enterocolitis, interventricular hemorrhage, cerebral palsy, and sepsis.7 Other complications include in utero umbilical cord compression, cord prolapse and fetal distress, fetal malpresentation, placental abruption, chorioamnionitis with subsequent endometritis, and risk of operative delivery from this multitude of factors.7 Maternal sepsis is a rare but life-threatening complication reported in nearly 1% of cases.7 For over 70 years, there has been controversy among health care professionals about the optimal approach to clinical assessment and diagnosis of prematurely ruptured membranes. In most cases, membrane rupture can be confirmed by documenting amniotic fluid leakage from the cervical os with visualization of pooling in the posterior vaginal fornix.14 However, the diagnosis of PROM is difficult if there is a slow fluid leak or any bleeding, or when the classic gush of fluid does not occur.15 In addition, the relatively small amount of amniotic fluid present early in gestation further challenges the diagnosis of ruptured membranes.16 Ladfors et al.17 showed that even later in pregnancy (after 34 weeks), speculum examination for visualization of amniotic fluid carries a 12% false negative rate when no fluid is seen. In 1951, Schumann18 described the two sac theory of membrane rupture: he stated that amniotic fluid can dissect between the two layers of membranes and produce a bulge that ruptures into the vagina, leaving one intact layer with remaining fluid enclosed and appearing to rupture at a later time. In such cases, women give a history of fluid leakage, and tests for leakage should be positive, although membranes appear clinically to be intact at delivery. Occasionally, women present with a history suspicious for membrane rupture, but clinical
ABBREVIATIONS AFI PROM amniotic fluid index premature rupture of membranes PAMG-1 placental alpha-microglobulin-1 PPROM preterm premature rupture of membranes

examination is negative or inconclusive; they subsequently return with clearly identifiable rupture of the membranes. Whether these cases represent preliminary minimal transudation of fluid across weakened membranes or minimal leakage around a firmly applied fetal presenting part cannot be determined. In approximately 20% to 25% of cases, rupture of membranes is not grossly apparent.10,19 A patients history may suggest membrane rupture, but test results are non-confirmatory, creating an obstetrical dilemma. Early and accurate diagnosis of membrane rupture would allow for gestational age-specific interventions to optimize perinatal outcome and minimize serious complications.20 Therefore, the search for an ideal test to diagnose membrane rupture definitively with no delay continues. Although initial studies may be encouraging, diagnostic techniques are usually found subsequently to be limited by inaccuracies from false positives and false negatives with poorer sensitivities and specificities than originally anticipated. The ideal test should be simple, rapid, inexpensive, and non-invasive. Optimally, the accuracy of the test should not be hampered by the presence of blood, semen, infected urine, or other contaminants. An accurate biochemical marker for membrane rupture should have a high concentration in the amniotic fluid, a low concentration in maternal blood, and an extremely low background concentration in cervicovaginal discharge with intact membranes.10 We sought to review the diagnostic methods currently used to confirm premature rupture of membranes and to assess their effectiveness in establishing the diagnosis.
METHODS

We searched Medline and the Cochrane databases for all relevant studies and reviews published in English on accuracies of diagnostic methods for the diagnosis of rupture of the membranes up to January 31, 2009. Key words used for the Medline search were premature rupture of membranes, amniotic fluid, diagnosis, and tests. Articles not available online were retrieved by manual search. We found no Cochrane reviews of diagnostic methods of PROM, as discussed in this report. The studies identified showed evidence from levels II-2, II-3, and III, according to ranking of The Canadian Task Force on Preventive Health Care.21,22
MICROSCOPIC FETAL CELL IDENTIFICATION

A literature review by Friedman et al.23 described the first microscopic technique for identification of fetal particles in amniotic fluid, developed by Philipp et al in 1929 and published in the German literature. In this technique, fetal lanugo hairs were identified in amniotic fluid. This was presumed to be incontrovertible evidence of membrane rupture when identified in vaginal secretions. However, because of the

562

l JUNE JOGC JUIN 2010

Diagnosis of Premature Rupture of Membranes: Inspiration From the Past and Insights for the Future

Accuracy of diagnostic tests for PROM

Discriminatory concentration Mean (range) N/A 6.5 (6.07.0)35,36,55,57,59 N/A N/A N/A N/A N/A N/A N/A
23

Test used for diagnosis Fetal cells (bromthymol blue)34,36 pH 19,23,31,39,40,46,54,55,57,59,62 Fetal cells (Masson stains) 24 Fetal cells (Papanicolaou stain)23,28 Fetal cells (Pinacyanol stain)23,30,31 Fetal cells (Acridene orange stain)23,32 Fetal cells (Nile blue sulphate)23,33 Crystallization23,37,38-40,43,54,74 History alone23 History + pH + crystalization History + pH + Nile Blue stain23 History + Crystallization + Nile Blue stain23 History + ultrasound + crytallization74 pH + Crystallization + Nile Blue stain23 AmnioInjection Evans blue T182447 DAO19,58,67 3.32 log glucose + log fructose49

Sampling site Vagina Vagina Vagina Vagina Vagina Vagina Vagina Vagina N/A Vagina Vagina Vagina N/A Vagina Amniocentesis Vagina N/A

Patients, n Mean (range) 239 (176239) 125 (46250) 275 75 (50100) 150 (100250) 200 (100300) 106 (100111) 198 (51509) 100 100 100 100 83 100 18 168 (100272) 215

SN, % Mean (range) 90.2 (85.794.7) 90.2 (81.3100.0) 94.2 92.8 (92.0-93.5) 88.3 (71.098.7) 75.5 (61.289.7) 89.5 (80.798.2) 90.8 (62.0-98.5) 90.3 90.8 87.1 87.1 85.1 90.8 100.0 88.9 (83.0100.0) 98.9

SP, % Mean (range) 98.4 (97.499.3) 79.3 (16.0100.0) 99.0 92.7 (91.394.0) 95.4 (91.798.8) 90 (87.392.7) 98.6 (97.1100.0) 95.3 (88.2100.0) 88.4 95.6 92.7 95.6 78.6 95.6 100.0 98.3 (95.0100.0) 96.0

FN, % Mean (range) 7.5 (0.614.3) 12.4 (033.3) 5.8 4.3 (2.06.5) 11.7 (1.329) 24.6 (10.338.8) 10.6 (1.819.3) 4.4 (1.212.9) 9.7 9.2 12.9 12.9 11.1 9.2 0.0 8.5 (017.0) 7.7

FP, % Mean (range) 1.7 (0.72.6) 28.2 (0100.0) 1.0 7.4 (6.08.7) 3.7 (1.28.3) 10.0 (7.312.7) 1.5 (02.9) 4.7 (0.011.8) 11.6 4.4 7.3 4.4 21.4 4.4 0 1.7 (05.0) 2.6

N/A N/A N/A N/A N/A N/A 22.5 U/L19, 67 17.7279

continued

scant amounts of fetal lanugo hair in amniotic fluid and the fact that such hair was present in amniotic fluid only later in pregnancy, this method never gained popularity. As this study was not published in the English language, statistical data are not shown in the Table. Subsequent diagnostic tests were based on cytologic inspection for fetal squamous cells in the vagina. Interest in cytologic diagnosis was based on the noted absence of stain within the cytoplasm and nucleus of vernix caseosa cells compared with vaginal squamous cells.24 Investigators searched for fetal cells in vaginal fluid using diverse stains including Masson stain,24 Sudan III to demonstrate fetal fat particles,2527 Papanicolaou stain,28,29 pinacyanole stain,30,31 acridene orange stain,32 and Nile blue sulphate stain.33 Recognizing the challenge to diagnose PROM in clinically equivocal cases, Averette et al.30 studied 100 patients in

whom the integrity of the chorioamniotic membrane could not be determined by visualization of fluid on speculum examination. Fetal cell staining with pinacyanole to identify vernix caseosa cells in such cases showed high accuracy rates (97%), similar to those in earlier studies among patients with clinically confirmed membrane rupture. Although fetal cell staining techniques were considered rapid, simple, and durable, concerns about inaccuracies emerged. Contamination from the powder on examination gloves gave false positive results as anuclear granules mimicked fetal cells,31 and hypercornified vaginal cells also simulated anucleate fetal cells.32 False negatives resulted from the subjectivity of inspection, having insufficient cellular material,32 or after a prolonged time interval from membrane rupture. Accuracy rates appear to diminish with increased duration between presumed membrane rupture
JUNE JOGC JUIN 2010 l 563

OBSTETRICS

Table continued
Discriminatory concentration Mean (range) 3mU/mL,59 2ng/mL,51 20.2IU/mL52 53.3 g/L (2-125) 77.5 g/L (30-125) 31 ng/mL 50 ng/mL 25 ng/mL 35 g/L (3-100) 400 g/L 4.5 mmol/L
52,60,61,62

Test used for diagnosis Prolactin RIA50,51,52

Sampling site Vagina

Patients, n Mean (range) 57 (20100) 122 (52312) 89 (46131) 52 160 (54406) 46 130 (54174) 20 200 137 (100188) 151 114 (88139) 139 194 (184203) 20 69 (34103)

SN, % Mean (range)

SP, % Mean (range)

FN, % Mean (range) 68.0

FP, % Mean (range) 15.0 (030.0)51,52 8.1 (2.620.0), NR60 13.2 (12.014.3) NR 17.0 (6.028.0) 7.1 7.3 (1.828.6) NR 87 10.5 (5.016.0) 9.0 0.0 0 6.3 (0.012.5) NR 30.0 (25.035.0)

88.0 85.0 (76.0100.0)51,52 (70.0100.0)51,52 80.6 (17.6100.0), NR60 92.0 (84.0100.0) NR 93.6 (90.098.2) 87.5 84.9 (74.4100.0) NR 86 83.9 (68.0100.0) 94.0 100.0 100.0 98.8 (98.798.9) NR 99.2 (98.3100.0) 91.9 (80.097.4), NR60 86.9 (85.788.0) NR 65.7 (26.897.0) 92.9 92.8 (71.498.2) NR 92 89.5 (84.295.0) 91.0 100.0 100.0 93.8 (87.5100.0) NR 70.0 (65.075.0)

AFP (RIA)60,67,76,77,80

Vagina

5.4 (0.016.0), NR60 8.0 (016.0) NR 13.1 (3.025.0) 11.9 14.7 (039.2) NR 92 9.1 (0.025.0) 6.2 0.0 0.0 1.2 (1.11.3) NR 0.85 (01.7)

AFP (RIA)57,58,59 human placental lactogen (hPL) RIA51 fFN monoclonal assay (ROM check)58,59,63,64,65 fFN monoclonal assay (ROM check)59 IGFBP-1 immunoenzyme (PromTest)19, 46, 65,71,72,73,74 IGFBP-1 dipstick (ActimPROM)10 Lactate68 b hCG (ELCIA)

Cervix Vagina Vagina Vagina Vagina, cervix72 Amniocentesis Vagina Vagina N/A Vagina Vagina Vagina Amniocentesis N/A

46.4 mIU/mL < 80 mm 0.36 mg/dL 12 mg/dL 5 ng/mL 5 ng/mL pH> 5.2

Ultrasound amniotic fluid index46 Creatinine concentration assay69,70 Urea concentration70 PAMG-1 (AmniSure)76,77 PAMG-1 (AmniSure)10 AmnioSense absorbent pad15,78

Legend: SN: sensitivity; SP: specificity; FN: false negative; FP: false positive; DAO: diamine oxidase; IGFBP-1: insulin growth factor binding protein-1; PAMG-1: placental alpha microglobulin-1; bhCG: beta subunit human chorionic gonadotropin; fFN: fetal fibronectin; AFP: alpha fetoprotein; RIA: radioimmunoassay; NR: not reported; N/A: not applicable

and diagnostic testing. Early authors including King34 and Bourgeois24 reported 57% and 81.8% accuracy rates respectively, beyond the initial 24-hour period of membrane rupture. This was later confirmed by others, and shorter intervals to testing were shown to affect rates of diagnostic accuracy.23 As novel methods were developed, older cytologic techniques began losing popularity, because they were time-consuming, required trained cytologists, were ineffective before 32 weeks gestation, and provided an uncertain diagnosis of membrane rupture.29
564 l JUNE JOGC JUIN 2010

LITMUS PAPER AND pH TESTING

Litmus paper testing began development at approximately the same time as fetal cell staining methods. Because of the difference in pH of vaginal secretions (4.5 to 5.5) and amniotic fluid (7.0 to 7.5), it was rightly assumed that the pH of vaginal secretions would rise when contaminated by escaping amniotic fluid.35 This assumption prompted preliminary experiments with bromthymol blue dye and, later, nitrazine applicators. Although similar in principle to bromthymol blue, a reported advantage of nitrazine is the complete change in colour of the applicator when exposed to amniotic fluid in vaginal secretions.36 Impregnated with

Diagnosis of Premature Rupture of Membranes: Inspiration From the Past and Insights for the Future

sodium-dinitro-phenylozonapthol-disulphonate, nitrazine paper showed promising early results in detecting membrane rupture, with accuracies of 100%35 and 98.9%36 in clinically ruptured cases, and similarly high rates of accuracy among intact cases. This preliminary high optimism regarding nitrazine testing decreased by the 1960s,31,37 because of false positive results from vaginal infections, blood, semen, alkaline urine or alkaline antiseptics, and false negatives in cases of minimal leakage from chronic membrane rupture or high leak of the membranes.15
AMNIOTIC FLUID CRYSTALLIZATION

Indeed, these authors showed a precipitous increase in error rate when the time interval between membrane rupture and sampling exceeded two hours.23 False negative tests were more commonly associated with prolonged rupture. None of the traditional procedures proved entirely satisfactory in isolation, but combinations of any three of the following produced a diagnostic accuracy of 93.1%23: 1. a positive history, 2. a positive nitrazine test, 3. fluid crystallization, or 4. Nile blue sulphate staining.
ULTRASOUND ASSESSMENT OF AMNIOTIC FLUID VOLUME

Amniotic fluid crystallization, created primarily by the sodium chloride and protein content, began to dominate cytologic stains, with reported accuracies in clinically ruptured cases ranging from 73.0% to 98.5%.26,27,32,3740 Screening for arborization provided an accuracy of 97.8% compared with 87.3% for litmus paper testing.37 Interestingly, Tricomi et al.37 suggested that fluid for examination should be aspirated from the vagina no further than 3 cm from the introitus to avoid contamination and false positive ferning results from cervical mucus lying in the posterior fornix. Initial optimism with low false negative rates changed as missed diagnoses in the presence of blood, meconium, or heavy leukorrhea became recognized.23 False positive results were subsequently attributed to fingerprints, semen, or cervical mucus.23,41,42 In most ferning specimens tested in early studies,23,26,3740 sensitivity and specificity analyses were performed on women in labour, in whom there was certainty about the diagnosis of membrane rupture.43 In most clinical situations, however, the test is intended for patients who have equivocal rupture. De Haan et al.43 showed that the modest diagnostic reliability of the fern test compared with earlier published data is at least partly due to differences in study populations (i.e., whether the women were in labour or not). These authors showed false positive and false negative rates of 11.8% and 2.0%, respectively, in labouring women tested for amniotic fluid crystallization, compared with rates of 21.2.% and 40.6%, respectively, in women not in labour.43 Thus, the result of the fern test became viewed as supportive rather than conclusive for non-labouring women with non-specific vaginal fluid loss.
COMBINATION OF DIAGNOSTIC TESTS

In the late 20th century, it was theorized that the sonographic identification of oligohydramnios would develop after membrane rupture,44 thereby facilitating diagnosis and subsequent management. Manning et al.45 initially described a technique for measuring by ultrasound the deepest vertical pool of amniotic fluid in patients with intrauterine growth restriction. This method was later used to assess membrane rupture; it was shown that ultrasound quantification of the deepest amniotic fluid pocket is of poor quality in confirming membrane rupture.44 No significant difference was found in the mean depth of amniotic fluid pocket between 100 patients with confirmed term PROM and 51 patients with intact membranes.44 Oligohydramnios may not be detected in patients with confirmed PROM, possibly because drainage may become intermittent or even stop once the presenting part descends and acts as a plug, preventing further drainage.44 Robson et al. suggested that a significant amount of amniotic fluid needs to drain rapidly and continuously for oligohydramnios to occur, especially because the fluid is replaced to varying degrees by the fetus.44 Erdemoglu et al.46 showed that a reduction in the four-quadrant AFI below 80 mm did not reliably identify cases of suspected membrane rupture by history with negative visualization of fluid by speculum examination. The measurement of AFI offers no advantage over measurement of a single vertical pocket of fluid in cases where ultrasound is used to evaluate possible membrane rupture.
INTRA-AMNIOTIC DYE INJECTION

Many early studies either tested one method for diagnosis of membrane rupture or compared two methods. Friedman et al.23 were the first to test individual methods as well as different combinations of tests, which more commonly reflects true clinical practice. They also attempted to correlate the time of testing with the time of presumed rupture.

By 1970, amniocentesis with injection of dye to confirm amniotic membrane rupture had become a commonplace procedure; it was thought to be safe and had high patient acceptability rates.47 Prior to amniocentesis, intravenous injection of radioisotope was performed for placental localization, and the amnio-injection was performed under local
JUNE JOGC JUIN 2010 l 565

OBSTETRICS

anaesthesia. Interestingly, the two reported disadvantages of the procedure at that time were related to difficulty in diagnosis in the presence of meconium-stained fluid and the possibility of neonatal skin staining for 48 hours after dye injection.47 Several types of stains have been used for amnio-injections, with safety hazards reported only for methylene blue (see below). Although ultrasonographically guided transabdominal instillation of indigo carmine dye (1 mL of dye in 9 mL of sterile normal saline) and observation for fluid passage transvaginally is designated an unequivocal diagnostic method for confirmation of membrane rupture,14 this invasive test carries increased maternal and fetal risk. Inherent risks of intra-amniotic dye injection include trauma, bleeding, infection, and preterm labour.48 While strengthening diagnostic certainty, a negative dye test may occur if the membranes seal after previous amniotic fluid leakage.
GLUCOSE AND FRUCTOSE MEASUREMENTS

testing complexity, and low test sensitivities in cases of equivocal rupture.

PLACENTAL ALPHA-MICROGLOBULIN-1

Gorodeski et al.49 examined the difference between amniotic fluid and cervicovaginal mucus in their concentrations of solutes. During pregnancy, glucose and fructose are present in high concentrations in cervical mucus, with reported means of 240 mg/100 mL and 30.4 mg/100 mL, respectively.49 Amniotic fluid concentrations are noticeably lower, with mean concentrations of 39 mg/100 mL and 3.3mg/100 mL, respectively. Gorodeski et al.49 showed that low values of glucose and fructose are found in amniotic fluid aspirated in a true case of membrane rupture. Best results are obtained with a calculated linear sum of the values (3.32 log glucose + log fructose). These authors reported no confounding by meconium or vaginal discharge. This technique is fairly impractical in testing for membrane rupture, and further studies have not expanded upon differential carbohydrate gradients.
MODERN METHODS

Initially isolated in Moscow in 1975,75 PAMG-1 has undergone recent evaluation for diagnostic testing in PPROM. This 34kDa placental glycoprotein is abundant in amniotic fluid (200025 000 ng/mL), with much lower concentrations in maternal blood (525 ng/mL).20 The protein is present in negligible amounts in cervicovaginal secretions with intact membranes (0.050.2ng/mL).76 The 1000- to 10 000-fold difference in concentration between amniotic fluid and cervicovaginal secretions stimulated interest in a PAMG-1 immunoassay. Marketed as AmniSure (AmniSure International, Cambridge, MA), the assays minimum detection threshold for PAMG-1 is 5 ng/mL, sufficient for 99% accuracy with minimal false negatives. PAMG-1 can be detected with as little as 0.25 L of amniotic fluid in 1 mL of vaginal secretions.76 In the presence of blood or vaginitis, the background level of PAMG-1 can occasionally reach a maximum of 3 ng/mL.76 False-positive results with use of the AmniSure assay seem very unlikely, although these may appear with increased use.77 Further, assay of PAMG-1 appears to be reliable over a wide range of gestational ages (11 to 42 weeks), and proved superior to conventional combined clinical tests involving visualization of fluid pooling in the posterior fornix, arborization, and nitrazine testing.77 This test is currently available in Europe and was recently approved by the Food and Drug Administration for use in the United States. AmniSure is a novel rapid, non-invasive bedside test that may be very helpful in diagnosis of difficult cases without visible leakage. Further studies are needed to assess the reliability of the test according to the time from membrane rupture.
NON-INVASIVE ABSORBENT PAD

Because of the limitations of available testing methods, investigators have sought alternative markers in vaginal amniotic fluid, such as prolactin5052, alpha-fetoprotein,5159 b-subunit of human chorionic gonadotropin,52,6062 fetal fibronectin,58,59,6366 diamine oxidase,19,58,67 lactate,68 creatinine,69,70 urea,70 and insulin growth factor binding protein-1, previously called placental protein 12.10,19,46,65,66,7174 Interest in assessment of these markers stems from their high concentrations in amniotic fluid compared with normal vaginal secretions, but all require special laboratory equipment and training.14 Although these markers are useful for patients with intact membranes or unequivocal membrane rupture, they remain unpopular due to cost,
566 l JUNE JOGC JUIN 2010

Efforts to be able to confirm chorioamniotic membrane rupture with minute amounts of amniotic fluid have recently led to the development of the absorbent pad, AmnioSense. This 12 4 cm pad has a central strip that changes colour on contact with fluid with a pH > 5.2.15,78 After contact with urine, the strip reverts to its original colour when dry. This is due to the detachment of conjugate-based nitrazine molecules by the urine ammonium ions.15 AmnioSense has undergone cytotoxicity and skin irritation and sensitization testing, and it complies with the US Pharmacopoeia Guidelines.15 In a study of 34 women presenting with suspected membrane rupture, the AmnioSense pad initially showed 100% sensitivity; overall specificity was 75%, but when women with bacterial

Diagnosis of Premature Rupture of Membranes: Inspiration From the Past and Insights for the Future

vaginosis or Trichomonas vaginalis were excluded from analysis, the specificity increased to 90%.15 In a recent study, Mulhair et al.78 compared the reliability of the absorbent pad test with a standard of amniotic fluid pooling in the posterior fornix on speculum examination in a cohort of 139 women. They found a specificity of 65.0% and a sensitivity of 98.3% for the AmnioSense pad. The two studies of the absorbent pad currently available15,78 suggest that a negative AmnioSense result indicates intact membranes in term and preterm gestations in 99% of cases. A positive result, however, suggests only a 70% chance of ruptured membranes, and thereby warrants confirmation or further investigations to identify infections (Table).15,78 It remains unknown whether potential confounding substances such as semen, blood, or meconium may be distinguished from amniotic fluid by the AmnioSense pad test. As women with negative pad checks are unlikely to have ruptured membranes, this would imply decreased need for an uncomfortable and intrusive speculum examination.
SAFETY

CONCLUSION

Early techniques for establishing the diagnosis of ruptured membranes included fetal cell staining, vaginal pH determination and amniotic fluid crystallization. Criticism of such methods emphasized subjectivity in interpretation, poor sensitivity, and numerous confounding factors cross-reacting with substances in the vaginal reservoir. The absence of a non-invasive gold standard for the diagnosis of PROM has led to technically advanced biochemical markers. Despite improved diagnostic value in cases of unequivocal membrane rupture or intactness, biochemical markers lack the sensitivity and specificity required in equivocal cases. They have also not become popular, in part, because of cost and complexity in testing. In the common clinical situation where a health care provider encounters a patient with possible ruptured membranes, diagnostic accuracy is the key to successful management and improved perinatal outcome. Nearly one quarter of patients do not present with overt clinical evidence of ruptured membranes. We believe that recent techniques involving placental markers such as PAMG-1 and the AmnioSense absorbent pad, with minute sampling required, non-invasiveness, rapidity of testing, and high accuracy rates, may provide a solution to the clinical challenge of diagnosing ruptured membranes.
REFERENCES
1. Lavery JP, Miller CE, Knight RD. The effect of labor on the rheologic response of chorioamniotic membranes. Obstet Gynecol 1982;60:8792. 2. McLaren J, Taylor JD, Bell SC. Increased incidence of apoptosis in non-labor affected cytotrophoblast cells in term fetal membranes overlying the cervix. Hum Reprod 1999;14:895900. 3. El Khwad M, Stetzer B, Moore RM, Kumar D, Mercer B, Arikat S, et al. Term human fetal membranes have a weak zone overlying the lower uterine pole and cervix before onset of labor. Biol Reprod 2005;72:7206. 4. Moore RM, Mansour JM, Redline RW, Mercer BM, Moore JJ. The physiology of fetal membrane rupture: insight gained from the determination of physical properties. Placenta 2006;27:103751. 5. Al Zaid NS, Bou-Resali MN, Goldspink G. Bursting pressure and collagen content of fetal membranes and their relation to premature rupture of the membranes. Br J Obstet Gynecol 1980;87:2279. 6. Parry S, Strauss JF. Premature rupture of the fetal membranes. N Engl J Med 1998; 338:66370. 7. American College of Obstetrics and Gynecology Practice Bulletin No. 80. Premature rupture of membranes. April 2007. 8. Duff P. Premature rupture of the membranes in term patients. Semin Perinatol 1996;20:4018. 9. Hannah ME, Ohlsson A, Farine D, Hewson SA, Hodnett ED, Myhr TL, et al. Induction of labour compared with expectant management for prelabor rupture of the membranes at termTERMPROM Study Group. N Engl J Med 1996; 334:100510. 10. Chen FCK, Dudenhausen JW. Comparison of two rapid strip tests based on IGFBP-1 and PAMG-1 for the detection of amniotic fluid. Am J Perinatol 2008;25:2436.

Safety data on diagnostic testing for rupture of chorioamniotic membranes show potential adverse neonatal effects related to intra-amniotic administration of methylene blue.79,80 Several cases of neonatal hemolytic anemia and hyperbilirubinemia, including sepsis and neonatal death, have been reported after intra-amniotic injection of methylene blue.80 These cases involved injection of 3.5 mL of 1% methylene blue solution, resulting in doses of 30 to 5mg.80 Even at a lower dose of 10 mg, Cowett et al.79 reported a case of hemolytic anemia and hyperbilirubinemia requiring two exchange transfusions and phototherapy. These adverse effects could not be attributed to other causes of neonatal jaundice. The neonates skin was discoloured for the first few days of life and methylene blue was identified in the urine during the first day of life. Intra-amniotic dye as a marker for confirmation of PROM also poses risk to the mother, including bleeding, infection, and preterm labour.48 It is also an invasive test used to assess a common complaint. We were unable to identify any safety data related to the use of vaginal swabs, either with or without use of a speculum, in screening for chorioamniotic membrane rupture. Potential risks of swabs might include ascending infection. A non-intrusive method, the AmnioSense absorbent pad, may conceivably lead to negligible rates of infections and skin irritation. Neither of the two published studies testing absorbent pads15,78 report any complications from use of the product.

JUNE JOGC JUIN 2010 l

567

OBSTETRICS

11. Myles TD, Espinoza R, Meyer W, Bieniarz A. Preterm premature rupture of membranes: comparison between twin and singleton gestations. J Matern Fetal Neonat Med 1997;6:15963. 12. Park JS, Yoon BH, Romero R, Moon JB, Oh SY, Kim JC, et al. The relationship between oligohydramnios and the onset of preterm labor in preterm premature rupture of membranes. Am J Obstet Gynecol 2001;184:45962. 13. Buhimschi CS, Buhimschi IA, Norwitz ER, Sfakianaki AK, Hamar B, Copel JA, et al. Sonographic myometrial thickness predicts the latency interval of women with preterm premature rupture of the membranes and oligohydramnios. Am J Obstet Gynecol 2005;193:76270. 14. Mercer BM. Preterm rupture of the membranes: diagnosis and management. Clin Perinatol 2004;31:76582. 15. Bornstein J, Geva A, Solt I, Fait V, Schoenfeld A, Shoham HK, et al. Nonintrusive diagnosis of premature ruptured amniotic membranes using a novel polymer. Am J Perinatol 2006;23:3514. 16. Brace RA. Physiology of amniotic fluid volume regulation. Clin Obstet Gynecol 1997;40:2809. 17. Ladfors L, Mattsson LA, Eriksson M, Fall O. Is a speculum examination sufficient for excluding the diagnosis of ruptured fetal membranes? Acta Obstet Gynecol Scand 1997;76:73942. 18. Schuman W. Double sac with secondary rupture of the bag of waters during labor; a clinical entity, and its explanation from examination of the membranes. Am J Obstet Gynecol 1951;62:6338. 19. Gaucherand P, Salle B, Sergeant P, Guibaud S, Brun J, Bizollon CA, et al. Comparative study of three vaginal markers of the premature rupture of membranesinsulin like growth factor binding protein 1, diamine-oxidase, pH. Acta Obstet Gynecol Scand 1997;76:53640. 20. Park JS, Lee SE, Norwitz ER. Non-invasive testing for rupture of the fetal membranes. Touch Briefings: US Obstetrics and Gynecology 2007; Issue 1: 136. 21. Woolf SH, Battista RN, Anderson GM, Logan AG, Wang E. Assessing the clinical effectiveness of preventative maneuvers: analytic principles and systemic methods in reviewing evidence and developing clinical practice recommendations. A report by the Canadian Task Force on the Periodic Health Examination. J Clin Epidemiol 1990;43: 891905. 22. Canadian Task Force on the Preventive Health Care. New grades for recommendations from the Canadian Task Force on Preventive Health Care. CMAJ 2003;169:2078. 23. Friedman ML, McElin TW. Diagnosis of ruptured fetal membranes. Clinical study and review of the literature. Am J Obstet Gynecol 1969;104:54450. 24. Bourgeois GA. The identification of fetal squames and the diagnosis of ruptured membranes by vaginal smear. Am J Obstet Gynecol 1942;44:807. 25. Von Numers C. A new method of diagnosis of rupture of the membranes. Acta Obstet Gynecol Scand 1936;16:24960. 26. Paavola A. Methods based on the study of crystals and fat staining: use in diagnosing rupture of membranes. Ann Chir Gynaecol Fenn 1958;47:228. 27. Volet B, Morier-Genoud J. The crystallization test in amniotic fluid. Gynaecologia 1960;149:15161. 28. Goldfine S. The detection of ruptured membranes by vaginal smear. Am J Obstet Gynecol 1955;70:10914. 29. Hopman BC, Wargo JD, Werch SC. Cytology of vernix caseosa cells. Obstet Gynecol 1957;10:6569. 30. Averette HE, Hopman BC, Ferguson JH. Cytodiagnosis of ruptured fetal membranes. Am J Obstet Gynecol 1963;87:22630. 31. Gall SA, Spellacy WN. Cyologic diagnosis of ruptured membranes. Obstet Gynecol 1964;24:7325.

32. Kushner DH, Chang IW, Vercruysse JM. Fluorescence microscopy for determination of ruptured fetal membranes by vaginal smear. Obstet Gynecol 1964;23:1969. 33. Brosens I, Gordon H. The cytological diagnosis of ruptured membranes using Nile Blue Sulphate staining. J Obstet Gynaecol Br Commonw 1965;72:3426. 34. King AG. The determination of rupture of membranes. Am J Obstet Gynecol 1935;30:8602. 35. Baptisti A. Chemical test for determination of ruptured membranes. Am J Obstet Gynecol 1938;35:68890. 36. Abe T. Detection of rupture of fetal membranes with Nitrazine indicator. Am J Obstet Gynecol 1940;39:4004. 37. Tricomi V, Hall JE, Bittar A, Chambers D. Arborization test for the detection of ruptured fetal membranesclinical evaluation. Obstet Gynecol 1966;27:2759. 38. Smith RW, Callagen DA. Amniotic fluid crystallization test for ruptured membranes. Obstet Gynecol 1962;20:65560. 39. Kovacs D. Crystallization test for the diagnosis of ruptured membranes. Am J Obstet Gynecol 1962;83:1257. 40. Ferron M, Bilodeau R. Amniotic fluid crystallization test for ruptured membranes. Can Med Assoc J 1963;89:10647. 41. Lodeiro JG, Hsieh KA, Byers JH, Feinstein SJ. The fingerprint, a false-positive fern test. Obstet Gynecol 1989;73:8734. 42. McGregor JA. Johnson S. Fig-leaf ferning and positive nitrazine testing: semen, a cause of misdiagnosis of premature rupture of membranes [letter to the editor]. Am J Obstet Gynecol 1985;151:11423. 43. de Haan HH, Offermans JPM, Smits F, Schouten HJA, Peeters LL. Value of the fern test to confirm or reject the diagnosis of ruptured membranes is modest in nonlaboring women presenting with non-specific vaginal fluid loss. Am J Perinatol 1994;11:4650. 44. Robson MS, Turner MJ, Stronge JM, OHerlihy CO. Is amniotic fluid quantitation of value in the diagnosis and conservative management of prelabour membrane rupture at term? Br J Obstet Gynaecol 1990;97:3248. 45. Manning FA, Hill FM, Platt LD. Qualitative amniotic fluid volume determination by ultrasound. Antepartum detection of intrauterine growth retardation. Am J Obstet Gynecol 1981;139:2548. 46. Erdemoglu E, Mungan T. Significance of detecting insulin-like growth factor binding protein-1 in cervicovaginal secretions: comparison with nitrazine test and amniotic fluid volume assessment. Acta Obstet Gynecol Scan 2004;83:6226. 47. Atlay RD, Sutherst JR. Premature rupture of the fetal membranes confirmed by intra-amniotic injection of dye (Evans blue T-1824). Am J Obstet Gynecol 1970;108:9934. 48. Gibbs RS, Blanco JD. Premature Rupture of the membranes. Obstet Gynecol 1982;60:6719. 49. Gorodeski I, Paz M, Insler V, Fishel J. Diagnosis of rupture of the fetal membranes by glucose and fructose measurements. Obstet Gynecol 1979;53:6116. 50. Koninckx PR, Trappeniers H, Van Assche FA. Prolactin concentration in vaginal fluid: a new method for diagnosing ruptured membranes. Br J Obstet Gynaecol 1981;88:60710. 51. Huber JF, Bischof P, Extermann P, Beguin F, Herrmann WL. Are vaginal fluid concentrations of prolactin, a-fetoprotein and human placental lactogen useful for diagnosing ruptured membranes? Br J Obstet Gynaecol 1983;90:11835. 52. Shahin M, Raslan H. Comparative study of three amniotic fluid markers in premature rupture of membranes: prolactin, beta subunit of human chorionic gonadotropin, and alpha-fetoprotein. Gynecol Obstet Invest 2007;63:1959.

568

l JUNE JOGC JUIN 2010

Diagnosis of Premature Rupture of Membranes: Inspiration From the Past and Insights for the Future

53. Rochelson BL, Richardson DA, Macri JN. Rapid assay-possible application in the diagnosis of premature rupture of the membranes. Obstet Gynecol 1983;62:4148. 54. Rochelson BL, Rodke G, White R, Bracero L, Baker DA. A rapid calorimetric AFP monoclonal antibody test for the diagnosis of preterm rupture of the membranes. Obstet Gynecol 1987;69:1635. 55. Hjertberg R, Belfrage P, Eneroth P. Latex agglutination test for a-fetoprotein in the diagnosis of premature rupture of the amniotic membranes (PROM). Acta Obstet Gynecol Scand 1987;66:4379. 56. Gaucherand P, Guibaud S, Rudigoz RC, Wong A. Diagnosis of premature rupture of the membranes by the identification of alpha-feto-protein in vaginal secretions. Acta Obstet Gynecol Scand 1994;73:4569. 57. Kishida T, Yamada H, Negishi H, Sagawa T, Makinoda S, Fujimoto S. Diagnosis of preterm rupture of the membranes using a newly developed AFP monoclonal antibody test kit. Eur J Obstet Gynecol Reprod Biol 1995;58:6772. 58. Gaucherand P. Guibaus S, Awada A, Rudigoz RC. Comparative study of three amniotic fluid markers in premature rupture of membranes: fetal fibronectin, alpha-fetoprotein, diamino-oxydase. Acta Obstet Gynecol Scand 1995;74:11821. 59. Kishida T, Yamada H, Negishi H, Sagawa T, Makinoda S, Fujimoto S. Diagnosis of premature rupture of the membranes in preterm patients, using an improved AFP kit: comparison with ROM-check and/or nitrazine test. Eur J Obstet Gynecol Reprod Biol 1996;69:7782. 60. Kim YH, Park YW, Kwon HS, Kwon JA, Kim BJ. Vaginal fluid b-human chorionic gonadotropin level in the diagnosis of premature rupture of membranes. Acta Obstet Gynecol Scand 2005;84:8025. 61. Anai T, Tanaka Y, Hirota Y, Miyakawa I.Vaginal fluid hCG levels for detecting premature rupture of membranes. Obstet Gynecol 1997;89:2614. 62. Esim, E, Turan C, Unal O, Dansuk R, Cengizglu B. Diagnosis of premature rupture of membranes by identification of b-HCG in vaginal washing fluid. Eur J Obstet Gynecol Reprod Biol 2003;107:3740. 63. Hellemans P, Verdonk P, Baekelandt M, Joostens M, Franx M. Gerris J. Preliminary results with the use of the ROM-check immunoassay in the early detection of rupture of the amniotic membranes. Eur J Obstet Gynecol Reprod Biol 1992;43:1739. 64. Eriksen NL, Parisi VM, Daoust S, Flamm B, Garite TJ, Cox SM. Fetal fibronectin: a method for detecting the presence of amniotic fluid. Obstet Gynecol 1992;80:4514. 65. Rutanen EM, Pekonen F, Karkkainen T. Measurement of insulin-like growth factor binding protein-1 in cervical/vaginal secretions: comparison with the ROM-check Membrane Immunoassay in the diagnosis of ruptured fetal membranes. Clinic Chimica Acta 1993;214:7381.

66. Guibourdenche J, Luton D, Andre E, Noel M, Pourquet D. Rapid detection of insulin-like growth factor-binding protein-1 and foetal fibronectin in cervico-vaginal secretions to diagnose premature membrane rupture. Ann Clin Biochem 1999;36:3889. 67. Elmfors B, Tryding N. The diagnosis of ruptured fetal membranes by measurement of the diamine oxidase (DAO) activity in vaginal fluid. J Obstet Gynaecol Br Commonw 1974;81:3612. 68. Wiberg-Itzel E, Cnattingius S, Nordstrom L. Lactate determination in vaginal fluids: a new method in the diagnosis of prelabour rupture of membranes. Br J Obstet Gynecol 2005;112:7548. 69. Gurbuz A, Karateke A, Kabaca C. Vaginal fluid creatinine in premature rupture of membranes. Int J Gynaecol Obstet 2004;85:2701. 70. Kafali H, Oksuzler. Vaginal fluid urea and creatinine in diagnosis of premature rupture of membranes. Arch Gynecol Obstet 2007;275:15760. 71. Lockwood CJ, Wein, R, Chien D, Ghidini A, Alvarez M, Berkowitz RL. Fetal membrane rupture is associated with the presence of insulin-like growth factor-binding protein-1 in vaginal secretions. Am J Obstet Gynecol 1994;171:14650. 72. Rutanen EM, Karkkainen TH, Lehtovirta J, Uotila JT, Hinkula MK, Hartikainem AL. Evaluation of a rapid strip test for insulin-like growth factor binding protein-1 in the diagnosis of ruptured fetal membranes. Clinic Chimica Acta 1996;253:91101. 73. Darj E, Lyrenas S. Insulin-like growth factor binding protein-1, a quick way to detect amniotic fluid. Acta Obstet Gynecol Scand 1998;77:2957. 74. Jeurgens-Borst AJCM, Bekkers RLM, Sporken JMJ, van den Berg PP. Use of insulin like growth factor binding protein-1 in the diagnosis of ruptured fetal membranes. Eur J Obstet Gynecol Reprod Biol 2002;102:114. 75. Petrunin DD, Griaznova IM, Petrunina IuA, Tatarinov IuS. Immunochemical identification of organ specific human placental alpha-microglubulin and its concentration in amniotic fluid [article in Russian]. Akush Ginekol (Mosk) 1977;1:645. 76. Cousins LM, Smok D, Lovett SM, Poeltler DM. AmniSure placental alpha microglobulin-1 rapid immunoassay versus standard diagnostic methods for detection of rupture of membranes. Am J Perinatol 2005;22:14. 77. Lee SE, Park JS, Norwitz ER, Kim KW, Park HS, Jun JK. Measurement of placental alpha-microglobulin-1 in cervicovaginal discharge to diagnosis rupture of membranes. Obstet Gynecol 2007;109:63440. 78. Mulhair L, Carter J, Poston L, Seed P, Briley A. Prospective cohort study investigating the reliability of the AmnioSens method for detection of spontaneous rupture of membranes. BJOG 2009;116:3138. 79. Cowett RM, Hakanson DO, Kocon RW, Oh W. Untoward neonatal effect of intraamniotic administration of methylene blue. Obstet Gynecol 1976;48(Suppl):74s-75s. 80. Plunkett GD. Neonatal complications [letter to the editor]. Obstet Gynecol 1973;41:4767.

JUNE JOGC JUIN 2010 l

569