Prelabor rupture of membranes before and at the limit of viability

INTRODUCTION — 

Prelabor rupture of the fetal membranes (PROM) before or at the lower limit of neonatal viability is associated with substantial serious pediatric morbidity and mortality. The lower limit of neonatal viability can be defined as the earliest stage of fetal maturity when there is a reasonable chance, although not a high likelihood, of extrauterine survival. Currently, this is considered to be approximately 22 to 23 weeks of gestation in facilities with a neonatal intensive care unit.

The etiology, diagnosis, and complications of PROM before or at the lower limit of neonatal viability will be reviewed here. The diagnosis and management of preterm and term PROM later are discussed separately.

●(See "Preterm prelabor rupture of membranes: Clinical manifestations and diagnosis".)

●(See "Preterm prelabor rupture of membranes: Management and outcome".)

●(See "Prelabor rupture of membranes at term: Management".)

INCIDENCE — 

PROM before or at the limit of neonatal viability complicates 0.1 to 0.7 percent of pregnancies [1-3].

CAUSES AND RISK FACTORS — 

PROM before or at the limit of neonatal viability may occur spontaneously or it may be related to an invasive procedure that breaches the fetal membranes.

Spontaneous — The pathogenesis of spontaneous PROM before or at the limit of neonatal viability is not well understood. Multiple etiologies are probably involved and operate along various pathways involving both mechanical and nonmechanical factors. It is likely these pathways lead to a final common pathway that culminates in membrane rupture. (See "Spontaneous preterm birth: Pathogenesis".)

Risk factors are similar to those for preterm labor (table 1). The major risk factors appear to be a prior history of preterm labor, PROM before or at the limit of neonatal viability, cervical insufficiency, or a current pregnancy complicated by multiple gestation or antepartum bleeding. It is likely that both PROM before or at the limit of neonatal viability and preterm labor are expressions of a more general process of intrauterine inflammation and thus share many antecedents [4]. Since the fetal membranes are avascular, they may be particularly vulnerable to the local effects of inflammatory reactions.

After invasive procedures

●Amniocentesis – Diagnostic amniocentesis is the most common procedure associated with PROM before or at the limit of neonatal viability. Fluid leakage has been reported after approximately 1 percent of these procedures [5,6]. In contrast to spontaneous PROM, leakage of amniotic fluid shortly after amniocentesis is typically transient and associated with a good outcome, possibly because of "resealing" of the membranes after amniocentesis [7]. (See 'Resealing and reaccumulation of fluid' below and "Diagnostic amniocentesis", section on 'Leakage of amniotic fluid'.)

●Fetoscopy, fetal surgery, and percutaneous umbilical vein blood sampling – These procedures are more likely to rupture membranes than amniocentesis. The risk is highest with fetal surgery and fetoscopy. For fetoscopy, clinical experience suggests that the risk is related to the number and diameter of ports and the duration of manipulation [8]. Single port operative procedures are associated with a 6 to 10 percent incidence of membrane rupture; the risk is lower for diagnostic procedures, but much higher (20 to 30 percent [9]) for operative procedures involving two or more ports and long operating times.

●Cerclage – Membrane rupture can occur at or after physical examination-, ultrasound-, or history-indicated cerclage placement, and has a poor prognosis. However, cerclage placement is not necessarily causal as these patients have underlying risk factors (eg, subclinical chorioamnionitis) for preterm birth [10,11]. (See "Transvaginal cervical cerclage", section on 'Removal of cerclage after PPROM' and "Transvaginal cervical cerclage", section on 'Complications'.)

CLINICAL MANIFESTATIONS AND DIAGNOSIS — 

The clinical manifestations and diagnosis of ruptured membranes are the same across gestation and are discussed in detail separately. (See "Preterm prelabor rupture of membranes: Clinical manifestations and diagnosis", section on 'Diagnostic evaluation and diagnosis'.)

The diagnosis of PROM before or at the limit of neonatal viability is based upon a characteristic history (leaking fluid from the vagina) and physical examination (visualization of fluid flowing from the endocervical canal and pooling in the vagina), supplemented by ultrasound examination and standard laboratory tests in cases of diagnostic uncertainty.

COURSE

Resealing and reaccumulation of fluid — Up to 14 percent of pregnant patients with spontaneous PROM before or at the limit of neonatal viability eventually stop leaking amniotic fluid, presumably due to resealing of the fetal membranes [12-14]. Cessation of leakage is probably not due to actual repair and regeneration of the membranes, but rather to changes in the decidua and myometrium that block further leakage [15]. This subgroup of pregnancies has outcomes comparable to pregnancies uncomplicated by amniorrhexis [12].

A second subset of patients continue to leak fluid but have partial reaccumulation of amniotic fluid. Approximately 25 percent of patients reaccumulate fluid on ultrasound examination during expectant management [16,17].

As noted above, patients with iatrogenic PROM after amniocentesis usually reseal membranes within a week and have a good pregnancy outcome. The relatively discrete and focal nature of the iatrogenic membrane disruption when compared with the more ragged and larger defect that occurs with spontaneous membrane rupture likely contributes to these differences in resealing rates. (See "Diagnostic amniocentesis", section on 'Leakage of amniotic fluid'.)

Latency/preterm birth — The time interval between membrane rupture and delivery is a critical factor in determining outcome in PROM before or at the limit of neonatal viability. In the author's review of studies in which the weighted mean gestational age at PROM was 23.5 weeks of gestation [12,13,16-28], the weighted median latency from PROM before or at the limit of neonatal viability to delivery was 6.8 days, and the weighted mean latency was 11 days. The large difference between median and mean latency is because the majority of these pregnancies deliver soon after presentation, but a few of those that do not deliver may remain undelivered for several weeks. Approximately 50 percent of patients give birth in the first week after rupture; the mean latency for pregnancies not delivering within the first 24 hours is 10 to 14 days, and 84 percent of pregnancies have delivered by 28 days post-rupture.

Several investigators have suggested that there is an inverse relationship between latency and gestational age at rupture: Ruptures occurring earlier in gestation appear to have longer latencies [16,19,21,23-25,29]. This inverse relationship may represent "right censoring"; patients who deliver shortly after arriving on the labor floor are often excluded from latency studies and thus these studies may exclude patients more likely to be at an advanced gestational age.

In addition, there appears to be a correlation between the amount of residual amniotic fluid and the latency period; severe oligohydramnios is associated with the shortest average latency [30].

PREGNANCY COMPLICATIONS AND OUTCOMES — 

Pregnancy complications include preterm birth, maternal and/or fetal/neonatal infection, placental abruption, umbilical cord prolapse, fetal/neonatal deformation, fetal or neonatal death, retained placenta, and need for cesarean birth via a classical hysterotomy incision.

Counseling the patient with PROM before or at the limit of neonatal viability about pregnancy complications and pediatric outcome is challenging because of the small size, different gestational age ranges, and retrospective nature of the multiple studies on this subject. The majority of these reports refer to a single institutional experience, frequently prior to the widespread use of antepartum antibiotic prophylaxis, antepartum corticosteroids to accelerate fetal lung maturation, and neonatal surfactant therapy. In addition, these studies often have differing inclusion/exclusion and diagnostic criteria and limited ability to adjust for confounders. Despite these limitations, the body of data provides remarkably consistent conclusions and estimates of the magnitude of risk with expectant management.

Chorioamnionitis — Chorioamnionitis complicates approximately 30 to 40 percent of PROM cases before or at the limit of neonatal viability [17,21,22,24,31-33]. It is more common in these pregnancies than in those without PROM delivering at a similar gestational age [13,25,34], or in pregnancies with PROM occurring later in gestation [35,36].

Chorioamnionitis can be a cause or a consequence of PROM, especially before or at the limit of neonatal viability. It may cause PROM because infection of intact membranes weakens them, thereby facilitating rupture; it can be a sequelae of PROM because ruptured membranes are less effective as a barrier to transcervical migration of cervicovaginal flora. (See "Spontaneous preterm birth: Pathogenesis".)

While subclinical chorioamnionitis may be present prior to membrane rupture, it typically becomes clinically apparent early in the latency period; more than one-half of cases occur within the first seven days after rupture [12,18,19,22]. The maximum clinical occurrence is on the second through fifth day after rupture [12,37]. After the first week of latency, the incidence of clinical chorioamnionitis falls dramatically [12,19,21,22,38,39]. The higher frequency of clinical chorioamnionitis early in the latency period suggests that subclinical infection was likely present prior to membrane rupture and may have been the cause.

Clinically evident chorioamnionitis is associated with onset of preterm labor and birth, thus it is a major cause of shortened latency. The risk of developing chorioamnionitis after PROM before or at the limit of viability is increased in patients with very low residual amniotic fluid volumes [17] and those who undergo digital cervical examination [40]. The observation that the incidence of chorioamnionitis falls with prolonged latency implies that intrauterine infection may play a smaller role in precipitating delivery in the longer latency pregnancies than in pregnancies that deliver shortly after membrane rupture. This observation has been corroborated by a report that noted that the frequency of inflammatory pathology observed on neonatal head ultrasounds was lower after a longer latency than after a shorter latency [37].

Maternal sepsis and death — Chorioamnionitis can lead to maternal sepsis [41], although rare with appropriate intervention (eg, antibiotic therapy and delivery). Rates of maternal sepsis vary markedly depending on the study and inclusion criteria: In one study, maternal sepsis complicated approximately 5 percent (5/104) of pregnancies with PROM between 20 and 24 weeks [31] and in another study it complicated 1.2 percent (2/174) of pregnancies with PROM between 14 and 23 weeks [32]. However, in a prospective study of 364 pregnancies with PROM between 16+0 and 22+6 weeks, maternal sepsis developed in 10 percent of patients who chose to undergo termination of pregnancy after diagnosis of early PROM and in 13 percent of those who initially chose to continue their pregnancy [42]. Two patients died. The UK and Ireland Confidential Enquiries into Maternal Deaths and Morbidity have also reported maternal deaths associated with midtrimester PROM; most were due to sepsis [43]. A French confidential enquiry into maternal deaths concluded that the main preventable factors were delayed diagnosis, delayed fetal extraction, and inappropriate antibiotic treatment [44]. They estimated maternal mortality in PROM prior to viability was 4.5/10,000.

Abruption — Abruption is more common in pregnancies with PROM before or at the limit of neonatal viability than in the general obstetric population, 2 to 44 percent versus 0.4 to 1.3 percent [17]. Among pregnancies with PROM before or at the limit of neonatal viability, the risk of abruption correlates inversely with gestational age at rupture [24,45]. As an example, the incidence of abruption was 40 to 50 percent in PROM before 20 weeks in two studies [24,45] but 17 percent in a study of PROM at 20 to 24 weeks in third study [31]. (See "Acute placental abruption: Pathophysiology, clinical features, diagnosis, and consequences".)

Chronic abruption is likely one cause of PROM before or at the limit of neonatal viability. There is abundant evidence that thrombin enhances fetal membrane and/or decidual cytokine and/or protease production, which could exacerbate membrane damage [46-50]. Abruption-associated thrombin also appears to decrease expression of decidual cell progesterone receptor, which promotes preterm birth [51]. (See "Spontaneous preterm birth: Pathogenesis", section on '#3 Decidual hemorrhage'.)

Cord prolapse — A review of nine studies including 731 patients with PROM before or at the limit of neonatal viability reported a 1.9 percent incidence of cord prolapse [52]. This risk should be considered when monitoring these pregnancies as they reach and go beyond a viable gestational age, given the potential for an adverse outcome. (See "Umbilical cord prolapse".)

Fetal death — The reported incidence of fetal death varies depending on the inclusion of pregnancy terminations versus spontaneous fetal losses. In one series of 93 pregnancies with PROM at 15 to 24 weeks, 27 percent underwent a pregnancy termination; of the remaining pregnancies, fetal demise occurred in 30 percent [33]. In another cohort, 7 and 1 percent of pregnancies were terminated at 22 and 23 weeks, respectively; the risk of stillbirth was 8 percent at both 22 and 23 weeks but fell to 3 percent by 25 weeks [28]. The risk of fetal death appears to be inversely related to gestational age at PROM [22,26,28]. In one study, the prevalence of fetal death prior to 20 weeks, prior to 24 weeks, and after 25 weeks was 33, 20, and 0 percent, respectively.

Fetal death in PROM before or at the limit of neonatal viability is typically caused by abruption, cord prolapse, or infection [22,26]. Use of emergency cesarean birth to avoid fetal death when these complications occur is less prevalent at the limit of neonatal viability than later in gestation, given the poor pediatric prognosis.

Residual amniotic fluid seems to be protective against fetal death. All deaths in one series occurred in pregnancies in which the largest fluid pocket was <2 cm in greatest diameter [53].

In twin pregnancies with PROM of one sac at <26 weeks, a retrospective study observed a significantly higher rate of fetal death in the twin with ruptured membranes (6/23 [26.1 percent] versus 1/23 [4.4 percent]) [54]. Other outcomes were similar for both twins.

Cesarean birth — The cesarean birth rate is increased in pregnancies with PROM before or at the limit of neonatal viability, due to an increased prevalence of fetal heart rate abnormalities and fetal malpresentation. A classical hysterotomy will be required for fetal extraction since it is highly unlikely that the lower uterine segment will be sufficiently developed remote from term [22].

Retained placenta — Retained placenta necessitating manual removal is more common with these pregnancies than later in gestation (incidence: 9 to 20 percent), especially if membrane rupture occurs prior to 20 weeks of gestation [12,19,22,42]. (See "Retained placenta after vaginal birth".)

Postpartum endometritis — Postpartum endometritis occurs in up to 40 percent of patients with PROM before or at the limit of neonatal viability; the author of this topic calculated a weighted cumulative average risk of approximately 13 percent upon review of the literature. Postpartum maternal sepsis is less common (0 to 3 percent) [12,13,16-26,39]. The risk of maternal infection appears to be inversely proportional to the latency period [19]. (See "Postpartum endometritis".)

PEDIATRIC OUTCOMES

Morbidity — Short- and long-term morbidity are common and primarily related to gestational age at birth [1,2,7,42,55,56]. In a contemporary review, the most reliable prognostic factors were gestational age at PROM and latency duration [57]. This is another way of stating that increased gestational age at delivery is the dominant determinant of neonatal outcome [55]. Although the National Institute of Child Health and Human Development (NICHD) Neonatal Research Network provides an online calculator to estimate the risk of neurodevelopmental impairment of extremely preterm birth based on gestational age at birth, birth weight, sex, singleton birth, and exposure to antenatal corticosteroids within seven days, it does not account for the altered intrauterine environment associated with early PROM, which uniquely contributes to neonatal morbidity. For example, PROM before or at the limit of viability increases the risks of pulmonary hypoplasia due to oligohydramnios and infection-related morbidity (eg, sepsis, meningitis, pneumonia) due to bacterial contamination of the intraamniotic cavity from the vagina.

The prognosis for intact survival after PROM before or at the limit of neonatal viability is only fair. In a prospective study of 364 pregnancies with PROM between 16+0 and 23+6 weeks, survival to hospital discharge was 26 percent (54 of 207 expectantly managed pregnancies) and survival to discharge without severe morbidity was 18 percent (38 of 207 expectantly managed pregnancies); however, long-term data were not reported [42].

In one study with long-term data, early PROM was associated with greater composite severe childhood morbidity at age two years than later PROM, even after controlling for gestational age at delivery and other confounders [2]. In three studies of pregnancies complicated by PROM before or at the limit of neonatal viability, two reported that approximately 50 percent of offspring had no severe morbidity at two years of corrected age [31,58] and one reported 70 percent had normal neurodevelopment but more than half had respiratory problems at two and/or five years of age [59]. However, in another study, 90 and 64 percent of 22- and 23-week PROM survivors were diagnosed with some form of cerebral palsy [28]. Multiple gestation does not appear to affect outcome [42,54].

Prolonged latency alone is not associated with an increasing frequency of abnormal neonatal cranial ultrasound examination, and there is no evidence that planned early delivery will prevent adverse sequelae [37,60].

Specific complications of prematurity are reviewed separately:

●(See "Overview of short-term complications in preterm infants".)

●(See "Overview of the long-term complications of preterm birth".)

●(See "Periviable birth (limit of viability)".)

Neonatal death — The neonatal mortality rate after PROM before or at the limit of neonatal viability is high, primarily due to the short latency period, which results in an early gestational age at birth [12,13,16-26,28,29,31,38,39,61-66]. However, gestational age-specific mortality is similar to that for controls without PROM [25,39]. Neonatal death after discharge remains low, between 1 to 2 percent [28].

Since mean latency in these pregnancies is 11 days, PROM occurring at 23 to 24 weeks of gestation is more likely to be associated with neonatal survival than PROM at or before 22 weeks [14,17,20,22,26,67]. Neonatal survival improves significantly after 25 weeks of gestation [21,23,28,39]. In a review of three studies of PROM at 14 or 16 to 24 weeks, neonatal survival after conservatively managed PROM at <22 weeks was significantly lower than PROM at 22 to 24 weeks (8/57 [14.4 percent] versus 30/52 [57.7] percent) [1]. A study of pregnancies with PROM at 20 to 23 weeks and latency of at least seven days reported neonatal survival in 90 percent (52/58) [68]. However, these data likely overstate survival because of exclusion of patients who delivered soon after PROM or elected pregnancy termination for persistent fluid leakage, oligohydramnios, abnormal ultrasound findings, or other issues associated with a poor prognosis. A study of PROM between 20 and 24 weeks that did not exclude patients with latency less than seven days reported neonatal survival in 49 percent (51/104) [31].

The risk of neonatal mortality is also correlated with residual amniotic fluid volume [17,26,62,65]. In one study, for neonates delivering at the same gestational age, neonatal survival was 98 percent when the largest amniotic fluid pocket was ≥2 cm but only 31 percent when <2 cm. [17]. The survival rate for pregnancies associated with iatrogenic PROM before or at the limit of viability is severalfold higher than that for spontaneous PROM before or at the limit of viability [69]. This may be related to a lower rate of infection, higher residual amniotic fluid volume, higher rate of resealing, or other factors.

Pulmonary hypoplasia — Newborns with pulmonary hypoplasia have small lungs with relatively normal compliance, high pulmonary vascular resistance, reduced pulmonary blood flow, and some degree of systolic and diastolic cardiac dysfunction [70]. The prevalence of pulmonary hypoplasia in neonates of pregnancies complicated by PROM before or at the limit of neonatal viability is approximately 30 percent [33]. The mortality rate for these neonates is 70 to 90 percent [24,26,38,53,61].

The gestational age at the time of membrane rupture is the critical factor in the risk of subsequent neonatal pulmonary hypoplasia [18,26,27,29,38,71]. In one study of pregnancies less than 29 weeks of gestation, as the gestational age at occurrence of PROM increased, the odds of pulmonary hypoplasia decreased by 46 percent per week [61]. Several series have reported a low incidence (less than 1.4 percent) of pulmonary hypoplasia when PROM occurred after 26 weeks of gestation [24,27,29,38,61]. This gestational age corresponds to the end of the canalicular stage of lung development, after which the developing acinar structure becomes less sensitive to external perturbations (table 2) [72].

The degree of oligohydramnios is an additional risk factor for pulmonary hypoplasia; lower volumes of residual fluid confer the highest risk [26,53,62]. In one study, the incidence of pulmonary hypoplasia with severe, moderate, and no to mild oligohydramnios was 43, 19, and 7 percent, respectively [62].

It is less clear whether there is an association between the length of the latency period and risk of pulmonary hypoplasia. One study found the median latency period was longer in cases of pulmonary hypoplasia than in those without (31 and 7 days, respectively) [25]. However, this finding may represent a relationship between earlier gestational age at rupture and longer latency rather than an independent effect of the latency period itself. Others have not confirmed this association [53,61].

Lethal pulmonary hypoplasia is difficult to diagnosis antepartum. A variety of sonographic methods have been evaluated (eg, thoracic circumference, thoracic:abdominal ratio), but no method has proven sufficiently reliable for clinical decision-making [73].

Musculoskeletal development — Most limb development is in the embryonic period [74]. The appendicular skeleton is premolded in cartilage by the end of this period and ossification centers first appear between 12 and 14 weeks postmenstrual age. The majority of limb growth occurs in the second and third trimesters. Asymmetric intrauterine pressure and restriction in fetal movement associated with PROM before or at the limit of neonatal viability can lead to deformities of variable severity in previously normally formed extremities [75]. An increased likelihood of skeletal deformities has been observed among infants diagnosed with pulmonary hypoplasia [29,38,61], suggesting that the two disorders have a common etiology. (See "Congenital anomalies: Approach to evaluation".)

The frequency of skeletal deformities varies widely among studies; the mean incidence is 7 percent [12,13,16-26,29,38,39,61,62]. In contrast to pulmonary hypoplasia, the gestational age at previable PROM is not a significant determinant of the risk of skeletal abnormalities [29,38,61,62]. This difference is likely due to the progressive and continuous development of the axial skeleton in contrast with the phases of biochemical maturation observed in the lung. Thus, a severe insult of prolonged duration, regardless of gestational age, should be more predictive of skeletal abnormalities than pulmonary abnormalities.

This hypothesis is supported by several series that observed that the duration of latency and the severity of oligohydramnios independently increase the risk of skeletal abnormalities and act synergistically to raise the risk beyond the effect of either variable alone [29,38,61]. As an example, one study reported the risk of skeletal abnormality was twofold higher in pregnancies with PROM before or at the limit of neonatal viability complicated by severe oligohydramnios (defined as largest pocket less than 1 cm) than in matched pregnancies with normal or mildly reduced fluid (26 versus 54 percent); the risk was highest after 14 days of latency [62].

Limb deformations related to PROM before or at the limit of viability generally do not require surgical correction as they gradually resolve with postnatal growth and development [38].

APPROACH TO PREGNANCY MANAGEMENT — 

Management of PROM before or at the limit of neonatal viability differs from membrane rupture later in gestation because the risks of expectant management versus delivery are quite different remote from term. As discussed above, the limit of neonatal viability is defined as the stage of fetal maturity that ensures a reasonable chance of neonatal survival. Determining the limit of neonatal viability is desirable so that costly and painful interventions can be avoided if the neonate does not have a chance of survival. However, deciding upon a threshold of viability is challenging because it remains uncertain which extremely preterm newborn has a reasonable chance of survival. Institutional variability in survival at the limit of viability is due, at least in part, to varying levels of aggressive management.

Previable pregnancy at membrane rupture — In previable pregnancies, clinicians need to discuss the risks and benefits of expectant management versus pregnancy termination with the patient. Shared decision-making is paramount with a high degree of sensitivity to the patient's values and circumstances [76,77]. It is important to be clear that expectant management carries a higher risk of maternal morbidity. In a retrospective cohort study of patients with PROM at 14+0 to 23+6 weeks, composite maternal morbidity was nearly twofold higher among patients managed expectantly versus those actively delivered (60.2 versus 33.0 percent; adjusted relative risk of 1.91, 95%CI 1.35-2.73) [78]. (See 'Pregnancy complications and outcomes' above.)

For patients who choose active intervention, dilation and evacuation (D&E) is associated with fewer adverse complications than induction. In another retrospective cohort study of patients with PROM at 14+0 to 23+6 weeks, time to uterine evacuation was shorter with D&E and the rate of adverse events was lower (16.9 versus 52.2 percent) [79].

Most patients at this gestational age who are stable and choose to continue their pregnancies are not admitted to the hospital and not given antenatal corticosteroids or tocolytics. Antibiotic prophylaxis may be considered as early as 20 weeks of gestation [80], although the author of this topic waits until 23 weeks (see 'Pregnancies at the limit of viability at membrane rupture' below)

A rectovaginal group B Streptococcus (GBS) culture is obtained. Treatment for a positive GBS culture is initiated when the patient is in labor at a gestational age where neonatal resuscitation would be initiated, if necessary. (See "Preterm prelabor rupture of membranes: Management and outcome", section on 'Administer prophylactic antibiotic therapy' and "Prevention of early-onset group B streptococcal disease in neonates", section on 'Preterm prelabor rupture of membranes'.)

Pregnancies at the limit of viability at membrane rupture — Management of pregnancies at the limit of neonatal viability varies depending on patient-specific factors and preferences [76,81]. These patients are admitted to the hospital for usual management of PROM at the gestational age when they would want aggressive neonatal intervention. A common practice among neonatologists is to not offer or provide neonatal resuscitation to neonates <22+0 weeks of gestation due to the low chance of intact survival. Likewise, between 22+0 and 23 or 24 weeks, it is also a common practice among neonatologists to offer neonatal resuscitation to parents if there is at least a small chance of survival based on available information and provide or withhold resuscitation based on the preference of informed parents. (See "Periviable birth (limit of viability)".)

All patients are therefore offered admission in the 23^rd week, and the author of this topic offers a course of antenatal corticosteroid therapy in the mid-22^nd week. Administration of antenatal corticosteroids in the 22^nd week is reasonable if delivery in the 23^rd week is anticipated and the patient desires aggressive neonatal intervention after thorough consultation with maternal-fetal medicine and neonatology specialists. Some experts consider not giving antenatal steroids a "missed opportunity" in the care of patients intending neonatal resuscitation [82]. The author feels strongly that the decision to administer prophylactic corticosteroids at this margin of viability should be considered in the context of shared decision-making with the patient and family. Further, it should be independent of the intention to provide emergency cesarean birth. By the 23^rd week of gestation, he routinely provides emergency cesarean birth intervention for standard obstetric indications. Parents should be informed that antenatal corticosteroids may provide a survival benefit while increasing the risk of survival with severe impairment. Also, if the pregnancy is not delivered, then a single repeat course of antenatal corticosteroids may be needed later in gestation. Evidence of the benefits and harms of antenatal steroid administration is reviewed separately. (See "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery", section on '<22+0 weeks' and "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery", section on '22+0 to 22+6 weeks' and "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery", section on '23+0 to 33+6 weeks'.)

Antibiotic prophylaxis to prolong latency is administered in the 23^rd week at admission according to the same protocol used in patients who present with preterm PROM at later gestational ages [81]. Although a case series of 213 pregnancies with PROM <23 weeks suggested that antibiotic therapy was associated with longer latency and improved neonatal outcomes, the authors acknowledged that the improved survival may have been a consequence of increased antibiotic use at later, rather than earlier, gestational ages and other confounding factors [83].

The author of this topic initiates daily fetal testing after viability. While many of the risks that preterm PROM poses to fetal well-being cannot be predicted by testing, abnormal test results can be an early indicator of impending clinical chorioamnionitis, and the test is interpreted with this in mind.

Magnesium sulfate for neuroprotection is administered after viability when delivery appears to be imminent. Similarly, prophylaxis for GBS should not be administered unless a delivery is anticipated at or beyond viability. (See "Preterm prelabor rupture of membranes: Management and outcome".)

Pregnancies at a viable gestational age — Management of pregnancies with PROM at a gestational age when neonatal resuscitation is routinely provided is similar to that for preterm PROM later in gestation, and reviewed separately. (See "Preterm prelabor rupture of membranes: Clinical manifestations and diagnosis".)

MANAGEMENT OF PREGNANCY COMPLICATIONS

Life-threatening complications — Infection and abruption are potentially life-threatening maternal complications. The French National Confidential Enquiry into Maternal Deaths identified seven deaths between 2001 and 2015 associated with previable PROM: six deaths were attributed to sepsis after genital infection by Gram-negative bacilli and one to postpartum hemorrhage due to placenta accreta [44].

●Infection — Clinical chorioamnionitis is an indication for expeditious delivery because expectant management increases the risks for maternal morbidity and death, fetal death, and neonatal morbidity and death. (See "Clinical chorioamnionitis".)

Expectant management with antibiotic therapy is not appropriate because approximately 10 percent of patients with clinical chorioamnionitis managed expectantly will be diagnosed with sepsis, which can evolve rapidly despite administration of antibiotics [84]. (See "Septic abortion: Clinical presentation and management" and "Evaluation and management of suspected sepsis and septic shock in adults" and "Sepsis syndromes in adults: Epidemiology, definitions, clinical presentation, diagnosis, and prognosis".)

●Abruption — Acute placental abruption is a significant cause of maternal morbidity, which can be serious. Prompt intervention can reduce this risk. Management of abruption is reviewed separately. (See "Acute placental abruption: Pathophysiology, clinical features, diagnosis, and consequences" and "Acute placental abruption: Management and long-term prognosis".)

Coexistent cerclage — It is unclear whether presence of a cerclage increases the risk of maternal or neonatal sepsis in the setting of PROM before or at the limit of neonatal viability; the most appropriate management in this clinical setting remains controversial. The author of this topic leaves the cerclage in situ in the absence of signs of labor or intrauterine infection. Timing of cerclage removal is reviewed separately. (See "Transvaginal cervical cerclage", section on 'Removal of cerclage after PPROM'.)

Delayed interval delivery of multiple gestations — Multiple gestations with PROM before or at the limit of neonatal viability are managed in the same way as singletons. While some clinicians believe delayed-interval delivery is an option for pregnancies at an early gestational age (<24 weeks) in which only the first (presenting) fetus spontaneously delivers, this has not been the policy of the author of this topic, given the heightened maternal risks from intrauterine infection. The safety and efficacy of delayed-interval delivery is reviewed separately. (See "Multifetal gestation: Approach to delayed-interval delivery".)

UNPROVEN INTERVENTIONS

Prophylactic antibiotics — The utility of latency antibiotics in the previable period remains controversial and unclear. The author of this topic never administers prophylactic antibiotics prior to 20+0 weeks of gestation for the indication of prolonging latency as the potential addition of one week at this gestational age, on average, rarely has benefit in the overall context of fetal development. However, use of broad spectrum antibiotic prophylaxis is a common practice [1,85] based on data from observational studies [39,65,86] and extrapolation of results from randomized trials at later gestational ages.

Data from randomized trials show that antibiotic prophylaxis after membrane rupture before 34 weeks and before 37 weeks increases latency by approximately one week; however, no similar trials have been performed among patients with previable PROM to support either the efficacy or safety of this therapy at previable gestational ages. (See "Preterm prelabor rupture of membranes: Management and outcome", section on 'Administer prophylactic antibiotic therapy'.)

Amnioinfusion — The author of this topic does not perform amnioinfusion to prevent complications from oligohydramnios. The use of serial amnioinfusions is controversial, with only sparse data and no convincing evidence of benefit available in PROM before or at the limit of neonatal viability [87-89]. Some groups have, however, observed an increased latency period with amnioinfusion. One such study reported median latency with versus without amnioinfusion was 13 and 4 days, respectively [90] and the other reported median latency of 5.5 and 3 weeks, respectively, as well as improved neonatal morbidity [91]. More studies of the safety and efficacy of the procedure in this setting are needed before prophylactic amnioinfusion can be recommended. (See "Preterm prelabor rupture of membranes: Management and outcome", section on 'Amnioinfusion'.)

Repair of leaks — A variety of tissue sealants (eg, fibrin glue, gelatin sponge, amniopatch) have shown some success in case reports [15]. Neither the safety nor the efficacy of these sealants has been established [92]. Furthermore, thrombin may trigger fetal membrane matrix and/or decidual metalloproteinase and inflammatory cytokine production and decrease decidual cell progesterone receptor expression, which can increase myometrial contractions and promote further membrane damage and premature delivery [46,48,50,93].

Intra-amniotic injection of platelets and cryoprecipitate (ie, amniopatch) has also been used to treat iatrogenic rupture [94-97]. Although sometimes successful, the procedure has been associated with unexpected fetal death and should be considered investigational.

An evaluation of resealing technologies and protocols by the Cochrane group concluded that there is insufficient evidence to recommend any resealing protocol in contemporary clinical practice [92].

RECURRENCE — 

Patients with PROM before or at the limit of neonatal viability are at high risk of recurrence in subsequent pregnancies. An early prior spontaneous preterm birth appears to be more predictive of recurrence than late preterm delivery, and highly associated with a subsequent early spontaneous preterm birth [98]. In a retrospective cohort study of 108 patients with a history of one or more singleton pregnancies complicated by PROM <24 weeks of gestation, 46 percent (50/108) gave birth preterm in the pregnancy immediately after the index birth and 17 percent (18/108) gave birth before 24 weeks [99]. In another study of 99 patients with PROM <27 weeks, recurrent preterm birth occurred in 35 percent and 9 percent had recurrent PROM <27 weeks [100].

Given the increased risk of recurrence, modifiable risk factors such as cigarette smoking and short interpregnancy interval should be addressed. In consultation with their providers, enhanced surveillance in a future pregnancy, potentially in collaboration with a clinic specializing in preterm birth prevention, might be useful.

Patients with a history of cervical insufficiency in a prior pregnancy are probably at increased risk of previable PROM in a subsequent pregnancy, given the common etiology of both entities. (See "Cervical insufficiency".)

SOCIETY GUIDELINE LINKS — 

Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Prelabor rupture of membranes".)

INFORMATION FOR PATIENTS — 

UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Preterm prelabor rupture of membranes (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Epidemiology and etiology – Prelabor rupture of membranes (PROM) before or at the limit of neonatal viability complicates up to 0.7 percent of pregnancies. It may be spontaneous or a complication of an invasive obstetric procedure. PROM after amniocentesis has a much better prognosis than spontaneous PROM because membranes often reseal after amniocentesis. (See 'Introduction' above and 'Causes and risk factors' above.)

●Mean latency – The time interval between membrane rupture and delivery is a critical factor in determining outcome. The mean latency from gestational age at PROM to delivery is approximately 17 days and the median latency is approximately one week, since the majority of pregnancies deliver soon after membrane rupture. (See 'Latency/preterm birth' above.)

●Complications – Pregnancy complications include increased risks of infection (chorioamnionitis, endometritis), abruption, cord prolapse, fetal death, preterm birth, and need for cesarean birth. Maternal sepsis occurs in a small proportion of patients with chorioamnionitis, rarely resulting in maternal death. (See 'Pregnancy complications and outcomes' above.)

●Resealing – Cessation of amniotic fluid leakage with reaccumulation of amniotic fluid confers a prognosis comparable to that of pregnancies without previable PROM. (See 'Resealing and reaccumulation of fluid' above.)

The safety and efficacy of tissue sealants for repair of PROM have not been established. Procedures to aid in the reaccumulation of amniotic fluid should be considered experimental. (See 'Repair of leaks' above.)

●Management

•We recommend antenatal corticosteroids at ≥23+0 weeks of gestation (Grade 1A). A course of antenatal corticosteroids in the 22^nd week is reasonable if delivery in the 23^rd week is anticipated and neonatal resuscitation is desired. (See "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery".)

•Infection and abruption are potentially life-threatening maternal complications that require expeditious delivery. (See 'Life-threatening complications' above.)

•Management of PROM before or at the limit of neonatal viability differs from PROM later in pregnancy.

-Previable pregnancy at membrane rupture – The risks and benefits of expectant management versus pregnancy termination are discussed. Most patients at this gestational age who are stable and choose to continue their pregnancies are not admitted to the hospital and not given antenatal corticosteroids, tocolytics, or prophylactic antibiotics, but we obtain a rectovaginal group B streptococcus culture. They are admitted to the hospital for usual management of PROM at the gestational age when they would want aggressive neonatal intervention. (See 'Previable pregnancy at membrane rupture' above.)

-Pregnancies at the limit of viability at membrane rupture – Management of these pregnancies depends on patient-specific factors and preferences. These patients are admitted to the hospital for usual management of PROM at the gestational age when they would want aggressive neonatal intervention. Between 22 and 23 or 24 weeks, a common practice among neonatologists is to offer neonatal resuscitation to parents if there is at least a small chance of survival based on available information and provide or withhold resuscitation based on the preference of informed parents.

We routinely administer a full course of antenatal corticosteroids to pregnancies at ≥23 weeks of gestation and offer corticosteroid therapy in the mid-22^nd week. By the 23^rd week of gestation, we routinely provide emergency cesarean birth for standard obstetric indications.

Parents should be informed that antenatal corticosteroids may provide a survival benefit while increasing the risk of survival with severe impairment. Also, if the pregnancy is not delivered, then a single repeat course of antenatal corticosteroids may be needed later in gestation. (See 'Pregnancies at the limit of viability at membrane rupture' above.)

-Pregnancies at a viable gestational age – Our management of pregnancies with preterm PROM at a gestational age when neonatal resuscitation is routinely provided is similar to that for preterm PROM later in gestation. (See "Preterm prelabor rupture of membranes: Clinical manifestations and diagnosis".)

●Neonatal outcome

•Neonatal survival is primarily related to gestational age at delivery, and is comparable to that in preterm deliveries matched for gestational age without PROM before or at the limit of neonatal viability. For these very early gestations, gains in fetal maturity with expectant management dramatically improve survival. (See 'Neonatal death' above.)

•The neonate is at risk for morbidity/mortality related to preterm birth, infection, pulmonary hypoplasia, and musculoskeletal deformation. (See 'Pediatric outcomes' above.)