Although the classic term “premature rupture of membranes” has been recently modified to “prelabor rupture of membranes”, the abbreviation (PROM) as well as the definition of the condition have not been changed and are satisfactorily consistent worldwide. The new term “prelabor rupture of membranes” has been adopted by the American College of Obstetricians and Gynecologists (ACOG) in 2018 and even earlier by the National Institute of Clinical Excellence (NICE) [1,2]. We believe that the new term is more accurate and should be widely used instead of the old term. Here, the old term “premature” may be misinterpreted as “fetal prematurity” and the use of both “preterm” and “premature” to define the condition is confusing. Therefore, we highly recommend the use of the new term instead of the old one. Prelabor rupture of membranes is classified according to the gestational age at which it occurs into; term, preterm and previable:

According to Websters Encyclopedic Unabridged Dictionary, the word “viability” refers to the stage of fetal development at which a fetus is capable of living outside the uterus [5]. Clinically, the age of viability is defined as the gestational age at which a fetus has a 50% chance to survive after delivery [6]. The age of viability is inconsistent worldwide as survival of the premature fetus is highly dependent on immediate neonatal care. Biomedical maturity and technological advancement are the determinants of viability [7]. Therefore, gestational age at viability in the Middle-East, Africa and other low resource countries is generally greater than that in developed countries [5]. In the United States, the limit of viability is 24 weeks’ gestation because of advanced neonatal care. Moreover, some institutes set the definition to 23 weeks’ gestation based on their internal data [8]. In Portugal, the limit of viability ranges between 25 and 26 weeks’ gestation [9]. In India, an Asian country, neonatal mortality is 55% when delivered at 26 to 27 weeks’ gestation [10]. On the other side, African countries like Nigeria typically set there limit of viability to 28 weeks’ gestation [11]. Although there is insufficient data, 28 weeks seems to be the most acceptable cut-off among obstetricians in the Middle-East.

Although diagnosis and management of PROM has been relatively consistent among internationally recognized obstetric committees, the practice in low-resource settings is typically challenged by several issues that need to be addressed. As mentioned before, gestational age at viability is a determinant of management plan and unfortunately, it significantly differs from viability limits in the developed countries. Furthermore, due to neonatal care advancement, neonatal outcomes are more favorable if delivered in high resource countries compared to resource challenged. Therefore, even newborns delivered at 34 weeks’ gestation do not necessarily present reassuring outcomes in low-resource settings. Lack of excellent dating presents another challenge that complicates the decision-making process.

Low resource settings may not have access to some tests due to their cost. Fortunately, current standard PROM management does not warrant expensive work-up. On the other side, other tests tend to be less expensive and easier to access in Middle-Eastern and African countries eg, obstetric ultrasound is done by the obstetrician as a part of prenatal visit with no or minimal extra-cost. Unfortunately, some non-expensive approaches are not commonly performed in these countries due to lack of either awareness and/or experience on their rule or technique (such as microscopic examination, ferning test,. etc).

Another challenge is related to availability and proximity of tertiary care in low-resource countries. In many areas, tertiary maternal care or neonatal intensive care may be more than 150 miles away from residential areas. Given the fact that air transportation is not typically an option, attention to these constraints should be paid to ensure patient safety and guide clinical management at location of presentation.

Many aspects of management of PROM have been adequately addressed either by internationally recognized guidelines or by local practice policies. Our approach starts with identifying clinical questions that need to be answered, achieved by thorough reviewing of the latest versions of high-quality guidelines that are representative of global practice. Most recent versions of published guidelines were considered (to February 2019). These guidelines include WHO recommendations (November 2015) [4], ACOG practice bulletin (January 2018) [1], the Society of Obstetricians and Gynecologists of Canada (SOGC) (September 2017) [12], the RANZCOG (March 2017), and NICE guidelines (November 2015) [2,3]. The second step, each clinical question is assigned to a panel of 2 researchers to review how a question is covered in these guidelines. The aim is to verify if a question is consistently answered, to identify the range of answers provided if inconsistent among different guidelines, and to assess the adequacy of evidence supporting these answers. The third step is to search the literature for newly published studies related to the topic as well as the original studies conducted primarily in the Middle-East or alternatively, in low resource countries that may share similar challenges. MEDLINE, EMBASE, SCOPUS and Cochrane library were searched for prospective studies, clinical trials, systematic reviews and meta-analyses from 2000 to June 2019. Search terms include “rupture of membranes”, “PROM”, “PPROM”, “previable”, “induction of labor”, “chorioamnionitis”, “latency antibiotics”, “antenatal steroids”, and “preterm labor”. The fourth step is to survey obstetricians in the Middle-East to identify their needs, challenges and actual practice. These surveys include obstetricians at different levels of experience, who are involved in either community based or university-based practice. The fifth and final step is to review the outcomes of these steps and fit the highest level of evidence that is appropriate for low resource settings.

The level of evidence is determined in accordance to Oxford Centre for Evidence-based Medicine – Levels of Evidence, which stratifies studies depending on their design to level 1 (A to C), level 2 (A to C), level 3 (A and B), level 4, and level 5 [13] (Appendix1 in the Online Supplementary Document). A KAST approach will be used to present this topic; Knowledge, Assessment, Sharing decision, and Treatment (Table 1).

Clinical assessment aims to confirm the diagnosis and minimize introduction of infection. History taking should include; time of onset of fluid leakage, description of the amount and color of leaking fluid, onset of uterine contractions (if present) in relation to fluid leakage, and the presence of abdominal pain or vaginal bleeding. History alone does not confirm or rule out the diagnosis and thus a thorough physical examination is essential in all cases.

It starts with assessment of vital signs, of which temperature is particularly important because of the risk of intra-amoniotic infection (IAI). Abdominal examination is performed to rule out abdominal tenderness that may indicate chorioamnionitis or placental abruption, to palpate any uterine contractions and to assess the fetal presentation. Sterile speculum examination is used to visualize the cervix. Diagnostic signs of PROM during speculum examination includes a direct leakage from the cervix and “pooling” of fluid at the posterior fornix. Valsava maneuver may be attempted if no fluid leakage can be appreciated. Speculum examination may also detect umbilical cord prolapse, cervical dilation, or vaginal bleeding. Vaginal swab should be obtained for culture if infection is suspected. Speculum examination can be performed shortly after presentation if there is subjective or objective evidence of ongoing leakage or if there is any clinical concern that warrants immediate assessment. Otherwise, the patient may be given a 30-minute period in a semi-Fowlers position to allow “pooling” of amniotic fluid in the vagina. Placement of a sterile pad for 30 minutes to 1 hour can also help to identify the color and odor of the leaking fluid. This can differentiate amniotic fluid from urine, which seems to be the most common differential diagnosis in these cases [4].

Digital examination is not superior to a speculum inspection and it should be avoided to reduce the risk of infection [14]. Digital examination may be warranted if the patient has apparently frequent and efficient contractions. However, if induction of labor is planned and no uterine contractions are appreciated, initiation of induction without a digital examination is justified. As a rule, digital examination is to be avoided unless it may alter the clinical decision.

This clinical approach is supported by ACOG, NICE and RANZCOG and is appropriate for low-resource settings [1-3]. Because of low-resource practice challenges including the possibility of delayed presentation, we would like to emphasize the WHO recommendations to further reduce the risk of infection [4]. In addition to the use of a sterile speculum, attention should be paid to hand washing and using gloves. On the other side, prevention of infection should extend to physicians who should handle any vaginal discharge including blood with caution [4], and universal precautions to prevent infection transmission is advised and is standard of care.

Diagnosis is made following history and physical examination in most cases [6]. Visualization of fluid pooling in the posterior cervix is the basis of diagnosis. However, several tests are available to confirm that it is amniotic fluid. These tests include “ferning” of dried vaginal fluid under microscopic examination, basic vaginal fluid pH (Nitrazine test), and the use of amniotic fluid detection strips, if pooling is not visualized.

Both microscopic examination of ferning and vaginal pH testing require some degree of fluid pooling to allow proper testing and interpretation. Microscopic examination is performed by collecting pooled vaginal fluid and smearing it over a clean microscope slide, the specimen is left to air dry without covering. Long standing PROM with absence of ongoing leakage may lead to false-negative results for both tests. For microscopic examination, heavy contamination with blood or vaginal discharge is another cause of false-negative results. Fingerprints on the slide may result in false-positive results. Overall, sensitivity and specificity of microscopic examination are 98% and 88.2% among laboring women, 51.3% and 70.5% among non-laboring women, respectively [15]. On the other hand, a basic vaginal pH is not specific; the presence of blood, semen or bacterial vaginosis, may produce false-positive results [16]. Therefore, Nitrazine test is not supported by NICE recommendations [2].

In low-resource settings, we highly recommend the implementation of these relatively inexpensive tests, which facilitates correct diagnosis and management of this common problem and hence, minimizes diagnostic errors. Given the fact that vaginal pH testing may be misled by false-positive result and the need for supplies, we recommend microscopic examination over Nitrazine test in low resource settings. Despite its simplicity, when 597 obstetricians from the Middle-East were surveyed whether ferning test is implemented in their practice; 575 (96.3%) reported that they did not endorse it as a part of their practice. Therefore, we believe implementation of this test should be further supported in low resource setting knowing that this test is supportive and that a negative test does not exclude diagnosis particularly in non-laboring women.

In the absence of pooling, commercially available detection kits can identify amniotic fluid-specific proteins in vaginal fluid, specifically growth factor binding protein-1 test (AmnioQuick Duo+, Biosynex, Strasbourg, France) and placental alpha-microglobulin-1 test (Amnisure®, Qiagen N.V., Venlo, the Netherlands]). A recent multicenter prospective cohort study on 99 women concluded that both tests are comparable in diagnostic accuracy [16]. These tests have high negative predictive value for exclusion of PROM. Nevertheless, they yield high false-positive results among term and preterm women (31% and 19%, respectively) [17,18]. Another test, called ROM Plus® (Clinical Innovations, Salt Lake City, UT USA), is designed to detect both placental protein 12 and alpha-fetoprotein as indicators of amniotic fluid. Compared to Amnisure, a prospective study on 111 women showed comparability in diagnostic accuracy [19]. These tests seem to have limited confirmatory role and are not only expensive but also unavailable at many low-resource settings. Therefore, their implementation is acceptable but not essential given their potential cost-ineffectiveness. And risk of false-positive results would be especially risky in lower resources areas if delivery is recommended at early fetal gestation.

Bed-side ultrasound is an essential part of clinical assessment, primarily to confirm fetal presentation. Although clinical diagnosis of fetal presentation, using Leopold maneuvers, is a common practice in low-resource settings, a cross-sectional study on 1633 women with a singleton pregnancy between 35 and 37 weeks' gestation showed that sensitivity of clinical examination for diagnosis of non-cephalic presentation is not ideal (70%) [20]. Because fetal presentation significantly impacts the decision-making process for delivery, a screening for fetal presentation is necessary regardless of abdominal examination. Assessment of amniotic fluid volume may be a useful adjuvant test if the diagnosis is not definite. However, absence of oligohydramnios should not be used to rule out diagnosis. On the other side, amniotic fluid index (AFI) is commonly used in the Middle East, which can lead to over-diagnosis of oligohydramnios. A Cochrane review that included 5 clinical trials in a meta-analysis, concluded that the use of maximum vertical pocket (MVP) is superior to AFI, as the later increases the rate of diagnosis of assumed oligohydramnios, without improving perinatal outcomes [21]. A recent clinical trial (SAFE trial) has supported the same conclusions among low-risk women [22]. Because amniotic fluid volume is not diagnostic, there is no adequate data whether AFI is more informative than maximum vertical pocket. Therefore, the later should still be used to indicate oligohydramnios.

Although trans-vaginal cervical length assessment is a part of preterm labor work-up because of its predictive value, its rule in PROM is doubtful. A prospective study on 101 women with PROM suggested a predictive value of transvaginal measurement of cervical length for early labor but not for chorioamnionitis [23]. Another prospective study on 55 women with PPROM evaluated the use of trans-labial technique and did not find cervical length informative on either latency or risk of infection [24].

If the diagnosis is still unclear, the ultrasound guided trans-abdominal instillation of indigo carmine dye is typically performed if all other tests are equivocal. A positive test is indicated by blue staining of a vaginal tampon or pad [6]. Although the procedure is diagnostic, it is invasive, and should be performed only by a physician who is expert in performing it. Therefore, conduction of such a test in low-resource settings may be challenging and should not be implemented unless feasibility and safety are ensured. Initial assessment of fetal status is an essential part of management planning. Electronic fetal heart rate monitoring with monitoring of uterine activity is warranted at the time of presentation. No laboratory tests are required if the mother and the fetus are stable [1], and no signs and symptoms of IAI present (Table 2).

PROM has a range of maternal, fetal and neonatal outcomes. It ranges from spontaneous sealing of the defect – being the most favorable outcome – to major maternal and fetal morbidity or mortality as the worst prognosis [25]. One of our priorities in this article is to reinforce the value of patient counseling and shared decision making in high volume low resource settings where this process can, sometimes, be deficient. We believe that after assessment is complete, potential management plan and prognosis should be discussed with the patient. It is important to involve the partner and other members of the family if the patient prefers them to be involved. Counseling should include anticipated maternal, fetal and neonatal outcomes, potential complications including the plan to prevent or diagnose them early, precautions for patient, and management plan. A period is given for the patient and the family to ask questions. Eventually, the patient either agrees to the treatment plan or rejects any part of it as long as her decision is made on clear understanding of the risks and benefits. Documentation should be a part of standard practice. Given the fact that availability of resources is limited in these settings especially human resources, we highly recommend the use of visual aids (eg, pamphlets) for patient education. Also, we recommend regular training for nurses to be able to respond to patients` basic questions and concerns. Overall, low-resource settings should pay more attention to nurse involvement in patient care process, being the first line of contact with the patient.

The process of counseling and decision making should be individualized based on several factors. The following 5 points should be discussed with the patient.

History of PROM is a major risk for PROM in a subsequent pregnancy [44,45]. Although this risk factor is not modifiable, there are other modifiable risk factors that are associated with increased risk of PROM that may be discussed with the patient prior to her next pregnancy. Examples include short inter-pregnancy interval, low body mass index, cigarette smoking, and illicit drug use [46-50].

In addition, in case of previous preterm PROM, early intervention in the next pregnancy is more critical. Cervical length surveillance is widely supported by ACOG, NICE and WHO [1,2,4]. If PROM occurs prior to 34 weeks, the patient should be counseled that she can be offered progesterone treatment starting from 16 to 24 weeks' gestation with the possibility of cerclage placement if the cervical length is less than 25 mm prior to 24 weeks of gestation [51-53]. Intramuscular injection of 17-hydroxyprogesterone has been supported over vaginal progesterone by ACOG among women with prior preterm labor including prior PROM [1]. However, a recent meta-analysis of 3 RCTs with a total of 680 women with prior preterm labor concluded that daily vaginal progesterone (gel or suppositories) is an appropriate alternative to intramuscular progesterone. Unfortunately, quality of evidence of these studies is low and interpretation of these results should be taken cautiously [54]. Given the fact that intramuscular form may not be available, and administration of the medication warrants good patient complaint and availability of local nursing, we recommend considering daily vaginal progesterone as an alternative, if weekly intramuscular progesterone is not feasible.

Because management of future pregnancy is particularly important, improper documentation or lack of medical records in low resource settings can be a significant issue. Therefore, thorough history taking in first prenatal visit is mandatory. We highly recommend enforcing next pregnancy management discussion immediately after PROM delivery. Thereby, a patient is aware and is more proactive to bring this discussion to her provider in the next pregnancy and to bring a printed copy of her records if transfer of medical records may present an issue (Table 3).

Once the diagnosis of term PROM is made, treatment plan should be discussed with the patient. As mentioned earlier, it is important to promote hospitalization regardless of labor status. Home management is not justifiable in this situation. Further management is influenced by maternal and fetal status, cervical dilation, fetal presentation, and presence of any indication for cesarean delivery.

If there is no contraindication to vaginal delivery and fetal status is reassuring, the goal is to expedite vaginal delivery if the patient is not in active labor on presentation.

Administration of antibiotics has been a practice issue in many countries in the world including many Middle-Eastern countries because prescription of antibiotics is not monitored, and they are available as over-the-counter medications. Laws and regulations are unlikely to change soon and therefore, it is the providers’ ethical and professional responsibility to limit antibiotic prescription and to enforce culture change among patients and younger physicians who may believe antibiotics are the first line of management. Drug resistance is a global threatening problem. Unfortunately, data from the Middle-East are sparse. However, a survey of antimicrobial susceptibility patterns from 5 hospitals in Cairo, Egypt, from 1999 to 2000 revealed high rates of resistance among most of the bacteria studied, specially gram-negative bacilli which were resistant to most relevant antibiotics [55].

Antibiotic administration is limited to 3 indications among women with term PROM: (1) prophylaxis against GBS; the indications are not any different from those conducted in uncomplicated labor; (2) Treatment of IAI if clinically suspected; if the patient experiences fever >39°C or persistently over 38°C along with leukocytosis, fetal tachycardia or purulent vaginal discharge according to ACOG committee opinion [1]. A key point to reduce the need for antibiotics is to minimize the number of digital pelvic examinations.

The third indication of antibiotic administration is prophylaxis against infection, and it presents the most controversial indication. A meta-analysis of 5 RCTs including 2699 women who were randomized to antibiotic group vs control group shows that antibiotic prophylaxis for term or near-term PROM is not associated with maternal or neonatal benefits. However, subgroup analysis of women with latency longer than 12 hours shows that women who received prophylactic antibiotics experience lower rates of chorioamnionitis and endometritis (risk ratio (RR) = 0.49, 95% CI = 0.27-0.91) and 0.12 (95% CI = 0.02-0.62), respectively [56]. Nevertheless, data from a local RCT conducted in Egypt in 2014 did not show statistical differences in the incidence of maternal infection or neonatal sepsis when prophylactic antibiotics were administered at or beyond 36 weeks of gestation [57]. We conducted a survey on 333 obstetricians; 291 (87.4%) stated that their practice supports prophylactic antibiotics for women with PROM from admission to delivery, 31 (9.3%) give antibiotics if latency exceeds 18 hours, 11 (3.3%) administer antibiotics only if GBS culture is positive or to treat chorioamnionitis.

Although the goal is to achieve expedited delivery, the decision whether to start induction of labor immediately or to consider initial watchful expectancy for possible spontaneous onset of labor is debatable among women who are not in active labor. Based on a Cochrane review of 23 RCTs in 2017, ACOG supports immediate induction of labor over expectant management because it reduces latency between PROM and delivery and thus, the risk of maternal and neonatal complications [1,58]. In women with intact membranes, cervical bishop score is used to assess cervical favorability and to determine whether intravenous oxytocin, misoprostol or Foley bulb is selected for induction of labor. Among women with term PROM, oxytocin may be superior to all other options. Among women with term PROM, induction of labor with oxytocin was associated with fewer digital vaginal examinations, shorter latent phase, shorter labor, and shorter hospital stay compared to prostaglandin induction [59]. Furthermore, application of vaginal prostaglandins can also be associated with the possibility of “washout”. On the other hand, the use of buccal prostaglandins may be associated with more systemic side effects including fever, which may interfere with monitoring of maternal infection. Placement of Foley bulb has been associated with theoretical concerns of chorioamnionitis risk due to the insertion of a foreign body. However, available evidence is limited. A retrospective study on 124 women showed increased risk of IAI with Foley bulb compared to oxytocin induction group. Yet, the risk is not statistically significant (P = 0.1) [60]. Preliminary results from Foley Bulb for Labor Induction in Premature Rupture of Membranes in Nulliparas (FLIP trial) which compared concurrent Foley bulb/oxytocin vs oxytocin alone did not show an increased risk of chorioamnionitis. Nevertheless, addition of a Foley bulb to oxytocin does not shorten induction to delivery time [61].

If cesarean delivery is indicated, preliminary results from a meta-analysis of 19 studies (including 6179 patients) undergoing cesarean delivery showed that vaginal irrigation with povidone-iodine 1% reduces the risk of endometritis and wound complications [62]. It is reasonable to consider this intervention given its simplicity and low cost, and increased risk of infection among women with PROM. Preterm breech presentation is an indication of cesarean delivery in the setting of PROM. ECV is an appropriate option in the absence of PROM. However, this procedure is associated with high risk of failure and significant maternal and fetal complications and is not recommended [63]. Although amnioinfusion may appear to be an option, a recent RCT (119 women) showed that it did not improve the success rates [64]. However, data on ECV following amnioinfusion in women with PROM are limited.

ACOG recommends immediate delivery once pregnancy is 34 weeks or beyond because conservative management is associated with an increased risk of amnionitis (16% vs 2%, P = 0.001), prolonged hospitalization (5.2 vs 2.6 days, P = 0.006), and lower mean umbilical cord pH at delivery (7.25 vs 7.35, P = 0.009) without improvement of prematurity related perinatal complications. More recent recommendations also support administration of a course of antenatal steroids between 34 0/7 and 36 6/7 weeks of gestation if it is not given earlier in pregnancy. The later recommendation is supported by an RCT (including 2827 infants) which showed that administration of betamethasone to these women significantly reduced the rate of neonatal respiratory complications [65]. We surveyed 449 obstetricians on gestational age beyond which antenatal steroids are not administered. The answer was 34 weeks by 294 obstetricians (65.5%), 37 weeks by 106 obstetricians (23.6%), 32 weeks by 42 obstetricians (9.3%), and 39 weeks by 7 obstetricians (1.6%). Therefore, we believe that administration of steroids for up to 37 weeks should be promoted in low resource setting where neonatal care of preterm babies is generally challenging. However, steroid usage may be contraindicated in some cases such as type I DM and sepsis.

Nevertheless, immediate delivery compared with expectant management after preterm pre-labor rupture of the membranes close to term (PPROMT trial) results were published in 2016. This RCT was conducted at 65 centers across 11 countries with a total of 1839 women randomized to either immediate intervention or expectant management. The study showed that neonates of patients assigned to immediate delivery had higher rates of respiratory distress (RR = 1.6, 95% CI = 1.1-2.3; P =  0.008), higher rates of mechanical ventilation (RR = 1.4, 95% CI = 1.0-1.8; P =  0.02) and longer admission to intensive care compared with neonates born to mothers in the expectant management group. However, women assigned to the expectant management group were at higher risks of antepartum or intrapartum hemorrhage (RR = 0.6, 95% CI = 0.4-0.9), intrapartum fever (RR = 0.4, 95% CI = 0.2-0.9), postpartum antibiotics (RR = 0.8, 95% CI = 0.7-1.0), and longer hospital stay. However, the risk of caesarean delivery was lower (RR = 1.4, 95% CI = 1.2-1.7) in the expectant group. Therefore, the study recommended expectant over immediate management at this gestational age [66]. A recent meta-analysis of 3 RCTs (2563 mothers, 2572 neonates) revealed comparable rates of the composite adverse neonatal outcomes between expectant and immediate managements. However, neonatal sepsis rates were 2.6% and 3.5%, respectively (RR = 0.74, 95% CI = 0.47-1.15). The risk of respiratory distress syndrome among neonates in the immediate delivery arm was higher (RR = 1.47, 95% CI = 1.10-1.97), and they were more likely to be admitted to the neonatal intensive care unit or special care nursery (RR = 1.17, 95% CI = 1.11-1.23). Immediate delivery was associated with lower risk of antepartum hemorrhage (RR = 0.57, 95% CI = 0.34-0.95) and IAI (RR = 0.21, 95% CI = 0.13-0.35), but higher risk of cesarean delivery (RR = 1.26, 95% CI = 1.08-1.47) [67].

Data from the Egyptian National Perinatal/Neonatal Mortality study in 2004 showed that neonatal mortality rate was 25 per 1000 live births; prematurity accounted for 39% of these cases. Perinatal mortality rate was 34 per 1000 births, 21% was attributed to prematurity. The mean gestational age of neonates diagnosed with early neonatal death (1-6 days) was 7.9 ± 1.1 months and 8.6 ± 0.8 months for late neonatal deaths (day 7 to 28). For stillbirths, the mean gestational age was 7.9 ± 1.3 months [68]. These numbers reflect that neonatal outcomes are less favorable among preterm neonates managed in low resource facilities compared to where the above studies were conducted. Therefore, we believe that expectant management should be adopted in low resource areas where management of neonatal complications is more challenging, highly expensive, and associated with poor prognosis compared to maternal complications, even among late preterm population. Gestational age at which the benefits of immediate delivery overweighs benefits of expectant management may be modified based on local perinatal data. Nevertheless, we recommend expectant management up to 37 weeks of gestation if PROM occurs beyond 34 weeks of gestation. RCOG has declared similar recommendations in 2019 based on these recent studies as well [69]. Regardless, delivery should be postponed at least 48 hours after administration of a course of prenatal steroids if maternal or fetal conditions are stable.

Administration of antibiotics should be similar to term PROM. However, women with unknown GBS status are treated as GBS positive and should be administered GBS prophylaxis. Latency antibiotics may be administered if expectant management to 37 weeks of gestation is elicited (see under PROM before 34 weeks of gestation).

As discussed earlier, gestational age at viability is inconsistent. A recent large joint workshop in the USA defined periviable birth as delivery between 20 0/7 weeks to 25 6/7 weeks of gestation [79]. Although the definition of viability is less standardized in low-income countries, it lies between 25 0/7 to 28 0/7 weeks of gestation in most instances [9-11]. Pulmonary hypoplasia is the major risk factor of neonatal mortality [80]. The two major determinants of pulmonary hypoplasia are early gestational age of PROM and low residual amniotic fluid [81,82]. Pulmonary hypoplasia is rare if PROM occurs after 23-24 weeks of gestation, which is the critical gestational age for alveolar development [83]. This explains why the age of viability corresponds to this gestational age in developed countries. Nevertheless, lack of advanced care and equipment among most facilities in low resource areas is associated with high neonatal mortality due to severe respiratory distress syndrome or neonatal sepsis, hence, the discrepancy in gestational age of viability. Therefore, we recommend that, unless there are a definite local data, pregnancy between 24 and 28 weeks should be considered periviable and pregnant women who presents with Preterm PROM or preterm contractions should be immediately transferred to a tertiary care center with the highest neonatal intensive care for hospitalization if both the patient and the fetus are stable to transfer. Home care should not be an option [79]. Patient should be counseled on survival rates, morbidities, short term and long term outcomes of periviable birth through an involved neonatologist [79]. Finally, there is no current evidence to support routine cesarean section for delivery of women with periviable PROM unless obstetrically indicated; neonatal outcomes seem to be comparable to vaginal delivery [84,85].

However, prior to the age of viability, expectant management is not the standard of caregiven the high risk of maternal infection with prolonged latency. This risk is not justified given the poor neonatal outcome secondary to pulmonary hypoplasia specially in the presence of oligohydramnios or anhydramnios. A discussion should be conducted with the patient to allow a shared decision based on realistic expectations. However, patient decision should be supported and a plan that ensures patient safety should be discussed regardless of her decision.

If PROM occurs prior to the age of viability, hospitalization is not medically necessary as long as the patient is aware of warning signs of maternal infection and transfer to a tertiary center is feasible in a reasonable amount of time, ideally 15 to 30 minutes. However, hospitalization may be an acceptable option if patient safety at home cannot be ensured. Home management should be accompanied by a good follow-up plan; patients may feel motivated not to seek medical advice if they develop symptoms to avoid the decision of pregnancy termination. Sequences of maternal infection should be thoroughly discussed including the possibility of maternal sepsis and maternal mortality especially with delayed presentation. It is not unreasonable to schedule weekly visits to check maternal status and fetus viability. However, no intervention is proven to be beneficial at this stage including antenatal steroids, magnesium sulfate administration or treatment of GBS. Administration of latency antibiotics may be offered if the patient is interested in expectant management to prolong pregnancy and reduce the risk of neonatal infection. The earliest window when antenatal steroids may provide neonatal benefit is for babies delivered between 23th to 25th weeks of gestation. Neonatal benefits include reduction in neonatal mortality, intraventricular hemorrhage, periventricular leukomalacia, and necrotizing enterocolitis. [84,86]. No benefit is expected prior to this gestational age [87].

Less data are known regarding magnesium sulfate use for neuroprotection prior to age of viability. The 3 large clinical trials (conducted in Australia, France and USA), which highlights the protective role of magnesium sulfate against cerebral palsy, recruited women with a median (interquartile range) of 27 weeks and 3 days (25 5/7 weeks – 28 5/7 weeks), women whose gestational age ranges between 24 weeks to 32 weeks 6 days, and average gestational age of 28.3 ± 2.5 weeks, respectively [88-90]. Therefore, the standard of care is to administer magnesium sulfate after the age of viability. If a patient is managed conservatively and she reaches the gestational age of viability, she should be admitted to the hospital and managed like women with preterm PROM but with knowing that the prognosis is not the same (Table 4).