In France, prolonged pregnancies (≥41+0weeks) involve 15–20% of pregnant women, and post-term pregnancies (≥42+0 weeks) approximately 1%. The frequency of post-term pregnancies is very heterogeneous: in Europe and the United States, it ranges from 0.5% to 10% according to country.
In prolonged pregnancies, the cesarean section rate—especially the emergency cesarean rate—is multiplied by approximately 1.5 (grade B). From 370–6 to 430–6 weeks, the risk of perinatal mortality increases regularly, from 0.7‰ to 5.8‰.
Meconium aspiration syndrome is responsible for substantial morbidity and mortality, and its incidence increases regularly between 38+0 and 42+6 weeks, from 0.24‰ to 1.42‰ (grade B). Similarly, the risks of neonatal acidosis (grade B), 5-min Apgar scores less than 7 (grade B) and admissions to neonatal intensive care (grade B) increase progressively between 38+0 and 42+6weeks. These risks appear to double for post-term growth-restricted newborns (grade C).
Ultrasound dating of the pregnancy makes it possible to reduce the risk that it will be incorrectly considered prolonged and that labor will therefore be induced unnecessarily. To harmonize practices, if the crown-rump length (CRL) is correctly measured (this measurement should be taken between 11+0 and 13+6 weeks, when CRL should measure from 45 to 84 mm), ultrasound dating based on it should be used to determine the official date pregnancy began, regardless of its difference from the date assumed by the patient or estimated based on the date of the last menstrual period. This rule does not apply to pregnancies by IVF, for which the date pregnancy began is defined by the date of oocyte retrieval (Professional consensus).
From 370–6 to 430–6 weeks, the risk of perinatal mortality increases regularly and there is no threshold at which a clear increase in perinatal mortality becomes visible. Fetal monitoring by cardiotocography (CTG) that begins at 41+0 weeks would cover approximately 20% of women and reduce perinatal morbidity compared with monitoring that begins at 42+0 weeks (grade C). The frequency recommended for this monitoring ranges between two and three times a week (Professional consensus).
For ultrasonography assessment, measurement of the largest fluid pocket is recommended, because measurement of the amniotic fluid index (that is, the sum of the four quadrants) is accompanied by more diagnoses of oligohydramnios, inductions of labor, and cesareans for fetal distress without any improvement in neonatal prognosis (grade A). The practice of assessing the Manning biophysical score increases the number of diagnoses of oligohydramnios and fetal heart rage (FHR) abnormalities and generates an increase in the rates of inductions and cesareans without improving neonatal prognosis. The use of this biophysical score in monitoring prolonged pregnancies is therefore not recommended (grade B).
In the absence of a specific disorder, induction of labor can be proposed in patients between 41+0 and 42+6 weeks (grade B). Nonetheless, the choice of prolongation beyond above 42+0 weeks appears to involve an increase in fetal risk, which must be explained to the patient and balanced against the potential disadvantages of induction (Professional consensus).
Stripping the membranes can reduce the duration of pregnancy by increasing the number of patients going into labor spontaneously during the week afterward (grade B). Compared to an expectant approach, it does not increase the cesarean section rate (grade A). It reduces recourse to induction by 41% at 41+0 weeks and by 72% at 42+0 weeks (grade B), without increasing the risk of either membrane rupture or maternal or neonatal infection (grade B).
Used as a tampon or vaginal gel, prostaglandins E2 (PGE2) are an effective method of inducing labor (grade A). They can be used to induce labor successfully, regardless of cervical ripeness (grade A). If misoprostol is chosen, the lowest dose is to be preferred, starting with a vaginal dose of 25 μg every 3–6 h (grade A). For misoprostol, more powerful studies remain necessary for better defining the doses, routes of administration, tolerance and indications. Misoprostol at any dose is contraindicated in women with uterine scars (grade B). Placement of an intracervical Foley catheter is an effective mechanical means of inducing labor, with less uterine hyperstimulation than prostaglandins and no increase in the cesarean section rate (grade A). Nonetheless, as the risk of infection might be increased, this technique requires more robust evaluation before entering general practice (grade B).
In cases of meconium-stained amniotic fluid, pharyngeal aspiration before delivery of the shoulders is not recommended (grade A). The team managing a post-term newborn with meconium-stained amniotic fluid at birth must know how to perform intubation and, if the intubation is not helpful, endotracheal aspiration (grade C) and ventilation with a mask. Routine endotracheal intubation of a vigorous newborn is not recommended (grade A).
The sponsor (French College of Gynecologists and Obstetricians (CNGOF)) appointed a steering committee (Appendix A) assigned to define the exact questions to be put to the experts, to choose them, follow their work and draft the synthesis of recommendations resulting from their work. The experts analyzed the scientific literature on the subject to answer the question raised. A literature review identified the relevant articles through mid-2011 by searching the MEDLINE database and the Cochrane Library. The search was restricted to articles published in English and French. Priority was given to articles reporting results of original research, although review articles and commentaries were also consulted. Guidelines published by organizations or institutions such as the American College of Obstetricians and Gynecologist (ACOG) [3], the Canadian Society of Gynecology and Obstetrics (SOGC) [4] and the World Association of Perinatal Medicine (WAPM) [5] were reviewed, and additional studies were located by reviewing bibliographies of identified articles.
For each question, each overview of validated scientific data was assigned a level of evidence based on the quality of its data, in accordance with the framework defined by the HAS (French Health Authority), summarized below.
1.1. Quality of evidence assessment
LE1: very powerful randomized comparative trials, meta-analysis of randomized comparative trials;
LE2: not very powerful randomized trials, well-run non-randomized comparative studies, cohort studies;
LE4: non-randomized comparative studies with large biases, retrospective studies, transversal studies, series of cases.
The steering committee drafted a synthesis of the recommendations, based on the replies by the experts. Each recommendation for practice was assigned a level defined by the HAS as follows:
1.2. Classification of recommendations
Level A: Recommendations are based on good and consistent scientific evidence.
Level B: Recommendations are based on limited or inconsistent scientific evidence.
Level C: Recommendations are based primarily on consensus and expert opinion.
Professional consensus: In the absence of any conclusive scientific evidence, some practices have nevertheless been recommended on the basis of agreement between members of the working group (Professional consensus).
All texts were reviewed by persons not involved in the work, i.e., practitioners in the various specialties concerned (listed in the Appendix A) and working in varying situations (public, private, university or non-university establishments). Once the texts were reviewed, changes were made, if appropriate, considering assessment of the quality evidence.
The texts are cited [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], and [13] but the individual references included in each text are not reported here because of the enormous amount of space that would require.
Depending on the method of calculation, genetic factors and some maternal characteristics, the duration of gestation ranges from 280 to 290 days from the first day of the last menstrual period (for regular 28-day cycles). This duration is generally expressed in France as weeks of amenorrhea, which corresponds to the English expression “weeks of gestation”. The duration of pregnancy therefore varies between 40+0 and 41+3 weeks. Setting an official due date for women appears illusory, for these vary according to author, mode of calculation and country. This generally does not present problems, however, for what is essential is that physicians explain to their patients the date by which specific surveillance for “prolonged pregnancy” should start and the date at which induction of labor might be envisioned.
Conventionally, and essentially arbitrarily, a pregnancy is considered to be “prolonged” after 41+0 weeks, but the infant is not considered “post-term” until 42+0 weeks (professional consensus). A term birth thus occurs during the period from 37+0 to 41+6 weeks (Fig. 1). In view of the moderate, progressive increase in perinatal morbidity during this period and beyond, the three days (at a maximum) that separate 41+0 weeks from 9 months do not appear to produce a significant variation in perinatal complications, and this approximation between 41+0 weeks and 9 months is acceptable (Professional consensus).
3. Frequency of prolonged and postterm pregnancies [7]
In France, prolonged pregnancies (≥41+0 weeks) involve 15–20% of pregnant women, and post-term pregnancies (≥42+0 weeks) approximately 1%. This frequency of post-term pregnancies is very heterogeneous: in Europe and the United States, it ranges from 0.5% to 10%, varying by and within countries (grade B). These variations simultaneously reflect the diversity of the populations studied and the variations in obstetric practices between countries: early pregnancy dating by ultrasound and increasingly frequent recourse to induction of labor have jointly contributed to a progressive diminution in the incidence of prolonged and post-term pregnancies in most countries (grade B).
4. Maternal complications of prolonged pregnancies [7]
In prolonged pregnancies, the cesarean section rate—especially the emergency cesarean rate—is multiplied by approximately 1.5 (grade B). Nonetheless, the lack of precision in the studies prevents us from being able to determine whether the increased cesarean risk is linked to the prolonged pregnancy or to policies that promote induction of labor starting at some specific gestational age. Prolonged pregnancy is also accompanied by a moderate increase of the risk of third- and fourth-degree perineal lesions, postpartum hemorrhages, chorioamnionitis and endometriosis (grade C).
5. Fetal and neonatal complications of prolonged pregnancies [8]
From 370–6 to 430–6 weeks, the risk of perinatal mortality increases regularly, from 0.7‰ to 5.8‰ (Fig. 1). In prolonged pregnancies, the fetus is at increased risk of oligohydramnios (grade C), fetal heart rate abnormalities (grade A), and meconium-staining in utero (grade B).
The risk of in utero death has dropped considerably in recent decades (grade B). In 2001, its frequency was estimated at 1.6‰ to 3.5‰ births in developed countries (grade B).
Meconium aspiration syndrome is responsible for substantial morbidity and mortality, and its incidence increases regularly between 38+0 and 42+6 weeks, from 0.24‰ to 1.42‰ (grade B). Similarly, the risks of neonatal acidosis (grade B), 5-min Apgar scores less than 7 (grade B) and admissions to neonatal intensive care (grade B) increase progressively between 38+0 and 42+6 weeks. These risks appear to double for postterm growth-restricted newborns (grade C). Post-term newborns have an increased risk of neurologic complications such as neonatal convulsions (risk multiplied by 1.5) (grade B), anoxic-ischemic encephalopathy (grade C), cerebral motor disorders (risk multiplied by 2.4) (grade B), psychomotor developmental disorders (risk multiplied by 2.2) (grade C) and childhood epilepsy (risk multiplied by 1.9) (grade B), but these complications cannot be directly attributed to the post-term status. The absolute incidence of these risks nonetheless remains low.
Post-term newborns have a risk of macrosomia five times higher than that of term newborns (grade B), and macrosomia (including in newborns of mothers without diabetes) increases the risk of hypoglycemia (grade B). Although their prevalence remains low (less than 1%), post-term newborns are at increased risk of obstetric injuries related to shoulder dystocia (grade B) and of bone fractures (grade C). Finally, postmaturity is a risk factor for polycythemia (grade C), and the risk of neonatal sepsis increases by a factor of 1.3 in post-term newborns (grade B).
6. Pregnancy dating [9]
Ultrasound dating of the pregnancy makes it possible to reduce the risk that it will be incorrectly considered prolonged and that labor will therefore be induced unnecessarily (grade A). For spontaneous conceptions, it is routinely recommended to use crown-rump length (CRL) measurement before 14 weeks to determine the date pregnancy began and to calculate gestational age (grade C). Ideally, this measurement should be taken between 11+0 and 13+6 weeks (when CRL should measure from 45 to 84 mm).
To harmonize practices, if the CRL is correctly measured, the ultrasound dating based on it should be used to determine the official date pregnancy began, regardless of its difference with the date assumed by the patient or estimated based on the date of the last menstrual period. It should be explained to the patient that this measurement provides the best determination of the theoretical ultrasound-determined date pregnancy began, to serve as the reference for monitoring the pregnancy.
If the quality criteria for CRL measurement before 14 weeks are not met, the date pregnancy began should be determined by the date of the last menstrual period and the usual duration of the patient's cycles. This rule does not apply to pregnancies by IVF, for which the date pregnancy began is defined by the date of oocyte retrieval (Professional consensus). Before any intervention for prolonged pregnancy, it should be verified that the best indicators were used for dating the pregnancy: date of oocyte retrieval, CRL with its quality criteria, or date of last menstrual period (Professional consensus).
In the absence of a first trimester ultrasound and other information for dating the pregnancy (e.g., when the date of the last menstrual period is unknown or cycles are irregular), ultrasound head measurements must be chosen to estimate gestational age (head circumference or biparietal diameter before 18 weeks) (grade B). When gestational age is still uncertain above 22 weeks, repeating the biometric measurements 15–20 days later makes it possible to reassess the growth process and sometimes to estimate gestational age more accurately (Professional consensus).
7. At what gestational age should monitoring begin and how often? [10]
From 370–6 to 430–6 weeks, the risk of perinatal mortality increases regularly and there is no threshold at which a clear increase in perinatal mortality becomes visible. Beginning fetal monitoring at 40+0 weeks would require close monitoring of nearly half of all pregnant women, and nothing indicates that such a policy would reduce neonatal morbidity or mortality. On the other hand, fetal monitoring that begins at 41+0 weeks would cover approximately 20% of women and reduce perinatal morbidity compared with monitoring that begins at 42+0 weeks (grade C). The frequency recommended for this monitoring is 2–3 times a week (Professional consensus).
7.1. Are there populations that should be particularly monitored? [10]
Fetal mortality among African–American women does not rise significantly at the end of pregnancy (grade B). Nonetheless, no data justify modifying practices for prolonged pregnancies in this population.
The combination of fetal growth restriction and prolonged pregnancy is a risk factor for perinatal mortality (grade C). Nonetheless, no specific data for this population allow us to define particular management for this situation.
8. What examinations are useful for monitoring prolonged pregnancies?
8.1. Counting fetal movements? [11]
In prolonged pregnancies, the perceived diminution of fetal movements appears to be associated with an increased risk of perinatal morbidity (grade C). Current data are nonetheless insufficient to show that asking parents to count fetal movements in prolonged pregnancies yields any benefits (Professional consensus).
No method for counting these movements in this situation has been assessed.
Despite the sparse data available, it nonetheless appears sensible to recommend that patients consult their obstetric professional when they perceive a reduction in these movements (Professional consensus). The investigations to perform in this situation have not yet been clearly established (Professional consensus).
8.2. Amnioscopy [11]
The diagnostic value of amnioscopy for predicting colored or meconium-stained amniotic fluid is mediocre (grade C). No association has been demonstrated between colored amniotic fluid on amnioscopy and unfavorable neonatal outcome in prolonged pregnancies (grade C). Amnioscopy is thus not useful in these cases (Professional consensus).
8.3. Visual or computerized analysis of fetal heart rate (FHR) [12]
Although visual analysis has not been associated with improved perinatal outcome in monitoring prolonged pregnancies (grade B), recording FHR is the first-line examination for monitoring fetal well-being, and its integration into monitoring protocols appears ineluctable (Professional consensus).
In computerized FHR analysis, a short-term variation equal to or greater than 4 ms may have a negative predictive value near 100% for predicting fetal acidosis (grade B). Nonetheless, no study has demonstrated the superiority of this type of analysis over visual FHR analysis in improving perinatal prognosis (grade B). Whether or not the situation involves prolonged pregnancies, it is nonetheless a potential alternative to visual analysis, because it offers advantages in terms of organization and development of monitoring protocols, by reducing, in particular, the duration and repetition of examinations (grade B). It may also be useful in the analysis of suspicious tracings. The data for computerized FHR analysis, like those for visual analysis, remain insufficient to demonstrate the value of its use in prolonged pregnancies (grade B).
8.4. Oxytocin tests [12]
Oxytocin tests make it possible to reduce the false-negative rate of FHR recordings but the procedure is long and has a false-positive rate of the order of 30%. Data are insufficient to recommend its use as a first- or second-line tool for fetal monitoring in prolonged pregnancies (grade C).
8.5. Ultrasound test for oligohydramnios [13]
In prolonged pregnancies, there is a significant association between oligohydramnios and the onset of FHR abnormalities, meconium staining, and an increased cesarean section rate. The diagnostic value of oligohydramnios in predicting unfavorable outcome is nonetheless only moderate (grade C).
For ultrasonography assessment, measurement of the largest fluid pocket is recommended, because measurement of the amniotic fluid index (that is, the sum of the four quadrants) is accompanied by more diagnoses of oligohydramnios, inductions of labor, and cesarean sections for fetal distress without any improvement in neonatal prognosis (grade A).
In view of the risk of oligohydramnios and of increased morbidity and mortality after 41+0 weeks, it seems reasonable to suggest prenatal ultrasound monitoring of the quantity of amniotic fluid (measurement of the largest amniotic fluid pocket) starting at 41+0 weeks and thereafter twice a week (Professional consensus). In cases of oligohydramnios, defined as less than 2 cm in the largest pocket, induction can be envisioned (Professional consensus). In the absence of induction, monitoring must be reinforced (Professional consensus).
8.6. Artery Doppler indexes [13]
Few studies have found a significant association between abnormal Doppler umbilical, cerebral or aortic indexes and unfavorable outcome in pregnancies equal to or longer than 41+0 weeks. Because the diagnostic value of these Doppler indexes is very low, their routine use for prolonged pregnancies is not recommended (grade C).
8.7. The biophysical (Manning) score of fetal well-being [13]
The practice of assessing Manning biophysical scores increases the number of diagnoses of oligohydramnios and FHR abnormalities and generates an increase in the rates of inductions and cesareans without improving neonatal prognosis. The use of this biophysical score in monitoring prolonged pregnancies is therefore not recommended (grade B).
9. At what point should delivery be induced? [14]
In the absence of a specific disorder, induction of labor can be proposed to patients between 41+0 and 42+6 weeks (grade B). The moment of induction will be determined according to the mother's characteristics (local cervical conditions, uterine scar, parity, body mass index, and age), her preference, and the organization of care in the maternity unit (Professional consensus).
Nonetheless, the choice of prolongation beyond 42+0 weeks appears to involve an increase in fetal risk and must therefore be explained to the patient and balanced against the potential disadvantages of induction (Professional consensus). This comment is especially important in that the generalized use of CRL measurement during the first trimester to estimate gestational age can increase the proportion of “true” post-term pregnancies compared with the proportion in a similar population that includes inadequately dated pregnancies (Professional consensus).
10. How should delivery be induced? [15]
10.1. Stripping the membranes [15]
Stripping (also called sweeping) the membranes can reduce the duration of pregnancy by increasing the number of patients going into labor spontaneously during the week afterward (grade B). Compared to an expectant approach, it does not increase the cesarean section rate (grade A). It reduces recourse to induction by 41% at 41+0 weeks and by 72% at 42+0 weeks (grade B), without increasing the risk of either membrane rupture or maternal or neonatal infection (grade B). Its inconveniences (contractions, bleeding, and pain) must nonetheless be clearly explained and accepted before it is practised. In view of these disadvantages, stripping the membranes should not be performed during a routine examination without advance information and consent (Professional consensus).
10.2. Oxytocin [15]
Oxytocin administration is currently the reference method for induction of labor with a ripe cervix at 41 weeks and later (grade A). When the cervix is unfavorable (Bishop score ≤5), oxytocin can still induce labor but is no longer the agent of choice (grade A).
10.3. E2 prostaglandins (dinoprostone) [15]
Used as a tampon or vaginal gel, prostaglandins E2 (PGE2) are an effective method of inducing labor (grade A). They can be used to induce labor successfully, regardless of cervical ripeness (grade A). When cervical conditions are unfavorable, PGE2 can reduce recourse to oxytocin and diminish the doses needed (grade A). The use of PGE2 is associated with a risk of tachysystole that may be accompanied by abnormal FHR (hyperstimulation syndrome) (grade A).
10.4. E1 prostaglandins (misoprostol) [15]
Although misoprostol is not officially approved for prolonged pregnancies, it is an effective and uncomplicated means of inducing labor, especially for a very unfavorable cervix (grade A). It is not yet clear whether vaginal or oral administration is optimal, but the vaginal route has been assessed and chosen most often. A risk of tachysystole with or without FHR modifications is clearly found in all studies with doses of 50 μg, but these do not appear to increase the rates of either neonatal morbidity or cesarean delivery (grade A). The lowest doses are to be preferred, starting with a vaginal dose of 25 μg every 3–6 h (grade A). The use of this substance therefore requires an environment where appropriate obstetric monitoring is available in case of FHR abnormalities. More powerful studies remain necessary for better defining the doses, routes of administration, tolerance and indications. Misoprostol at any dose is contraindicated in women with uterine scars (grade B).
10.5. Mechanical means [15]
Placement of an intracervical Foley catheter is an effective mechanical means of inducing labor, with less uterine hyperstimulation than prostaglandins and no increase in the cesarean rate (grade A). Nonetheless, as the risk of infection might be increased, this technique requires a more robust evaluation before entering general practice (grade B).
10.6. Other methods [15]
The current data do not show that acupuncture is effective for inducing labor either at term or in prolonged pregnancies (grade C). Data concerning nipple stimulation and/or sexual activity are contradictory and insufficient in this indication (grade C). Finally, the use of nitric oxide (NO) donors is under study.
10.7. Special case of the uterine scar
A uterine scar is a particular situation that exposes the woman and her fetus to the risk of uterine rupture, a risk of the order of 0.5% in cases of spontaneous labor (grade A). This risk is increased, but moderately (0.7%), by induction with oxytocin and multiplied by five for inductions by prostaglandins (2.45%) (grade A). It is therefore logical to avoid non-medically indicated inductions in patients with uterine scars, to avoid prostaglandins, and to take the Bishop score into account for inductions with oxytocin. We note again that misoprostol is contraindicated in women with uterine scars (grade B).
11. What management for the newborn? [16]
In cases of meconium-stained amniotic fluid, pharyngeal aspiration before delivery of the shoulders is not recommended (grade A). The team managing a post-term newborn with meconium-stained amniotic fluid at birth must know how to perform intubation and endotracheal asp
