Prematurity is defined as birth before 37 weeks of amenorrhea. It is the leading cause of perinatal morbidity and mortality. It is estimated that premature births to 15 million worldwide in 2010, including 60,000 children in France (7,4% of births) and 12000, born before 32 weeks of amenorrhea (term defining great prematurity). More than half of these births follow spontaneous work.

Before 34 weeks of amenorrhea, prematurity requires specific maternal-fetal management centered primarily on antenatal corticosteroid treatment for fetal maturation including 2 intramuscular injections at 24 hour intervals. This cure can only be done once and its benefit is proven in the 7 (to 14) days before the birth (recommendation for the clinical practice of the French National College of Obstetricians and Gynecologists of December 2016). The single course of antenatal corticosteroids before 34 weeks of amenorrhea is associated in the neonatal period with reduction in morbidity and mortality, and in the long term with an improvement of the survival without sequelae, if the full cure is administered in the 7, to 14, days before birth. The second key element of the management of a premature delivery threats is the admission in a pediatric structure adapted to the degree of prematurity.

The care of a premature delivery threats before 34 weeks of amenorrhea care leads to hospitalization with possible transfer to a maternity of pediatric adapted level, tocolysis, a biological and bacteriological assessment, and, sometimes, other examinations and treatments.

The prediction of premature labor is a challenge. Current methods, such as vaginal examination, cervicometry and detection of fetal fibronectin, make it possible to obtain a negative predictive value (NPV) approaching 100% but a poor positive predictive value (PPV), thus 8 out of 10 patients hospitalized and treated for premature delivery threat no not give birth within 7 days, or even before 34 weeks of amenorrhea. A better prediction of preterm delivery has two benefits: administer antenatal corticosteroid therapy in high-risk patients at the right time and not treat not-at risk patients.

The PartoSure® test, which detects the placental protein vaginal alphamicroglobulin-1 (PAMG-1), has satisfactory metrological qualities according to observational studies : NPV 98% and PPV 75% of delivery within 7 days.

However this test has never been evaluated in real clinical condition. In our study, the result of the test becomes decisional for the care.

The hypothesis is that use of this test will improve the prediction of preterm delivery during a first consultation for preterm delivery threat and thus allow more frequent prenatal corticosteroid treatment in optimal period, avoiding treating non-at risk patients.

Inclusion Criteria:

• Major patient
• Pregnant of a singleton
• Followed at Brest hospital and resident unless 30 kms of Hospital
• Patient consultant for preterm delivery threat before 34 weeks of amenorrhea with more than 6 contractions per hour (ie 1 in 10 minutes) felt and / or sensed by external toco-ergometry associated with cervical changes objectified by a measurement of the cervix by ultrasound between 15 and 25 mm
• With capacity of a free and informed consent
• Affiliated member of the French social security system

Exclusion Criteria:

• Patient transferred for preterm delivery from a peripheral hospital center
• Patient not followed at Brest hospital
• Patient doesn't lives within 30 km of the Brest hospital
• Clinical premature rupture of membranes
• Circled patients
• Important metrorrhagia
• Cervix <15 mm
• Cervical dilatation > 3 cm
• Isolated short cervix defined by a short cervix in measurement of the cervix by ultrasound without uterine contraction felt or captured in external toco-ergometry
• Placenta previa, placenta percreta
• Multiple pregnancy
• Suspicion of chorioamnionitis
• Associated maternal-fetal pathology that may induce prematurity
• No prenatal corticosteroid treatment for this pregnancy