[Tocolysis: indications and contraindications. When to start and when to stop].

Clinical trials have demonstrated that most tocolytic agents such as beta mimetics, atosiban and indometacine -exception made for magnesium sulfate- extend the duration of pregnancy, when compared to placebo. They reduce the rate of delivery at 24 hours, 48 hours and at 7 days. There is no proof however for their beneficial effect on perinatal or neonatal outcomes. Usual obstetrical contraindications of tocolytic treatments (infection, genital haemorrhage, fetal distress and certain maternal conditions) are often determined on a case by case basis, rather than upon evidence based medicine. Tocolysis may be considered in case of maternal infection without chorioamniotitis or during moderate hemorrhage due to placenta preavia (NP 4). There is no objective argument to refute a tocolytic attempt due and in accordance with a low gestationnal age limit. The risk for neonatal respiratory distress syndrome is higher at 34 weeks compared to 35 and 36 weeks. The studies were however realized before the corticosteroid era of fetal lung maturation. There are no available randomized trials concerning the benefits of tocolysis after 34 weeks of amenorrhea. Furthermore, there are no satisfactory trials preconising a specific tocolytic agent based on amniotic liquid study of lung maturity. In case of advanced cervical dilatation, rare studies have concluded the usefulness of tocolysis to allow fetal lung maturation by corticosteroids. Most studies have evaluated the effectiveness of tocolytic treatment during 48 hours. There is no evidence for continuing tocolytic treatment after an effective tocolysis prescribed during the first 48 hours. For intermediate situations, between franc success or failure of tocolysis (such as reduction but persistent painful uterine contractions or major cervical modification at a low gestational age despite reduction of uterine contractile activity), literature guidelines are poorly contributive. The management of such patients is most often empirical.