Objetivo: evaluar la eficacia y seguridad de atosiban en gestantes con amenaza de parto pretérmino comparado con nifedipino, indometacina, terbutalina, fenoterol y placebo.Materiales y métodos: se realizó una revisión sistemática de la literatura en ocho bases de datos electrónicas (Medline, Central, Embase, entre otras), mediante términos de búsqueda libres y estandarizados. Los desenlaces evaluados incluyeron tiempo de retardo del parto, mortalidad neonatal, proporción de eventos adversos maternos y proporción de complicaciones neonatales. Se evaluó la calidad de la evidencia por estudio y para el cuerpo de evidencia, y se sintetizó la información mediante metaanálisis, cuando fue posible; de lo contrario, se resumió de forma narrativa.Resultados: se incluyeron once estudios. Atosiban no mostró diferencias estadísticamente significativas en retardo del parto contra otros uteroinhibidores. Mostró menor mortalidad neonatal que la indometacina (RR = 0,21; IC 95 %: 0,05 a 0,92), y menor proporción de eventos adversos maternos totales que el fenoterol (RR = 0,16; IC 95 %: 0,08 a 0,31), el nifedipino (RR = 0,48; IC 95 %: 0,3 a 0,78) y la terbutalina (RR = 0,44; IC 95 %: 0,28 a 0,71).Conclusiones: atosiban tiene una eficacia similar para retardar el parto ante la amenaza de un parto pretérmino con otros comparadores (certeza  moderada), con ventajas frente a indometacina en mortalidad neonatal (certeza baja) y frente a fenoterol, nifedipino y terbutalina en eventos adversos maternos (certeza moderada).

Objectives: To assess the efficacy and safety of atosiban in pregnant women under threat of preterm birth. MethOds: A systematic review was carried out, inclusion criteria were adult pregnant women diagnosed with preterm labor threat with the following characteristics: regular uterine contractions lasting at least 30 seconds, contraction frequency greater than 4 every 30 minutes, cervix dilatation from 1 to 3 cm, gestational age from 24 to 33 weeks and normal fetal heart rate (110 -160 bpm). Atosiban was set as intervention and the comparators were nifedipin, indomethacin, terbutaline, fenoterol and placebo. Assessed outcomes were: Gestational age at birth, delay in the timing of birth, neonatal mortality, neonatal respiratory distress syndrome, intraventricular hemorrhage, periventricular leukomalacia, rate of maternal adverse events, rate of neonatal complications. Quality of evidence was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) tool. The comparison measurements found on studies were extracted in order to perform a network meta-analysis. Results were expressed through forest plots. Results: 11 studies were included in network meta-analysis, all of them were randomized controlled clinical trials. The results showed that atosiban has the greater probability of being the most effective alternative on the following outcomes: delay in the timing of birth (0.81), neonatal mortality (0.85) and rate of maternal adverse events (0.99). cOnclusiOns: Atosiban showed similar efficacy for delaying labor in women with preterm labor threat with comparators. However, it showed a greater safety profile with respect to maternal adverse events and a similar level compared to nifedipine for neonatal complications.