TABLE 18.

Antibiotics for Preterm Labor

SourceLocation and Type of TrialInterventionMaternal OutcomePerinatal/Neonatal Outcome
Oyarzun et al369Chile; urban hospital setting; DBRPCTWomen in labor between 22–36 wk gestation (n = 173) were given either amoxicillin and erythromycin for 6–7 d, plus tocolytic therapy (n = 83) or placebo (n = 90).No significant differences between the groups were found for maternal outcomes, including duration of randomization-to-delivery interval and frequency of clinical chorioamnionitis and endometritis. Rate of cesarean section was significantly higher in the placebo group (28% vs 12%).No significant differences in neonatal outcomes were detected between the groups, including preterm birth rate, neonatal death, respiratory distress syndrome, proven sepsis, and birth weight. Suspected sepsis was significantly more frequent in the placebo group (6/90 vs 0/78).
Norman et al356South Africa; multicenter trial, urban slum setting; RCTThe study recruited 81 pregnant women. The study group (n = 43) received ampicillin and metronidazole for 5 d. The control group (n = 38) received no antibiotics. In all women, contractions were suppressed with hexoprenaline and indomethacin for 24 hours, and betamethasone was given for fetal lung maturity.In those receiving ampicillin and metronidazole, pregnancy was significantly prolonged (median 15 d vs 2.5 d, P = .04) with significantly more women still pregnant after 7 d (63% vs 37%, P = .03, OR: 0.34; CI: 0.13–0.94).Significantly fewer infants in the treatment group developed necrotizing enterocolitis than in the control group (0 vs 5, P = .02).
Kenyon et al368 (ORACLE II trial)Multicountry, multicenter trial; urban hospital setting; RPCT6241 women in suspected or confirmed preterm labor (<37 wk gestation) were randomized to receive 1 of 4 possible treatments: 1) co-amoxiclav (325 mg) plus erythromycin (250 mg); 2) co-amoxiclav plus erythromycin placebo; 3) erythromycin plus co-amoxiclav placebo; or 4) co-amoxiclav placebo plus erythromycin placebo and followed for pregnancy outcomes.Antibiotic use did not prolong pregnancy, as most women did not deliver within 48 h (89.9%) or 7 d (84.6%).The only significant difference between the antibiotic and placebo groups was that the rate of oxygen dependence at 36 wk post-conception was higher with erythromycin use than with no erythromycin (P = .04). Importantly, the other measure of chronic lung disease (oxygen dependence at >28 d of age) did not show this difference. There were no significant differences between the babies with respect to birth weight or need for intensive or special care. Co-amoxiclav was associated with a (non-significant) doubling of risk for suspected or proven necrotizing enterocolitis. In the subgroup of infants born at <32 wk gestation, there were no significant differences in neonatal outcomes (respiratory distress syndrome, oxygen dependence, sepsis, necrotizing enterocolitis, abnormal cerebral scan, or death) between antibiotics and placebo.