Antenatal steroids and neonatal outcome after chorioamnionitis: a meta-analysis

Background: There is debate concerning the safety and efficacy of antenatal steroids in preterm labour with suspected intrauterine infection (chorioamnionitis).

Objectives: We performed a systematic literature review and meta-analysis aimed at evaluating the efficacy and safety of antenatal steroids in clinical and histological chorioamnionitis.

Search strategy: MEDLINE, EMBASE, BioMed Central and the Cochrane databases were searched using the terms 'chorioamnionitis OR intrauterine infection' and '*steroids OR *corticoids'.

Selection criteria: Studies that reported selected neonatal outcome measures in preterm infants with clinical or histological chorio-amnionitis, according to antenatal steroid exposure, were eligible.

Data collection and analysis: Study selection, data extraction and data analysis were performed by two independent investigators. The meta-analysis techniques used included: Mantel-Haenszel analysis; an assessment of study heterogeneity using the Q statistic; and Egger's regression test and funnel plots, to assess publication bias.

Main results: Seven observational studies were included. In histological chorioamnionitis (five studies), antenatal steroids were associated with reduced mortality (OR = 0.45; 95% CI = 0.30-0.68; P = 0.0001), respiratory distress syndrome (OR = 0.53; 95% CI = 0.40-0.71; P < 0.0001), patent ductus arteriosus (OR = 0.56; 95% CI = 0.37-0.85; P = 0.007), intraventricular haemorrhage (IVH; OR = 0.35; 95% CI = 0.18-0.66; P = 0.001) and severe IVH (OR = 0.39; 95% CI = 0.19-0.82; P = 0.01). In clinical chorioamnionitis (four studies), antenatal steroids were associated with reduced severe IVH (OR = 0.29; 95% CI = 0.10-0.89; P = 0.03) and periventricular leucomalacia (OR = 0.35; 95% CI = 0.14-0.85; P = 0.02).

Conclusions: Antenatal steroids may be safe and reduce adverse neonatal outcome after preterm birth associated with chorioamnionitis. There is a need for randomised clinical trials to address this issue.