Prophylactic indomethacin: systematic review and meta-analysis.

AIMS To examine the effectiveness of prophylactic intravenous indomethacin in reducing the mortality and morbidity associated with patent ductus arteriosus and intraventricular haemorrhage in infants weighing less than 1750 g at birth. METHODS A literature search from 1980 onwards was made of three databases: Medline; Embase; and the Oxford Database of Perinatal Trials. Using strict criteria applied to randomised controlled trials only, two observers independently selected 14 studies for inclusion in the review. The methodological quality of each study was assessed independently by two observers using explicit criteria. Data on relevant outcome measures were extracted on two separate occasions. Where appropriate, the results of individual trials were combined using meta-analysis techniques to provide a pooled estimate of effect. RESULTS There is a trend towards reduced neonatal mortality in infants receiving prophylactic indomethacin, pooled estimate of risk difference -0.025 (95% confidence interval (CI) -0.061, 0.010). The incidence of symptomatic patent ductus arteriosus is significantly reduced in treated infants, pooled estimate of risk difference -0.217 (95% CI -0.275, -0.160), but there is no evidence that treatment affects respiratory outcomes. Prophylactic indomethacin significantly reduces the incidence of grades 3 and 4 intraventricular haemorrhage in treated infants, pooled estimate of risk difference -0.039 (95% CI -0.066, -0.011). However, there is no sound evidence assessing the long term effect of prophylaxis on neurodevelopmental outcome. Although there is a trend in treated infants towards an increased incidence of necrotising enterocolitis, pooled estimate of risk difference 0.015 (95% CI -0.002, 0.033), and some evidence that treatment may transiently impair renal function, there is no evidence that haemostasis is disturbed. CONCLUSION Prophylactic treatment with indomethacin has several immediate benefits. However, more data are needed on the incidence of possible adverse effects and neurodevelopmental outcomes before routine use of this therapy can be recommended.