Abstract

AIM To review systematically the evidence to determine whether dexamethasone treatment of very low birthweight infants begun within 14 days of age prevents chronic lung disease (CLD) without clinically significant side effects.

METHODS Randomised controlled trials of dexamethasone started within this time frame were identified through a search of electronic databases, proceedings of scientific meetings, and personal files. Meta-analyses using event rate ratio (ERR), event rate difference (ERD), and if significant, numbers needed to treat (NNT) for benefits and numbers needed to harm (NNH) for adverse effects were calculated. Weighted mean difference were used for continuous variables. Three prespecified subgroup analyses were performed for; (i) dexamethasone begun within 36 hours (hours) of birth; (ii) dexamethasone initiated between 7–14 days of age; or (iii) if surfactant treatment was used.

RESULTS Ten studies were included in the review; six where dexamethasone was initiated within 36 hours of age, four studies for dexamethasone started between 7 and 14 days and six studies using surfactant. Mortality ERR and NNT with 95% confidence intervals for dexamethasone initiated at 7–14 days of age were 0.35 (0.16, 0.74) and 8 (4, 30). ERRs and NNTs for CLD at 28 days and 36 weeks of postmenstrual age were 0.71 (0.61, 0.84), 8 (5, 17), and 0.57 (0.44, 0.76), 10 (6, 23) in the overall analyses. When dexamethasone was started at 7 to 14 days of age ERR and NNT for CLD at 36 weeks were 0.63 (0.47, 0.85) and 3 (2, 9). Clinically significant side effects included increased risk of hypertension, hyperglycaemia, and increased time to regain birthweight.

CONCLUSIONS These meta-analyses show a significant reduction in risk of CLD at 28 days and 36 weeks of postmenstrual age. In the subgroup where dexamethasone was started between 7 and 14 days of age mortality was significantly reduced. Caution is warranted in the routine use of dexamethasone because of lack of data on long term neurodevelopmental outcomes.