To evaluate the long-term neurological complications following postnatal glucocorticoid treatment in preterm infants a systematic review of randomised controlled trials was performed. A search strategy identified studies which randomised premature infants to dexamethasone treatment to prevent chronic lung disease. To be included in the review, articles had to report on mortality, cerebral palsy and/or developmental delay. Trials were divided according to the start of the therapy in early, moderately early or late treatment analyses. We identified eighteen clinical trials of overall good methodological quality. In the early and late treatment trials a 1.91 (95% ci 1.38-2.64) and 1.31 (95% ci 0.85-2.01) fold increased risk of cerebral palsy was found with a number needed to harm (nnh) of 7 and 20, respectively. In the sub-analyses of the trials with a low contamination rate the nnh was 5 for both equations. In the moderately early trials there was no difference in cerebral palsy rate between the treatment compared to the control group. The relative risks of neurodevelopmental delay was not significant in any of the treatment groups if compared with the controls.
The use of postnatal glucocorticoid therapy should be restricted to patients who cannot be weaned from assisted ventilation of at least fourteen days. The minimum dosage and duration of the therapy in order to facilitate detubation needs still to be determined. More randomised controlled trials with a low rate of contamination are needed to investigate whether glucocorticoids given moderately early can prevent chronic lung disease of the premature without causing long term neurological complications.
