Background
It is important to establish whether research recommendations regarding magnesium sulphate for neuroprotection can be readily translated into clinical practice and achieve the dual objectives of good coverage of the target group, while minimising unnecessary or prolonged exposure to treatment.

Methods
This retrospective cohort study included all women admitted to a tertiary obstetric centre at 23–32 weeks gestation in the first 12 months following implementation of the guideline ‘Magnesium sulphate for the prevention of cerebral palsy’. We determined the number triaged to receive magnesium sulphate, the proportion of infants who received magnesium sulphate prior to delivery and the total number of doses administered.

Results
A total of 330 women were admitted at a mean gestational age of 28.2 weeks, and 132/330 (40%) were prescribed magnesium sulphate, of whom 123/132 (93%) delivered.
142/191 (74%) infants born at <32 weeks' gestation received magnesium sulphate prior to delivery, with no significant differences seen by plurality or gestational age. Of the 145 doses administered, only 13 women received more than one dose, and only nine of 145 (7%) doses proved to be unnecessary. The median treatment duration was 3 h 58 min. The infusion was discontinued as result of side effects in 2% of women. Conclusion Research recommendations regarding administration of magnesium sulphate with neuroprotective intent can be successfully translated into clinical practice. Appropriate triaging of women at high risk of imminent preterm birth is feasible, enabling a high level of magnesium sulphate coverage for infants that deliver prior to 32 weeks gestation, with minimal toxicity and a low rate of unnecessary maternal exposure.