Validation of a Model for Predicting Magnesium Concentration in Women with Preeclampsia: A Retrospective Cohort Study

Abstract

Objective. To validate a model for predicting magnesium concentration in magnesium sulfate treatment in preeclampsia. Design. Retrospective cohort study. Setting. Three secondary care hospitals, one accepting neonates from gestational week $28+0$. Population. Women with preeclampsia undergoing magnesium sulfate treatment. Subjects initially received Zuspan treatment (4 g bolus and 1 g/h maintenance dose), commonly increased by individual titration. Main Outcome Measures. Difference in mean between measured and predicted magnesium concentration. Proportion of women reaching target concentration (>2 mM) in 25 h. Results. 56 women were included, with 356 magnesium measurements available. Mean magnesium concentration was 1.82 mM. The prediction model overestimated magnesium concentration by 0.10 mM (CI 0.04–0.16) but exhibited no bias for weight, creatinine, or treatment duration. Weighted mean infusion rate was 1.22 g/h during 30 hours. Overall success rate in reaching target concentration was 54%, decreasing to 40% in $\text{women}>95$ kg. Overall success rate at 8 hours was 11%. No toxic concentrations were found. Conclusions. Zuspan regimen is very safe, but slow to reach therapeutic concentrations—despite efforts of individual titration. Success rate is lower in heavy women, which is of particular importance considering their predisposition to develop preeclampsia. The validated pharmacokinetic model performs well and may be used to individually tailor treatment from the outset.