Participants and study design
A randomized, multicenter, open study was conducted between November 2013 and October 2016 in three teaching maternity hospitals in Panama. These maternity units had experience in the management of severe pre-eclampsia. Verbal and written consent to participate was necessary in all women. The ethical committee of “Complejo Hospitalario Dr. Arnulfo Arias Madrid” approved this research. The other hospitals accepted the research.
Pregnant women diagnosed with severe pre-eclampsia or pre-eclampsia with severe features at more than 20 weeks gestation were invited to participate in the MAG-PP (MAGnesium-Post Partum) study.
Severe pre-eclampsia was diagnosed when a patient had arterial pressures of 140/90 mmHg or higher at least twice separated by 4 h with or without proteinuria ≥0.3 g and had one or more of the following clinical criteria: severe hypertension (160/110 mmHg), neurological symptoms, epigastric pain, or laboratory abnormalities (increase in creatinine, thrombocytopenia, elevation of liver enzymes, changes in clotting times other than HELLP syndrome) [4, 5] When the woman was known to have chronic hypertension and had one of the aforementioned criteria plus the presence of proteinuria ≥0.3 g, it was considered pre-eclampsia with severe features [4, 5].
Exclusion criteria were eclampsia before delivery, HELLP syndrome, epilepsy, and pre-eclampsia with additional pathology such as renal failure, acute pulmonary edema, decompensated diabetes mellitus, autoimmune diseases, or hypertensive encephalopathy.
Eclampsia was defined as the presence of a generalized clonic/tonic seizure when no other cause could be confirmed.
Postpartum hemorrhage was defined as postpartum bleeding greater than 500 ml or post caesarean section bleeding greater than 1000 ml.
Respiratory insufficiency was defined as onset of shortness of breath as demonstrated by tachypnea or dyspnea or an oxygen saturation less than 90%.
The ethics committee and teaching and research authorities of the three hospitals approved the study before commencing recruitment.
Randomization and masking
After obtaining oral and signed consent to participate in the MAG-PP study, patients were assigned to one of two study groups in a 1:1 ratio: A- continue Mg for 24 h after birth (control group), or B- receive Mg after birth for 6 h (experimental group). Randomization was done with a computerized program using blocks of 4, and sealed envelopes were used for the results of the randomization. The coordinator or investigator of each hospital did not have access to the randomization sequence.
All patients were advised of the benefits and risks of continuing postpartum Mg or stopping after 6 h.
Any patient with severe pre-eclampsia or pre-eclampsia with severe features who had received less than 8 h of Mg for seizure prophylaxis before birth was a candidate to enter the study. The research was explained to candidate patients who agreed to be randomized at birth or immediately postpartum.
All patients received the same management and monitoring (the hospital had similar routine), except that group B was only maintained on Mg for 6 h postpartum. The patients who received Mg for 24 h postpartum maintained a bladder catheter and bed rest. The patients who received Mg for 6 h were allowed to ambulate and breastfeed as tolerated after 6 h. The magnesium sulfate was administrated Intravenously by means of an infusion pump in each hospital. When systolic blood pressures of 160 mmHg or more or diastolic blood pressures of 110 mmHg or more occurred, the following was administered according to hospital availability or the preferences of the treating physician: hydralazine 5 mg every 20 min, up to a maximum of 5 doses if needed; labetalol 20 mg, and the dose was doubled if hypertension persisted 20 min later, and if a third, fourth or fifth dose was required, then 80 mg was given; or the third option was nifedipine 10 mg orally every 20 min for up to 5 doses until the blood pressure was lowered. The use of drugs for elevated postpartum blood pressures in patients without hypertensive crisis was permitted at the discretion of the physicians responsible for the patients. Maternal complications, if any, were recorded.
In our hospitals breastfeeding is allowed in patients with cesarean from 6 h postpartum, in patients with magnesium sulfate is allowed after 24 h when magnesium has been removed.
The primary endpoint and variable was seizure (eclampsia) in the first 72 h postpartum. The secondary variables evaluated were complications such as postpartum bleeding, respiratory distress, maternal death, total time on Mg prior to birth, postpartum convulsion time, number of seizures, admission to the intensive care unit due to a complication, use of postpartum antihypertensives, use of antihypertensives for hypertensive crises, time to onset of ambulation, and breastfeeding start time.
There are no previous similar studies, and it is difficult to use an odds ratio (OR) to calculate an adequate sample, so we decided to use the MAGPIE study  in which countries from Latin America participated, which showed that the incidence of eclampsia in the group treated with Mg was 0.8%. Assuming a sample in which the incidence of eclampsia was similar for each group, with a maximum of one patient with eclampsia per group, we could make the comparison. If the experimental group (6 h postpartum Mg) had two patients with eclampsia prior to accruing the desired sample size, the study was to be stopped. Assuming an incidence of seizure of 0.8%, one patient with eclampsia would be present in a sample of 125 patients with severe pre-eclampsia per group. Adding 15% to account for possible protocol alterations or losses, the number of patients should be 286 (143 per group).
The baseline characteristics of the participants in each group were compared with Student’s t test for continuous variables with a normal distribution. Dichotomous variables were compared using Fisher’s exact test and the chi-squared test when appropriate. Values of p less than 0.05 were considered significant.
Statistical analysis was performed with Epi Info software version 7 (Centers for Disease Control and Prevention, Atlanta, GA) under the intent-to-treat principle.