Pilot study of intravenous magnesium sulfate in refractory cardiac arrest: safety data and recommendations for future...

Recent case reports have evidenced a temporal association between administration of i.v. magnesium sulfate (M) and resuscitation from prolonged cardiac arrest refractory to standard (S) ACLS attempts. However, speculation has arisen that M as a vasodilator, may decrease aortic diastolic and coronary perfusion pressure (CPP), aortic systolic and cerebral perfusion pressures (CePP), which may decrease resuscitation rates and neurologic recovery, as compared to standard ACLS alone (SA).

To resolve positive beginning evidence vs. negative theory, we conducted a pilot study of M+S vs. SA in refractory cardiac arrest on resuscitation rates (% R, return of stable pulses > 30 min without CPR, first in-hospital cardiac arrest > 5-min duration) and neurologic recovery/survival to hospital discharge (SHD).

All patients from 1 January 1990-31 December 1991 at Rose Hospital, in cardiac arrest refractory to S through the first epinephrine dose (including 3 defibrillation attempts with pulseless VT/VF) were included in the data analysis, except: (1) patients with trauma, known poisoning, < 18 years, pregnancy excluded; (2) Standard ACLS alone patients with cardiac arrest < 5-min duration were not included in the SA comparison group, because the shortest cardiac arrest time before i.v. MgSO4 administration in the M+S group was 5 min. M+S (N = 29) and SA (N = 33) groups were also comparable on mean age (72-73 years) in this open-label prospective case-matched control group study.

SHD rates were nearly equivalent between M+S (5.2%) and SA (4.5%). Complete or partial neurologic recovery, as best neurologic status post-R, occurred in 21% (6/29) M+S patients vs. 9% (3/33) SA (P = 0.17), even though cardiac arrest time on the study code call for resuscitated patients averaged shorter with SA (14.2 min) than M+S (19.8 min). M was frequently administered late in the resuscitation attempt--code call to M administration averaged 16.5 min (< 10 min in only 4/28 patients). A trend toward increased % R with M was evidenced: 21% (7/33) SA vs. 35% (10/29) M+S (P = 0.21). A temporal association between M administration and first return of spontaneous circulation (ROSC) was also documented in 4 of 10 M+S patients (pulseless electrical activity (3)/pulseless VT (1)), who had first ROSC/R occur within 0.5-2.25 min following first i.v. M bolus delivery, after 11-30 min (mean = 20 min) of continuous pulseless rhythm refractory to standard ACLS. All M+S resuscitations occurred within the dose range 2.5-5 g (i.v. push): 3/6 (50%) and 7/13 (54%) R occurred with 1-3 g and 4-5 g MgSO4, respectively (at least 11/13 patients had peripheral i.v. delivery with 4-5 g M). Analyzing post-ROSC hypotension proved important, as 50% of pts with first recorded systolic BP post-ROSC < 90 mmHg were resuscitated vs. 83% with > 90 mmHg (P = 0.10). A trend toward increased post-ROSC hypotension was evidenced with i.v. MgSO4: Recorded first or second systolic BP < 90 mmHg post-ROSC occurred in 66% of M+S vs. 42% of SA patients. All 3 M+S patients having a wide open i.v. levophed infusion as vasopressor support, started immediately post-ROSC/i.v. MgSO4 with systolic BP < 90 mmHg and continued at least 15 min (titrating to a systolic BP approximately 110 mmHg), had a temporal association between M delivery and R after 14-30 min of continuous pulselessness refractory to S.

Human research determining whether i.v. MgSO4 increases long-term survival from refractory cardiac arrest should be vigorously pursued, as it is safe to proceed given the above described considerations.