Magnesium Sulfate as an Analgesic and Sedative-sparing Agent in Patients under Mechanical Ventilation: A Randomized, Double Blind Clinical Trial

Abstract

Objectives:

The efficacy of intravenous magnesium sulfate in reducing sedative and analgesic requirements was examined in patients under mechanical ventilation.

Methods:

In this randomized clinical trial, patients requiring mechanical ventilation received either magnesium sulfate (serum magnesium target level: near 3 mg/dL) or placebo concurrently with the hospital analgosedation protocol. The calculated dose of magnesium in the intervention group was infused over 24 h. The primary outcome of the study was the cumulative opioid requirement at 24 and 48 h after intubation.

Findings:

A total of 110 patients were randomly assigned to either the placebo (n = 54) or the magnesium sulfate (n = 56) group. The results showed that the cumulative opioid consumption in the magnesium group was significantly less than in the control group at 24 h postintervention (120 mg vs. 214 mg, P < 0.001). This difference remained significant at 48 h postintervention (240 mg vs. 433 mg, P < 0.001). In addition, in the adjusted analysis using analysis of covariance, after adjustment for the model for acetaminophen dose and additive sedatives as confounders, this still remained statistically significant (P < 0.001).

Conclusion:

Administration of magnesium sulfate targeting a serum magnesium level near 3 mg/dL in patients under mechanical ventilation reduced opioid requirements and may be considered as an analgesic and sedative-sparing agent.

INTRODUCTION

Analgosedation is an approach based on the priority of pain control to achieve the desired level of sedation. Analgesia and sedation are required for most patients undergoing mechanical ventilation.[1] Current agents used in intensive care units (ICUs) are associated with significant adverse effects such as prolonged mechanical ventilation, delirium, impaired cognitive function, hemodynamic instability, prolonged hospitalization, increased costs, and higher mortality rates.[2,3]

Management and tailoring of analgesic and sedative regimens during mechanical ventilation is a complex issue. There are various protocols, which make the issue even more controversial.[4,5] The sedation requirements for patients following trauma, surgery, or medical illness and those with or without a history of opioid use disorder vary significantly; thus, a “one-size-fits-all” approach is not suitable for sedation in all settings.[6]

Magnesium acts as a second messenger and as a coenzyme for hundreds of cellular processes.[7] Magnesium has been extensively used in clinical practice for the management of cardiac dysrhythmia, eclampsia, preeclampsia, and severe asthma.[8,9,10] Magnesium has shown analgesic and sedative effects through antagonizing N-methyl-D-aspartate (NMDA) receptors and reducing the release of catecholamines from neurons.[11]

Few studies have investigated the efficacy of magnesium in the sedation of mechanically ventilated patients. However, the optimal dose, preferred method of administration, and duration of magnesium therapy are not defined. Furthermore, its safety in this setting remains unclear. To help cover some of these gaps in critical care medicine, the efficacy of intravenous magnesium sulfate in reducing sedative and analgesic requirements was examined in patients under mechanical ventilation.

METHODS

This randomized controlled trial was designed as a double-blind, placebo-controlled, randomized clinical trial. The study was conducted in a Tertiary Care Teaching Hospital (Imam Khomeini Hospital Complex), affiliated with Tehran University of Medical Science, Tehran, Iran.

It was performed from October 2024 to November 2025. The Ethics Committee of Tehran University of Medical Sciences approved the study (IR.TUMS.TIPS.REC.1403.143). Furthermore, it was registered as a clinical trial (register ID: IRCT20100228003449N33). Written informed consent was obtained from the patient’s first-degree family member after describing the study protocol.

Patients aged 18 years or older were included if they were expected to require invasive mechanical ventilation for at least 48 h following ICU admission. Considering the pharmacokinetic properties of propofol and fentanyl, which are often used for the intubation procedure at our center, patients were enrolled in the study 2 h after intubation, to eliminate the effects of the agents used for intubation.

Patients with any acute neurological disorder (including traumatic brain injury, stroke, intracranial hemorrhage, meningitis, seizure disorders, or neuromuscular paralysis), estimated glomerular filtration rate <30 mL/min/1.73 m², anuria or oliguria, active bleeding, history of cardiac conduction abnormalities, baseline bradyarrhythmia, or baseline serum magnesium level >2.5 mg/dL were not eligible for inclusion. Patients who experienced any possible adverse effects related to magnesium during the study were excluded.

Using the Randomization.com Website, patients were assigned to two groups in a 1:1 ratio, either the intervention group or the control group [Figure 1]. The permuted block randomization table was generated electronically in block sizes of four.

Figure 1.

Flow diagram of the study. RRT = Renal replacement therapy, ESRD = End-stage renal disease, AKI = Acute kidney injury

Patients were assigned to receive either a placebo or intravenous magnesium concurrently with the sedation protocol of the study. Intravenous magnesium was prepared by a nurse not involved in outcome assessment, using identical containers labeled with code A or B. Assessment of analgesia and sedation was performed by an intensivist unaware of group allocation.

Both groups received a local sedation protocol [Supplementary Figure 1 ^{(1.5MB, tif)} ].[12,13] Patients in the intervention group received magnesium sulfate as an intravenous infusion. The target serum magnesium level in this group was set at 3 mg/dL. Magnesium sulfate dosage was calculated based on the difference between the target serum magnesium level (3 mg/dL) and the baseline level, using a rule-of-thumb formula. Based on this, each gram of magnesium sulfate was estimated to increase serum magnesium concentration by approximately 0.18 mg/dL.[11] One-third of the calculated magnesium dose was administered as an intravenous short infusion over 30 min, and the remaining two-thirds were infused continuously over 24 h. Each gram of magnesium sulfate was diluted in at least 50 mL of normal saline. Serum magnesium levels were measured twice daily in all patients.

In the control group, patients received isotonic saline with a similar volume as placebo. If hypomagnesemia occurred during the study, magnesium sulfate was administered for correction to achieve a target serum magnesium level of at least 1.7 mg/dL in this group.

The primary outcome of the study was the cumulative opioid requirement at 24 and 48 h. All opioid medications were converted to intravenous morphine milligram equivalents.[13]

The main secondary outcomes were the amount of requirement for other analgesics and sedatives, and the incidence of adverse events related to magnesium administration. Other secondary endpoints were time, duration of ICU stay, and agitation and over-sedation episodes during the study.

Patients were under close observation and continuously monitored during the study period. Demographic characteristics (age, sex, height, and weight), underlying comorbidities, drug and substance use history, and baseline vital signs and laboratory data were recorded. In addition, vital signs, hemodynamic parameters, and all medical and nonmedical interventions were collected during the study period.

Depth of sedation was assessed according to the Richmond Agitation Sedation Scale (RASS) score.[14] The sedation target for each patient was determined by an intensivist before the start of the study. The sedation status of each patient was assessed twice daily, and the doses of analgesic and sedative medications were adjusted to achieve the target RASS score. Finally, the RASS compliance for each patient was calculated. Compliance was calculated as the proportion of patients who reached the target RASS score.[12] The prescribed dose of all opioids was recorded. Finally, all opioids were converted to intravenous morphine equivalent doses.[13,15]

During the study, episodes of agitation or over-sedation were recorded. The Behavior Pain Scale (BPS) was used to assess patients’ pain. Patients were evaluated twice daily for this purpose, and if the score was >6, which reflects moderate to severe pain,[16] intravenous acetaminophen was administered as a rescue analgesic per protocol.

Considering the lack of a similar study to assess the effect size, based on the primary outcome of the study, the sample size was calculated by assuming an effect size of 0.5, a type one error of 0.05, and 80% power. With G*Power 3.1.lik (Heinrich Heine University Düsseldorf, Düsseldorf, Germany), it was estimated as 51 patients for each group.

All statistical analyses were performed using IBM SPSS Statistics for Windows, Version 27.0 (Armonk, NY, USA: IBM Corp). A two-tailed P < 0.05 was considered statistically significant.

The distribution for all continuous variables was assessed using the Shapiro–Wilk test. Descriptive statistics for normally distributed continuous variables were presented as mean ± standard deviation, whereas nonnormally distributed (skewed) variables were summarized using the median and interquartile range. Categorical variables were reported as frequencies (n) and percentages (%).

For bivariate comparisons between the two study groups, the independent samples t-test was employed for continuous variables. For variables that did not follow a normal distribution or violated the assumption of equal variances, the nonparametric Mann–Whitney U-test was used.

To assess the potential effects of confounding variables, analysis of covariance (ANCOVA) was performed.

For the analysis of repeated measures data, different longitudinal models were applied based on the nature of the data. For normally distributed dependent variables measured over time across the two groups, linear mixed models were applied to account for intrasubject correlations. In cases that the assumption of normality was not met, generalized estimating equations (GEE) were applied to model the longitudinal outcomes. The GEE models were implemented via SPSS syntax to ensure appropriate specification of the model’s correlation structure and distribution.

RESULTS

A total of 110 patients were enrolled in the study and randomly assigned to either the placebo (n = 54) or the magnesium sulfate (n = 56) group. The baseline demographic, clinical, and laboratory characteristics of the participants are summarized in Table 1.

Table 1.

Comparison of patients’ baseline characteristics

Characteristic	Group		P b
	Placebo (n=54)	Magnesium (n=56)
Malea	35 (64.8%)	33 (58.9%)	0.52
Female	19 (35.2%)	23 (41.1%)
Age (year)	59.00 (38.00–74.00)	63.50 (45.50–71.00)	0.80
Weight (kg)	77.00 (70.00–85.00)	72.00 (68.50–84.50)	0.61
Height (m)	172.67 (7.37)	173.05 (7.00)	0.78
BMI (kg/m2)	25.75 (23.90–27.40)	24.65 (23.05–26.75)	0.35
Diagnosis*
Postabdominal surgery	40 (74.1)	42 (75.0)	0.91
Medical	14 (25.9)	14 (25.0)
GCS†
GCS: 14,15	43 (79.6)	42 (75.0)	0.56
GCS: 13,12,11	11 (20.4)	14 (25.0)
Substance user
Negative	48 (88.9)	46 (82.1)	0.60
Opioids	5 (9.3)	8 (14.3)
Amphetamine	1 (1.9)	2 (3.6)
Opioid dose (IVME)‡	10.26 (6.85)	15.87 (8.52)	0.24
SOFA score	7.00 (5.00–9.00)	6.00 (4.00–7.00)	0.13
APACHII II	15.85 (6.09)	15.84 (4.01)	0.69
WBC (109/L)	11.95 (8.10–18.00)	13.55 (9.65–17.70)	0.48
NLR (%)	12.21 (7.17–18.20)	10.31 (7.14–17.50)	0.55
Hgb (g/dL)	9.91 (1.96)	10.26 (1.89)	0.34
Platelet (109/L)	184.00 (104.00–279.00)	200.00 (144.00–297.00)	0.22
Baseline serum magnesium	2.30 (2.00–2.40)	2.00 (1.90–2.30)	0.99
Urea, (mg/dL)	54.00 (39.00–96.00)	45.00 (33.00–81.00)	0.23
Creatinine (mg/dL)	1.20 (1.00–1.60)	1.00 (0.80–1.50)	0.05
AST (units/L)	33.00 (23.00–47.00)	33.00 (24.00–65.00)	0.41
ALT (units/L)	25.00 (15.00–34.00)	25.50 (15.50–36.00)	0.53
ALP (units/L)	229.50 (159.00–326.00)	231.00 (164.00–297.00)	0.88
INR	1.29 (1.08–1.53)	1.20 (1.05–1.37)	0.27
PTT (s)	33.50 (28.00–39.00)	33.00 (28.50–40.00)	0.70
Furosemide dose (mg)	60.0 (40.0–90.0)	40.0 (40.0–60.0)	0.183

*Reason for needing care in the ICU, ^†Patients with a score of<11 or fluctuating GCS were not eligible, ^‡Just for whom had history of dependency, ^aNumber (%) or Median (Q1–Q3), ^bMann–Whitney test. Independent samples t-test, Fisher’s exact test, Pearson’s Chi-squared test, P<0.05 considered significant. BMI=Body mass index, GCS=Glasgow Coma Scale, IVME=IV morphine equivalent, APACHE=Acute physiology and chronic health evaluation, ED=Emergency department, OR=Operating room, ICU=Intensive care unit, SOFA=Sequential organ failure assessment, NLR=Neutrophil lymphocyte ratio, AST=Aspartate aminotransferase, ALT=Alanine aminotransferase, ALP=Alkaline phosphatase, PTT=Partial thromboplastin time, WBC=White blood cell, HgB=Hemoglobin

There was no statistically significant difference between the two groups regarding age, gender, weight, body mass index, primary diagnosis, intubation site, baseline Glasgow Coma Scale score, or history of opioid use disorder. Furthermore, the severity of illness at the time of admission, assessed by Sequential Organ Failure Assessment and Acute Physiology and Chronic Health Evaluation-II scores, showed no significant disparity between the groups.

Baseline laboratory data, including complete blood count, inflammatory markers, renal and hepatic function tests, and coagulation parameters, were comparable between the two groups [Table 1].

Table 2 shows a comparison of the desired RASS scores of patients in the two groups. The Richmond scale for each patient was compared to the desired goal of the patient. At 6, 24, 30, and 48 h of the study, the RASS compliance was comparable between groups [Table 3].

Table 2.

Comparison of the desired Richmond Agitation Sedation Scale goal between groups

Desired Richmond Scale (goal)	Group		P
	Placebo (n=54), n (%)	Magnesium (n=56), n (%)
−4	9 (16.7)	17 (30.4)	0.03
−3	21 (38.9)	10 (17.9)
−2	23 (42.6)	29 (51.8)
−1	1 (1.9)	0

Fisher’s exact test, P<0.05 considered significant. “Intervention group=Magnesium sulfate, Control group=Normal saline as placebo”. “Score range, −5 (unresponsive) to 4 (combative)”

Table 3.

Comparison of Richmond Scale compliance at different time intervals between the groups

Richmond Scale compliance (h)	Group		P
	Placebo (n=54), n (%)	Magnesium (n=56), n (%)
6	19 (35.2)	21 (37.5)	0.80
24	28 (51.9)	32 (57.1)	0.57
30	34 (63.0)	38 (69.1)	0.49
48	44 (81.5)	45 (81.8)	0.96

Chi-squared test, P<0.05 considered significant. “Intervention group=Magnesium sulfate, Control group=Normal saline as placebo.” “Score range, −5 (unresponsive) to 4 (combative)”

The GEEs model showed that the effect of intervention on the BPS score was significant (P < 0.001), so the mean pain score in the two groups was significantly different. The effect of time was also reported to be significant (P = 0.04), indicating changes in pain score over the measurement time. However, the interaction of group and time was not statistically significant (P = 0.07), indicating that the pattern of pain changes over time in the two study groups was similar [Supplementary Tables 1a-1d].

Supplementary Table 1a.

Descriptive statistics for the Behavior Pain Scale

Characteristic	Group
	Placebo (n=54)a	Intervention (n=56)a
Behavior Pain Scale
1st assessment	7 (6–8)	6 (5–7)
2nd assessment	7 (6–8)	7 (5–8)
3rd assessment	7 (6–8)	6 (5–7)
4th assessment	7 (7–8)	6 (5–7)

^aMedian (Q1–Q3). Mean±SD. A score >6 reflects moderate to severe pain. SD=Standard deviation

Supplementary Table 1d.

Comparison of Behavior Pain Scale between measured times

I-time	J-time	Mean difference (I−J)	P	95% Wald CI for difference
				Lower	Upper
1	2	−0.30a	0.022	−0.55	−0.04
	3	−0.23	0.130	−0.52	0.07
	4	−0.37a	0.017	−0.67	−0.07
2	1	0.30a	0.022	0.04	0.55
	3	0.07	0.641	−0.22	0.36
	4	−0.07	0.645	−0.38	0.24
3	1	0.23	0.130	−0.07	0.52
	2	−0.07	0.641	−0.36	0.22
	4	−0.14	0.246	−0.38	0.10

^aThe mean difference is significant at the 0.05 level. P<0.05 considered significant. It was also found that the BPS at the first time point was significantly different compared with the second time point (mean difference=0.30, P=0.02) and the fourth time point (mean difference=0.37, P=0.02), whereas the differences between the other measurement time points were not statistically significant (P>0.05). BPS=Behavioral Pain Scale, CI=Confidence interval

Supplementary Table 1c.

Between-group comparisons for Behavior Pain Scale

Pairwise comparisons
I-group		Mean difference (I−J)		P		95% Wald CI for difference
						Lower		Upper
Placebo; intervention		0.70a		<0.001		0.30		1.10

^aThe mean difference is significant at the 0.05 level. P<0.05 considered significant. Given the significant interaction between group and time, Simple main effects analysis and pairwise comparisons were reported. In between-group comparisons, the mean score of the BPS was significantly higher in the placebo group than in the magnesium sulfate group (mean difference=0.70; 95% CI: 0.30–1.10, P<0.001). BPS=Behavioral Pain Scale, CI=Confidence interval

The mean administered dose of magnesium sulfate in the intervention group was 6 g. The results showed that there was no significant difference between the two groups at the baseline (P = 0.99). However, from the second to the fifth time point, serum magnesium levels in the intervention group were significantly higher than those in the placebo group (P < 0.001 for all comparisons) [Figure 2].

Figure 2.

Changes in the serum magnesium levels during the study period

The results showed that the cumulative opioid consumption in the magnesium group was significantly less than in the control group at 24 h postintervention (120 mg vs. 214 mg, P < 0.001). This difference remained significant at 48 h postintervention (240 mg vs. 433 mg, P < 0.001) [Figure 3]. In addition, in the adjusted analysis using ANCOVA, after adjustment for the model for acetaminophen dose and additive sedatives as confounders, this still remained statistically significant (P < 0.001) [Table 4].

Figure 3.

Cumulative opioid requirement at 24 and 48 h. The stars in the figure indicate statistical significance levels for the comparisons shown

Table 4.

Primary outcome

Parameter	Group		P b	P c
	Placebo (n=54)	Magnesium (n=56)
Cumulative opioid* 24 h (mg)a	214.00 (124.57–220.0)	120.00 (110.00–127.02)	<0.001	0.001
Cumulative opioid* 48 h (mg)	433.00 (246.66–463.33)	240.00 (223.13–313.0)	<0.001	<0.001
Diff- cumulative opioid	225.0 (120.0–244.3)	120.0 (110.0–127.6)	<0.001
P d	<0.001	<0.001

*Based on IV morphine equivalent. ^aMedian (Q1–Q3), ^bMann–Whitney test, ^cANCOVA test (adjusted for acetaminophen dose and sedative agent), ^dWilcoxon test. P<0.05 considered significant, Independent samples t-test. Mann–Whitney U-test was used to compare between two groups. Intragroup changes between 24 and 48 h were examined using the Wilcoxon test. Also, ANCOVA model was used to compare values by adjusting for the effect of the acetaminophen dose and additive agent

Regarding the other sedative (midazolam) requirement, no significant difference was observed between the two groups. Midazolam supplementation was required for 16.7% of patients in the placebo group and 21.4% of patients in the intervention group. Furthermore, among patients who received midazolam, there was no significant difference in the cumulative dose of midazolam between the two groups [Table 5].

Table 5.

Secondary outcomes

Characteristic	Groups		P b
	Placebo (n=54)	Magnesium (n=56)
Acetaminophen dose1 (g)	4.00 (3.00–4.00)	2.00 (2.00–4.00)	<0.001
Midazolam requirement
No	45 (83.3)	44 (78.6)	0.52
Yes	9 (16.7)	12 (21.4)
Midazolam cumulative dose* (mg)	84.0 (43.0–92.50)	44.0 (30.25–100.0)	0.49
Additive sedative
None	43 (79.6)	50 (89.3)	0.07
Dexmedetomidine	9 (16.7)	4 (7.1)
Propofol	0	2 (3.6)
Ketamine	2 (3.7)	0
ADR (safety profile)
Bradycardia (HR<50)	0	1 (1.8)	0.97
Hypotension (MAP<65 mmHg)	5 (9.3)	4 (7.1)	0.16
Agitation episodes†
Zero episode	13 (24.1)	28 (50.0)	0.01
One episode	18 (33.3)	15 (26.8)
Two episodes	8 (14.8)	10 (17.9)
Three episodes	10 (18.5)	3 (5.4)
Four episodes	3 (5.6)	0
Five episodes	1 (1.9)	0
Six episodes	1 (1.9)	0
Over-sedation episodes†
None	35 (64.8)	22 (39.3)	0.02
One episode	12 (22.2)	13 (23.2)
Two episodes	5 (9.3)	16 (28.6)
Three episodes	1 (1.9)	4 (7.1)
Four episodes	1 (1.9)	1 (1.8)
ICU stay duration (days)	11.50 (4.00–20.00)	12.50 (5.00–28.00)	0.58
Reintubation
No	46 (85.2)	53 (94.6)	0.12
Yes	8 (14.8)	3 (5.4)

*Just in whom had received midazolam, ^†Based on the comparison of each patient’s target RASS, ^aMedian (Q1–Q3) or mean (SD) or n (%), ^bMann–Whitney test. Independent samples t-test, Fisher’s exact test. Pearson’s Chi-squared test. P<0.05 is considered significant. Quantitative data are presented as median and IQR (Median [Q1–Q3]). Qualitative data are presented as a number (percentage). The Mann–Whitney U-test was used to compare quantitative variables between the two groups. Qualitative variables were compared using the Chi-square test or Fisher’s exact test. ADR=Adverse drug reaction, HR=Heart rate, MAP=Mean arterial pressure, SD=Standard deviation, ICU=Intensive care units, RASS=Richmond Agitation Sedation Scale, IQR=Interquartile range

The results indicated that the dose of acetaminophen administered in the placebo group was significantly higher than that in the magnesium sulfate group (4 g vs. 2 g; P < 0.001).

As shown in Table 5, the requirement for additive sedative agents, including dexmedetomidine, propofol, and ketamine, did not differ significantly between the groups (P = 0.07). However, they used slightly more in the placebo group. However, its impact on the primary outcome was adjusted after the final statistical analysis.

Supplementary Table 1b.

Comparison of the trend of changes in the Behavior Pain Scale over time between the two study groups using generalized estimating equations

Source	Wald Chi-square	df	P
Intercept	154.052	1	<0.001
Group	14.901	1	<0.001
Time	8.274	3	0.040
Group × time	7.131	3	0.070

Dependent variable: Behavior pain scale, Model: (Intercept), group, time, group×time, P<0.05 considered significant. Data analysis was performed using the GEE model. Fixed effects, including group, time, and the interaction group×time, were entered into the model. GEE: Generalized estimating equations

Regarding the occurrence of the adverse events, cases of hypermagnesemia (>3 mg/dL) were reported significantly more frequently in the intervention than the placebo group (39.3% vs. 1.9%). However, the incidence of hypomagnesemia, bradycardia, and hypotension did not differ significantly between the two groups.

The incidence of agitation episodes differed significantly between the two groups (P = 0.01). Patients in the intervention group were more frequently classified as having no agitation episodes compared with the placebo group (50% vs. 24.1%). However, multiple agitation episodes were observed more commonly in the placebo group. Furthermore, the analysis of over-sedation episodes revealed a significant difference between the groups (P = 0.02). Patients in the intervention group experienced a greater number of oversedation episodes than the placebo group.

No significant difference was observed between the two groups in other clinical outcomes as well. Specifically, the ICU stay duration and reintubation rate did not differ significantly between the groups.

DISCUSSION

In this study, administration of intravenous magnesium sulfate as continuous infusion for a serum target level of 3 mg/dL reduced total opioid dose requirement in mechanically ventilated patients without any considerable safety issues. Optimizing serum magnesium levels showed a sparing effect regarding sedative medications, especially opioids, in critically ill patients under mechanical ventilation. Intravenous magnesium sulfate is a potential option to reduce the adverse effects of sedative and analgesic regimens.

Fentanyl and other opioids are the backbone of sedation regimens in patients requiring mechanical ventilation.[17] Magnesium has shown an effect in reducing postoperative analgesic requirements and pain intensity,[18] postprocedural pain,[19,20] nerve block,[21,22] and anesthesia during surgeries.[23,24] However, sedation required for mechanical ventilation is a more complex intervention than short-term anesthesia. Reducing opioid requirements in this population prevents drug-related side effects, costs, and drug interactions.

It was not the first time that magnesium had been used in a sedation regimen. Magnesium decreased sufentanil requirement in mechanically ventilated patients in the Memiş et al. trial.[25] However, that was a small study with a duration of only 6 h in 30 participants. Results of another study revealed that the addition of magnesium to the sedation protocol reduced midazolam requirement.[26] The duration of intervention was 48 h, but the study used midazolam as the main sedative, which is not recommended now.[27]

Magnesium could play this role through various pathways. As it exerts anticonvulsant effects by blocking NMDA receptors,[28] magnesium likely exerts its analgesic effects through the same mechanism. Antispasmodic effects of magnesium have been reported in eclampsia, in patients with tetanus, and through various inflammatory pathways, including those involved in pain.[29,30] Magnesium also modulates free radicals as well as the nuclear factor kappa-B pathway.[31] In addition, antiadrenergic effects and modulation of peripheral pain terminals via calcium channel blockade have also been suggested for magnesium.[31,32,33] Consistent with a previous study,[32] magnesium reduced acetaminophen requirement in the present study as well.

Various doses of magnesium have been used.[25,26] In the present study, the magnesium sulfate dose was calculated based on a simple rule. The mean prescribed dose in patients was 6 g within 24 h. Although much higher doses (even up to 40 g/24 h) have been used for seizure prophylaxis in eclampsia,[8] the results of the present study suggest that much lower doses as low as 70–80 mg/kg/day and target serum magnesium levels <3 mg/dL are sufficient to achieve analgesic effects in mechanically ventilated patients. However, given the limitations of measuring serum magnesium levels and considering that magnesium is an intracellular ion, caution must be exercised in this regard, and future studies with larger sample sizes should be conducted.

Selecting an appropriate sedation assessment tool is a challenging issue in critically ill patients. Assessment of patients with subjective tools is a limitation in some studies.[2,33] We used RASS, like landmark trials such as A2B[34] and MIDEX and PRODEX trial,[35] which is the most valid tool.[26]

In the current trial, the RASS compliance scale, just like Phan et al.’s study, was implemented.[12] Need for sedation and analgesia varies significantly between patients, and it appears to be a dynamic process in critically ill patients. We only assessed patients’ sedation level twice a day, which does not reflect patients’ needs well.

In the A2B trial,[34] the target RASS score of patients was in the range of −2–+1, which was considered (restless) to (light sedation). However, during the study, medical staffs were permitted to consider deeper sedation to the level of −5 (unresponsive), which achieved by administering propofol.

In our patients, the desired RASS goal ranged from −1 (drowsy) to −4 (deep sedation) and showed a significant difference between the two groups, that may explained by the practitioners’ different approaches. In fact, some of our patients required deep sedation, although trauma patients were not included in the study. To achieve the required deeper sedation, in addition to continuous intravenous fentanyl, patients received other analgesics and sedatives. Opioid analgesics were used instead of propofol as the first-line agent. This emphasizes the fact that analgesia is the first approach for our mechanically ventilated patients, exactly in line with the Society of Critical Care Medicine guideline.[27]

In our study, 12 patients with opioid dependency and three individuals with a history of amphetamine abuse were included. Because their numbers were small, they were not analyzed as a subgroup, and this will be an area for future research in these populations.

Magnesium is an inhibitor of the glutaminergic pathway and NMDA, which plays a role in various areas of addiction.[36] As an adjuvant, it may improve control of withdrawal symptoms from opioids, alcohol, and even amphetamines.[37,38] Adjunctive administration of 3 g of magnesium sulfate followed by 10 mg/kg/h for up to 2 h decreased the Clinical Opiate Withdrawal Scale in patients after buprenorphine-induced withdrawal symptoms.[39] The role of magnesium in the management of withdrawal syndrome or sedation regimens in patients with a history of opioid dependency or substance abuse who require mechanical ventilation could be investigated in future studies. The choice of sedation for this population remains controversial.[38,40]

Magnesium has a high safety profile. Concern is about the prescribed dose and kidney function. The standard biochemical definition of hypermagnesemia in the literature is often cited as >2.7 mg/dL. However, a threshold of 3 mg/dL was applied in our study as its safety profile, patients’ close monitoring, and to simplify bedside calculations. Hypermagnesemia is rare and transient in patients with preserved kidney function.[41] Hypermagnesemia was reported in 22 (39.3%) patients in the intervention group (defined as levels above 3 mg/dL), although levels above 3.8 mg/dL were not recorded in any patient in our study. In addition, patients were under continuous hemodynamic monitoring during the study period.

Hypomagnesemia (defined as serum magnesium <1.7 mg/dL) did not differ between the groups. Magnesium, known as a forgotten electrolyte, and hypomagnesemia are associated with longer length of hospital stay, especially in critically ill and type 2 diabetes patients.[42]

The impact of the level of sedation on outcomes in critically ill patients is a controversial issue, and further studies are needed in the future.[43] Our secondary outcomes, such as duration of ICU stay, did not differ significantly between the groups. Such variables are very complex and influenced by many risk factors in critically ill patients.

The current trial has several strengths. A cheap, available, and safe agent was selected as an adjuvant for sedation in patients under mechanical ventilation support. A distinct coordination between the research team and ICU staff was established, and a sedation protocol based on international guidelines was applied to all patients. Adjustment for confounders, including interfering sedative and analgesic medications, was considered in the modeling and statistical analysis.

There are several limitations to our study. Equianalgesic conversions of opioids represent only estimations and may not fully account for pharmacokinetic variability, altered organ function, opioid tolerance, or nonlinear potency (particularly for methadone and tramadol) in critically ill patients. As an intracellular cation, serum magnesium concentration is not an accurate measure of the body’s magnesium status. In addition, the weaning protocol from mechanical ventilation was not standardized, although this was a secondary outcome. The delirium status of patients was not investigated. The sedation protocol was based on the routine clinical practice in our hospital. A small sample size and some heterogeneity in patient characteristics are other limitations.

The role of magnesium in the management of withdrawal or sedation regimens in patients with a history of opioid dependency or substance abuse who require mechanical ventilation could be investigated in future studies. Furthermore, the efficacy of magnesium sulfate in other populations requiring sedation in the ICU, specifically patients under noninvasive ventilation is uncertain.

CONCLUSION

Administration of magnesium sulfate targeting a serum magnesium level near 3 mg/dL in patients under mechanical ventilation reduced opioid requirements and may be considered as an analgesic and sedative-sparing agent.

AUTHORS’ Contributions

AMY: Data gathering, EM: Data analysis and preparation of the primary draft of the manuscript, HK: Concept and design of the study, Final approval of the manuscript, MR, SFG, and MM: Clinical evaluation of the patients.

Data availability

Available per request from the corresponding author.

Conflicts of interest

There are no conflicts of interest.

Supplementary Figure 1

Study protocol of sedation and analgesia at our center. ICU = Intensive care unit, PRN= ‘pro re nata’

Funding Statement

Nil.