Hypomagnesaemia (low magnesium level) is common in hospitalized patients, especially in critically ill patients adjacent to other electrolytes abnormalities [1]. In general, as the magnesium level drops and decelerates more, the hypomagnesemia associated complications becomes more severe and the mandatory for intravenous magnesium sulfate (MgSO₄ IV) is unavoidable [2]. Hypomagnesaemia may cause severe fatal complications if not early diagnosed and promptly treated with possible consequences of neuromuscular and neurologic dysfunctions, heart arrhythmias and alterations in other electrolytes, especially potassium and calcium levels [3]. Data have shown that critically ill patients with magnesium level <2 mg/dL have a significantly higher mortality rate than patients with magnesium level >2 mg/dL [4]. According to the clinical evidence and meta-analysis; hypomagnesaemia associated with prolonge need of mechanical ventilation and ICU stay days in addition to increased mortality rate of 2-3 times compared to normo-magnesemic patients [5,6].

Many causes of hypomagnesaemia, including but not excluded to, gastrointestinal related hypomagnesemia (secretory loss, impaired absorption or reabsorption, acute pancreatitis) and renal loses related hypomagnesemia (alcohol, hypercalcemia, volume expansion, loop or thiazide diuretics, nephrotoxic medications, renal tubular dysfunction, inborn disorders) [7]. Generally, mild-moderate hypomagnesaemia with mild symptoms can be effectively treated with various oral magnesium supplements whereas severe symptomatic hypomagnesemia or the magnesium level drops below 1 mg/dL should be promptly treated with MgSO₄ IV [8]. 

Physiology, renal magnesium excretion rate depends primarily on the extracellular magnesium level available before shifting intracellularly is pursued which is time dependent [9]. In patients with normal to relatively normal kidney function (CrCl >60 mL/min), approximately half of infused magnesium will be renally excreted if the MgSO₄ IV is infused over less than 4 hours [10]. Due to the thick ascending loop of Henle Ca⁺²/Mg⁺² sensing receptors directed plasma magnesium level and the time-dependent intracellularly magnesium shifting, extended MgSO₄ infusion over more than 4 hours has been postulated in many studies to minimize renal loss and subsequently decrease overall magnesium dose requirement for replenishing the intracellular magnesium store, normalization the plasma magnesium level and mitigation the likelihood of hypomagnesemia recurrence once MgS0₄ IV discontinuation [11,12].

Practically, it is difficult to assess patients for hypomagnesemia because of the unreliable relationship between serum and tissue magnesium levels and the serum magnesium level is not always accurate at detecting magnesium deficiency. Patients may appear to be normo-magnesemic based on their serum magnesium level, yet have an underlying magnesium deficiency. Also, normal serum magnesium levels vary by laboratory which ranges between 1.6-2.6 mg/dL [13,14]. Approximately 99% of total body magnesium is distributed among the bones, muscles and soft tissues while the remaining 1% is primarily found in the extracellular fluid. Approximately 60% of the serum magnesium is free ions; 33% is bound to proteins and 7% is complexed with anions [15]. Whatever, the simplest and commonly used test to diagnose hypomagnesemia is still the total serum magnesium level which reflects free magnesium along with complexed and protein bound magnesium [16].

However, a paucity of clinical evidence exists to support extended infusion strategy (Infusion over at least 6-8 hours per MgSO₄ dose) compared to standard infusion strategy (Infusion over 4 hours per MgS0₄ dose). In this study we compared between the two MgSO₄ infusion strategies, Standard Infusion Strategy versus Extended Infusion Strategy, to investigate the biochemical and clinical impacts outcomes across the two strategies in admitted hospitalized patients who suffered from the refractory hypokalemia associated hypomagnesemic critically ill patients.

This is a retrospective study of critically ill patients who experienced a hypokalemic associated hypomagnesemia and received two comparative MgSO₄ IV strategies; Strategy I including the critically ill patients who received the whole replacement MgSO₄ IV dose of 4 g/day once daily for at least 3 days while the comparative strategy, Strategy II including the critically ill patients who received the replacement MgSO₄ IV dose of 4 g MgS0₄ at two divided extended infusion each involved 2 grams MgSO4 infused over at least 6 hours for 3 days.

This study targeted all eligible hypokalemic associated hypomagnesemia who were admitted to our multidisciplinary, Medical and Surgical, Intensive Care Unit (MSICU) in King Hussein Medical Hospital (KHMH) at Royal Medical Services (RMS) between Jan 2018 and May 2021. Retrieved data were retrospectively performed from our Electronic Medical Record System. And this study was approved by the Institutional Review Board (IRB) and the consent forms were waived owing to the study’s retrospective design. This study was finally conducted on 110 refractory hypokalemia associated hypomagnesemic critically ill patients with the inclusion and exclusion criteria that are fully described in Table 1.

The serum magnesium level (Mg⁺²) accompanied with serum potassium level (K⁺) measurements were serially assessed, starting from baseline serum magnesium and serum potassium levels (Mg₁ and K₁, respectively), serum magnesium and potassium levels at the 3rd day of starting MgSO₄ infusion (Mg₂ and K₂, respectively), serum magnesium and potassium levels at least, 6 days after starting MgSO₄ infusion or at least, 3 days after discontinuation MgSO₄ infusion (Mg₃ and K₃, respectively), which one comes first. In our study, we defined refractory hypokalemic associated hypomagnesemia, as it was defined in many studies as “persistent potassium level below 3.5 mEq/l after giving at least 1.5-2 mEq/day” [17]. And we mathematically corrected the Mg levels by adjusting with serum albumin levels, using the following equation “cMg = Mg+0.08*(4-ALB)” [18].

All patients’ continuous variables were analyzed using Independent T-Test to express the results as Mean±SD or as Mean difference±SEM. Also, One Sample T-test was used to express Mean±SD for the whole refractory associated hypomagnesemia parametric studied variables. One the other hand, Chi Square Test was used to analyze the dichotomous studied variables and to express the result outcomes as Number (Percentage) and as Odd Ratio (OD). Statistical analyses were performed using the IBM SPSS version. 25 (IBM Corp., Armonk, NY, USA) and p-values ≤0.05 were considered statistically significant.

This study finally included 110 eligible refractory hypokalemia associated hypomagnesemic critically ill patients admitted to our adult ICU at KHMC/RMS/Amman/Jordan between Jan 2018 and May 2021 via the Emergency Department (ED) or via other hospital wards with any medical or surgical problem. 

Table 1: The Inclusion and Exclusion Criteria of the Refractory Hypokalemia Associated Hypomagnesemic Critically Ill Patients

Table 2: The Comparative Variables of the Refractory Hypokalemia Associated Hypomagnesemic Critically Ill Patients Across Strategy I (Standard MgSO₄ Infusion Strategy) and Startegy II (Extended MgSO₄ Infusion Strategy)

The comparative variables between Group I and Group II were statistically analyzed by Independent T- Test and the results were expressed as Mean±SD and as Mean difference±SEM. While the comparative variables for the total sample were analyzed by One Sample T-Test and the results were also expressed as Mean±SD (at p-value<0.05*). Strategy I: Refractory hypokalemia associated hypomagnesemic critically ill patients who were infused 5 g MgSO4 over 4 hours as a Standard Infusion Strategy. Strategy II: Refractory hypokalemia associated hypomagnesemic critically ill patients who were infused 2 g MgSO4 over 12 hours twice daily as Extended Infusion Strategy. ICU: Intensive care unit. K₀: Last reported low potassium level before it was unsuccessfully normalized by KCL 2 mEq/mL. 1: Baseline levels after unsuccessful KCL management and before MgSO4 initiation. 2: Levels at the 3^rd after MgSO4 was infused using either of the proposed MgSO4 infusion strategies. 3_: Levels at the 3rd day after standard or extended MgSO4 infusion was stopped. KCl: Potassium chloride 2 mEq/ml vial. K: Potassium level. Mg: Magnesium level. cMg: Magnesium level after correction with ALB (cMg = Mg+0.08*(4-ALB). ALB: Albumin level. H.ALB: Human albumin 20%. Avg: average. SBP, DBP and MAP: Measured hemodynamics of systolic, diastolic and arterial blood pressures. HR: Heart rate. SI: Shock index.

Totally, 2045 ICU patients were excluded from our study because they had one or more of the exclusion clinical criteria during the study period [on Sympathomimetic agents (688, 33.64%), on Hydrocortisone (544, 26.60%), on Insulin (1454, 71.10%), on Furosemide (972, 47.53%), had an estimated CrCl <30 ml/min (421, 20.59%) and had a missed data (280, 13.69%).

The mean age of the whole studied refractory hypokalemia associated hypomagnesemic critically ill patients was 52.97±17.20 years and Strategy I patients were insignificantly younger than Strategy II patients (52.72±16.99 years versus 53.25±17.59 years, respectively, p-value = 0.874). Male patients were distributed in the study in approximately 1.34: 1 ratio to female patients [63 (57.3%) versus 47 (42.7%), respectively, p-value = 0.543] in which 56.9% (33 Men) and 43.1 (25 Women) were belonged to Strategy I compared to 57.7% (30 Men) and 42.3% (22 Women) in Strategy II. The refractory hypokalemia associated hypomagnesemia risk estimate for Female compared to Male patients in our study was 1.03 (95% CI; 0.49-2.20). There were an overall 61 (55.5%) medically and 49 (44.5%) surgically ICU patients which respectively distributed to 34 (58.6%) and 24 (41.4%) in Strategy I versus 27 (51.9%) and 25 (48.1%) in Strategy II with an odd ratio of 1.31 (95% CI; 0.62-2.79).

The overall 28-day ICU mortality was detected in 92 refractory hypokalemia associated hypomagnesemic critically patients with an overall incidence rate of 83.6% during an average of 10.95±2.128 days and 17.24±4.212 days of ICU and overall hospital admission days, respectively. Strategy I patients compared to Strategy II patients, had an overall 28-day ICU mortality, ICU stay days and overall hospital admission days of 47 (81.0%), 11.05±2.27 days and 17.00±4.11 days compared to 45 (86.5%), 10.83±1.97 days and 17.50±4.35 days, respectively, p-value>0.05. Based on Charlson Comorbities Index (CCI), there were insignificant differences in comorbidities status across the two tested MgSO4 infusion strategies.

Table 3: The Dichotomous Comparative Variables of the Refractory Hypokalemia Associated Hypomagnesemic Critically Ill Patients Across Strategy I (Standard MgSO₄ Infusion Strategy) and Strategy II (Extended MgSO₄ Infusion Strategy)

The comparative variables between Group I and Group II were statistically analyzed by Chi Square Test and the results were expressed as Number (Percentage) while the risk estimates were expressed as Odd Ratio (at p-value<0.05*). Group I: Refractory hypokalemia associated hypomagnesemic critically ill patients who were infused 5 g MgSO₄ over 4 hours as a Standard Infusion Strategy. Group II: Refractory hypokalemia associated hypomagnesemic critically ill patients who were infused 2 g MgSO₄ over 12 hours twice daily as Extended Infusion Strategy. Med: Medical. Sur: Surgical. M: Male. F: Female. 1: Baseline levels after nsuccessful KCL management and before MgSO₄ initiation. 2: Levels at the 3^rd after MgSO₄ was infused using either of the proposed MgSO₄ infusion strategies. 3_: Levels at the 3rd day after standard or extended MgSO₄ infusion was stopped. K: Potassium level. cMg: Magnesium level after correction with ALB (cMg = Mg+0.08*(4-ALB). CCI: Charlson Comorbidities Index. MORT: Mortality.

All eligible patients had K level (K0) below 3.5 mEq/l after receiving at least 100 mEq/day of KCL, 2.806±0.119 mEq/l and 143.25±14.39 mEq/day, in which it was insignificantly higher in Strategy I refractory hypokalemia associated hypomagnesemic studied patients (2.81±0.124 mEq/l) compared to Strategy II refractory hypokalemia associated hypomagnesemic studied patients (2.799±0.116 mEq/l).

Hemodynamically, there were insignificant differences in all selected studied hemodynamics across the two studied strategies. The average of the whole studied refractory hypokalemia associated hypomagnesemia, regarding Systolic Blood Pressure (SBP), Diastolic Blood Pressure (DBP), Mean Arterial Pressure (MAP), Heart Rate (HR) and Shock Index (SI), were 100.67±1.439 mmHg, 59.29±0.545 mmHg, 73.09±0.716 mmHg, 93.36±1.317 bpm and 0.928±0.025 bpm/mmHg, respectively.

Before the MgSO₄ infusion was pursued, all our studied patients had hypokalemia and hypomagnesemia with an average of 3.284±0.072 mEq/l and 1.886±0.070 mg/dl for baseline potassium level (K₁) and baseline corrected magnesium level (cMg₁), respectively. There were also insignificant differences in baseline levels of the two electrolytes between the two comparative strategies, in which Strategy I studied patient had slightly higher K1 (3.29±0.075 mEq/l) and cMg₁ (1.89±0.071 mg/dL) compared to Strategy II studied patients, 3.28±0.069 mEq/l and 1.88±0.069 mg/dL, respectively.

Once the diluted MgSO₄ 50% was infused via either Strategy I or Strategy II, while keeping the standard KCL maintenance replacement as it is, the average potassium level in Strategy I patients was nearly corrected (3.47±0.092 mEq/l) and in Strategy II patients was completely corrected (3.56±0.086 mEq/l) after at least 3 days of either MgSO4 infusion strategies. In case of magnesium level correction after 3 days of MgSO₄ infusion, magnesium levels (Mg₂) were completely corrected in both tested strategies with significantly higher level in case of the extended infusion strategies (Strategy II) compared with the standard infusion strategy (Strategy I) [2.83±0.098 mg/dL versus 2.69±0.103 mg/dL, respectively, p-value<0.05] with Mean difference±SEM of -0.139±0.019 mg/dL. The comparative variables of the refractory hypokalemia associated hypomagnesemic critically ill patients across Strategy I (Standard MgSO₄ Infusion Strategy) and Strategy II (Extended MgSO₄ Infusion Strategy).

In this study we compared two strategies of IV MgSO4 infusion in refractory hypokalemia associated hypomagnesemic critically ill patients. The first strategy was the standard infusion of 4 g MgSO₄ once daily for 3 days (Strategy I) while the comparative studied interventional strategy included critically ill patients who received the replacement MgSO₄ IV dose of 4 g MgS0₄ at two divided extended infusion each involved 2 grams MgSO4 infused over at least 6 hours for 3 days (Strategy II). The uniqueness of our, as we know, emphasized on its over 12 hrs twice daily, the second strategy is the traditional one over 4 hrs once daily. We asses and compared magnesium levels in different timing 4 hrs after infusion (early) and at least 72hrs after infusion (late). we found that we achieved Mg = 2 mg/dL in both strategies 4 hrs after infusion, fortunately the extended one has a higher percentage of retaining magnesium above 2 mg/dL at least 72hrs after the end of infusion. Most of the body's magnesium stores are intracellular, principally within bone, serum magnesium levels usually rise quickly with therapy, intracellular stores take longer to replete, it is therefore advisable in patients with normal renal function to continue magnesium repletion for at least one to two days after the serum magnesium concentration normalizes. Serum magnesium concentrations are regulated solely by renal excretion. Intravenous doses in particular can result in hypermagnesemia when GFR is severely impaired. our results explained by the theory that around 50% of the Mg⁺² replaced intravenously will be retained by the body, due to slow cellular uptake of Mg⁺² and replenishment of intracellular Mg⁺² stores, leading to serum concentrations exceeding the renal threshold.

Refractory hypokalemia associated with hypomagnesemia and intravenous magnesium therapy reverses the hypokalemia. Magnesium deficiency impairs Na-K-ATPase, which would decrease cellular uptake of K⁺. A decrease in cellular uptake of K⁺, if it occurs along with increased urinary or gastrointestinal excretion, would lead to K⁺ wasting and hypokalemia. Little K⁺ is excreted by the gastrointestinal tract normally; therefore, hypokalemia in magnesium deficiency is likely associated with enhanced renal K⁺ excretion. Hypomagnesemia may coexist with hypokalemia, hypocalcemia and hypophosphatemia [4]. Persons with chronic alcohol use disorder are prone to a variety of acid–base disturbances and some with mixed disturbances. Hypomagnesemia occurs in almost one-third of patients with chronic alcohol use disorder [5].

Prolonged magnesium infusion rates did not decrease magnesium replacement requirements. These results have been used to propose revision of our current magnesium infusion protocol to reduce infusion length [18]. Regarding clinical outcomes of LOS and mortality, patients in Strategy I had significantly lower ICU LOS, overall hospital LOS and 28-day ICU mortality compared with Strategy II, unfortunately we didn’t assess other cofounders that may affect these clinical outcomes. A larger, multisite and prospective study is needed to control for multiple confounders.

Symptoms of hypomagnesemia are non-specific and are often unrecognized. Mild hypomagnesaemia (0.5-0.7 mmol/l) does not generally give rise to problems in the short-term. However chronic suboptimal serum magnesium levels may predispose to dysrhythmias in cardiovascular disease; this is particularly important in patients on chronic diuretic therapy who may have co-existing hypokalemia. Correction of magnesium deficiency is necessary in order to correct hypokalemia. Magnesium deficiency must be corrected in hypocalcemic patients. Muscle weakness, cramp, carpopedal spasm or seizures may accompany hypomagnesaemia with or without hypocalcaemia. Mental changes and cerebellar signs may be associated with more severe magnesium depletion [12].

Hypomagnesemia has been associated with a significantly greater mortality rate in critically ill medical patients compared to normomagnesemia patients. In a study conducted by Rubeiz et al. 46% (17/37) of hypomagnesemic patients in the medical ICU died compared to 25% (37/147) of normomagnesemic patients (p-value <0.05) [13].

Magnesium replacement depends on the clinical situation and manifestations. In critical conditions such as pre-eclampsia, arrhythmias and tetany, large doses of IV magnesium are rapidly bolused and often followed by a continuous IV infusion. In asymptomatic patients, magnesium may be replaced by the oral or IV route depending on the clinical situation. The dose required to return patients to the normal magnesium range is variable and replacement may take several doses. Serum magnesium levels are primarily controlled by glomerular filtration and tubular reabsorption at the sites of the Loop of Henle and distal tubule. When faced with an increased filtered load of magnesium, the kidney is capable of increasing its excretion rate. Following intravenous (IV) administration, cellular magnesium uptake is slow and approximately 50% or more of the infused dose is lost due to increased excretion by the kidneys and decreased tubular reabsorption [14, 16].

The investigators current practice in the Medical and Neuroscience ICUs at CAMC General Hospital is to order 8g of magnesium sulfate for replacement in patients with hypomagnesemia. When IV magnesium sulfate is ordered the pharmacy automatically sets the rate to run at 2g per hour unless otherwise specified. Often times the physician will specify for 8g to be infused over eight hours. The basis of using an extended infusion is that a slower magnesium infusion rate may increase magnesium retention by allowing a longer period of time for magnesium uptake by cells and by decreasing the magnesium load delivered to the kidneys at any given time. As far as the investigators are aware, there have been no studies completed to date that assess the rate of IV magnesium infusion on the magnesium retention rate [14,17].

1. A.B. Clemmons et al. "Prolonged versus short infusion rates for intravenous magnesium sulfate administration in hematopoietic cell transplant patients." Biology of Blood and Marrow Transplantation, vol. 24, no. 3, 2018.

2. B. S. Charles et al. "Hypomagnesemia in the ICU – Does correction matter?" Journal of the Association of Physicians of India, vol. 64, Nov. 2016.

3. DePriest, L. Jack "Effect of magnesium sulfate infusion rate on magnesium retention in critically ill patients." National Library of Medicine, 31 Jan. 2014.

4. Di Nicolantonio, J. James et al. "Subclinical magnesium deficiency: A principal driver of cardiovascular disease and a public health crisis." Open Heart, vol. 5, no. 1, 13 Jan. 2018, e000668.

5. S. Doshi, J. Waller and A. Clemmons. "Prolonged versus short infusion rates of iv magnesium in hospitalized general medicine patients with hypomagnesemia." Hospital Pharmacy, vol. 55, no. 1, Feb. 2020, pp. 64–68.

6. D.A. Hammond et al. "Effectiveness and safety of magnesium replacement in critically ill patients admitted to the medical intensive care unit in an academic medical center: a retrospective, cohort study." Journal of Intensive Care Medicine, 2017, doi:10.1177/088506661772063.

7. B.A. Hansen and Bruserud. "Hypomagnesemia in critically ill patients." Journal of Intensive Care, vol. 6, 2018, p. 21.

8. W. Jahnen-Dechent and M. Ketteler. "Magnesium Basics." Clinical Kidney Journal, vol. 5, suppl. 1, 2012, pp. i3–i14.

9. P.M. Ku et al. "Prolonged versus short infusion rates for intravenous magnesium sulfate administration in hematopoietic cell transplant patients." Supportive Care in Cancer, vol. 26, no. 8, 2018, pp. 2809–2814.

10. C.S. Limaye et al. "Hypomagnesemia in critically ill medical patients." Journal of the Association of Physicians of India, vol. 59, 2011.

11. Panahi, Yunes et al. "The role of magnesium sulfate in the intensive care unit." EXCLI Journal, vol. 16, 5 Apr. 2017, pp. 464–482, doi:10.17179/excli 2017-182.

12. R.A. Reinhart and N.A. Desbiens. "Hypomagnesemia in patients entering the ICU." Critical Care Medicine, vol. 13, no. 6, 1985, pp. 506–507.

13. G.J. Rubeiz et al. "Association of hypomagnesemia and mortality in acutely ill medical patients." Critical Care Medicine, vol. 21, no. 2, 1993, pp. 203–209.

14. Sharpe, Peter et al. “Treatment of hypomagnesaemia in adult”. Southern Health and Social Care Trust, Dec. 2019, CG0089 (4).

15. M. Snyder et al. "Impact of intravenous magnesium infusion rate during ambulatory replacements on serum magnesium concentrations after allogeneic stem cell transplant." Supportive Care in Cancer, vol. 24, no. 10, 2016, pp. 4237–4240.

16. H.M. Soliman et al. "Development of ionized hypomagnesemia is associated with higher mortality rates." Critical Care Medicine, vol. 31, 2003, pp. 1082–1087.

17. D. Velissaris et al. "Hypomagnesemia in critically ill sepsis patients." Journal of Clinical Medicine Research, vol. 7, no. 12, 2015, pp. 911–918.

18. I.D. Weiner and C.S. Wingo. "Hypokalemia: Consequences, causes and correction." Journal of the American Society of Nephrology, vol. 8, 1997, pp. 1179–1188.