Hyper- and hypomagnesemia as an initial predictor of outcomes in septic pediatric patients in Egypt

Abstract

Background

Critically ill septic children are susceptible to electrolyte abnormalities, including magnesium disturbance, which can easily be neglected. This study examined the potential correlation between serum magnesium levels upon admission to the pediatric intensive care unit (PICU) and the outcomes of critically ill septic patients.

Methods

This prospective study, conducted from May 2023 to November 2023, included 76 children with sepsis who underwent clinical and lab assessments that included initial magnesium levels. The outcome of sepsis was documented. Predictors of mortality were identified through multivariate logistic regression models, with discrimination and calibration assessed using the area under the curve (AUC).

Results

The median magnesium level upon PICU admission was 2.0 mg/dl (range 1.1–4.9), and it was slightly higher in non-survivors than survivors (2.1 mg/dl; interquartile range [IQR], 1.9–2.5 vs. 2.0; IQR, 1.8–2.6, respectively), Hypermagnesemia was observed to have a negative effect on critically ill septic patients. It was also found that hypermagnesemia was associated with low C-reactive protein levels (P=0.043). With a cutoff of 5.5, the pediatric Sequential Organ Failure Assessment score strongly predicted mortality (AUC=0.717, P<0.001), with a sensitivity of 64.3% and specificity of 68.8%.

Conclusions

As an initial predictor of mortality, the serum magnesium level cannot be used alone; however, hypermagnesemia has a negative impact on critically ill septic patients. Thus, healthcare professionals should be cautious with magnesium administration.

INTRODUCTION

Sepsis is an uncontrolled reaction to infection that leads to serious organ impairment [1]. Globally, the incidence of pediatric sepsis is 1.2 million cases yearly, and the mortality rate is as high as 5% [2]. Prompt intervention is crucial, and early identification of sepsis risk factors upon admission could be beneficial [3]. In critically ill pediatric patients, electrolyte imbalances are frequent and could significantly affect patient outcomes. In pediatric intensive care units (PICUs), emergency care aims to keep the body in an equilibrium state, including both fluid and electrolyte distribution, to support the function and maintenance of the organs [4]. Electrolyte disorders may be caused by a primary illness mechanism, organ dysfunction, fluid and electrolyte interventions, and the utilization of certain medications [5]. Magnesium is an essential mineral for metabolism. It acts as a cocatalyst for several enzymes and controls several processes, such as the contractile function of muscles, neuromuscular conduction, blood glucose regulation, heart contractility, and vascular pressure control [6]. Metabolic disarrangements involving magnesium levels are common in hospitalized cases and can influence the prognosis of certain diseases [7]. Few studies have evaluated serum magnesium levels in pediatric septic patients, even though it is imperative to properly evaluate and predict changes in serum magnesium levels in those patients to improve their clinical outcomes. Therefore, this study assesses the independent association between serum magnesium levels upon PICU admission and mortality in critically ill septic patients.

MATERIALS AND METHODS

This research was approved by the Research Ethics Committee of Kasr Al-Ainy Faculty of Medicine, Cairo University (No. N-78-2023). The parents or guardians of all enlisted patients provided written informed consent.

This observational prospective research enrolled 76 patients from age 1 month to 14 years who presented to the PICU of Children’s Hospital from May 2023 to November 2023. In this study, 76 patients from a total of 144 PICU admissions met the 2005 International Pediatric Sepsis Consensus Conference definition of pediatric sepsis [8]. Patients younger than 1 month or older than 14 years, patients admitted to the PICU with causes other than sepsis, and patients with no serum magnesium data were excluded.

A thorough history and clinical examination were performed for each patient upon admission. The severity of illness was assessed using the pediatric Sequential Organ Failure Assessment (pSOFA) score. All enrolled patients received a sepsis investigation, arterial blood gas analysis, and electrolyte measurements, including magnesium levels. The serum magnesium level was detected by the Xylidyl blue method using a sample taken upon admission. It was analyzed as normomagnesemia, hypomagnesemia, and hypermagnesemia according to reference ranges from 1.5 to 2.2 mg/dl [9]. The duration of PICU admission and whether the patient left the PICU alive or dead were recorded.

Statistical Analysis

Data were analyzed using the SPSS version 28 (IBM Corp.). Quantitative data are presented using the median and interquartile range (IQR). Categorical data are presented using frequency (count) and relative frequency (percentage). The non-parametric Kruskal-Wallis and Mann-Whitney tests were used to compare quantitative variables [10], and the chi-square test was used to compare categorical data. If the expected frequency was less than 5, the exact test was performed [11]. The cutoff values for C-reactive protein (CRP), magnesium, and the pSOFA score for predicting mortality were determined using an area under the curve (AUC) analysis of the receiver operating characteristic (ROC) curve. A logistic regression was done to detect independent predictors of mortality. P-values less than 0.05 were considered statistically significant.

RESULTS

This prospective study analyzed 76 patients. Forty-four (57.8%) of them were male, and 42% were female. Figure 1A shows the classification of the admitted patients into serum magnesium level groups. The baseline characteristics and parameters of the study population according to the magnesium level groups are provided in Table 1. According to the definitions of pediatric sepsis [8], two patients (2.63%) had systemic inflammatory response syndrome, 21 (27.63%) had sepsis, 48 (63.15%) had septic shock, and 5 (6.57%) had multi-organ dysfunction syndrome, with the cardiovascular and respiratory systems most involved. The source of infection was the respiratory tract or blood, and 27.6% of the children had co-morbidities upon admission.

Figure 1.

(A) Flowchart of admitted patients according to their magnesium levels. (B) Flowchart of patients included in this study classified according to the definitions of pediatric sepsis. Each percentage in Fig. 1B has been calculated based on the total number of magnesium level classifications in Fig. 1A. SIRS: systemic inflammatory response syndrome; MOSF: multi-organ dysfunction syndrome.

Table 1.

Baseline characteristics and parameters of the studied population according to different categories of magnesium (Mg) level

Variable	Normal magnesium level	Hypomagnesemia	Hypermagnesemia	P-value
Age (yr)	2 (1–6)	2 (1–8)	1 (0–4)	0.417
SBP (mm Hg)	88.0   (75.0–100)	94.5 (78.5–108.5)	90.0 (80.0–100)	0.479
DBP (mm Hg)	50.0 (40.0–60.0)	56.0 (40.0–72.0)	55.0 (45.0–60.0)	0.916
Temperature (°C)	38.5 (38.3–39.0)	38.5 (38.0–39.0)	38.5 (38.0–39.0)	0.489
HR (beats/min)	140 (120–160)	130 (120–155)	140 (130–150)	0.823
RR (rate/min)	35.0 (30.0–45.0)	34.5 (27.0–35.5)	35.0 (30.0–45.0)	0.608
Capillary refill (sec)	3 (3–3)	3 (2.5–3)	3 (3–3)	0.630
GCS	14 (10–15)	15 (14–15)	15 (11–15)	0.413
pSOFA score	5.0 (3.0–10.0)	3.5 (2.0–8.0)	5.0 (3.0–9.0)	0.692
Hemoglobin (g/dl)	10.0 (8.6–11.6)	8.1 (6.6–9.45)	9.95 (8.9–10.9)	0.219
TLC (/μl of blood)	17.6 (10.9–21.0)	6.7 (4.3–34.0)	17.9 (14.0–24.0)	0.297
PLT (/μl of blood)	345.0 (173–435)	160.0 (104–544)	324.5 (155–433)	0.780
CRP (mg/dl)	48.0 (6.0–170.0)	24.0 (0–72.0)	12.0 (0–96.0)	0.043
Platelet count (%)	54.0 (40.5–68.5)	42.5 (40.0–45.0)	54.0 (41.0–70.0)	0.509
PT (sec)	18.2 (15.9–26.0)	19.5 (18.0–21.0)	17.0 (14.0–19.8)	0.366
INR	1.4 (1.3–2.0)	1.75 (1.6–1.9)	1.4 (1.3–1.7)	0.521
pH	7.36 (7.30–7.42)	7.40 (7.35–7.50)	7.35 (7.22–7.42)	0.436
PCO2 (mm Hg)	33.0 (26.0–37.0)	20.5 (20.0–21.0)	35.0 (30.0–41.0)	0.089
HCO3 (mmol/L)	19.0 (15.7–23.0)	16.5 (11.0–22.0)	19.0 (14.0–24.0)	0.898
Na (mmol/L)	137.0 (134.0–144.0)	143.0 (140.0–157.0)	141.5 (136.0–149.0)	0.07
K (mmol/L)	4.5 (3.9–5.1)	3.7 (2.9–4.5)	4.4 (3.3–4.9)	0.301
PICU length of stay (day)	7.0 (3.0–12.0)	5.5 (3.0–9.5)	7.0 (5.0–13.0)	0.659

Values are presented as median (interquartile range).

SBP: systolic blood pressure; DBP: diastolic blood pressure; TEMP: temperature; HR: heart rate; RR: respiratory rate; GCS: Glasgow Coma Scale; pSOFA: pediatric Sequential Organ Failure Assessment; HB: hemoglobin; TLC: total leukocyte count; PLT: platelets; CRP: C-reactive protein; PC: platelet count; PT: prothrombin time; INR: international normalized ratio; PICU: pediatric intensive care unit.

As the outcome, 28 patients (36.8%) died, and 48 patients (63.2%) survived to hospital discharge (Table 2). Hypomagnesemia was found to be more associated with sepsis (75%) than normomagnesemia and hypermagnesemia (26.8% and 22.6%, respectively) (Figure 1B). Table 2 shows the underlying causes of sepsis and the need for mechanical ventilators and inotropic support. As shown in Table 1 and Figure 2, hypermagnesemia was associated with lower CRP levels (P=0.043, P=0.016 if compared to normomagnesemia). More patients with low magnesium levels (25%) than hypermagnesemia (19.4%) required mechanical ventilation, but no statistically significant differences in the need for mechanical ventilation and inotropic support were found across the different magnesium level groups. However, the need for those medical interventions correlated significantly with mortality (P=0.041 for mechanical ventilation and 0.044 for inotropic support). The mortality rate in the group with normal magnesium levels was slightly higher but nearly the same as in those with hypermagnesemia.

Table 2.

Diagnosis and clinical course according to the magnesium (Mg) levels

Variable	Normal magnesium level	Hypomagnesemia		Hypermagnesemia	P-value
Diagnosis
Sex					0.402
Female	15 (36.6)	1 (25.0)		16 (51.6)
Male	26 (63.4)	3 (75.0)		15 (48.4)
Gastroenteritis	2 (4.9)	1 (25.0)		1 (3.2)	0.302
Encephalopathy	1 (2.4)	-		5 (16.1)	0.126
Acute renal injury	3 (7.3)	-		1 (3.2)	0.703
Diabetic ketoacidosis	2 (4.9)	-		3 (9.7)	0.732
Pneumonia	11 (26.8)	-		7 (22.6)	0.673
Cardiogenic shock	2 (4.9)	-		-	0.554
Coagulopathy	2 (4.9)	-		-	0.554
GBS	1 (2.4)	-		1 (3.2)	1.000
Heart failure	1 (2.4)	-		3 (9.7)	0.445
Autoimmune encephalitis	1 (2.4)	-		-	1.000
Clinical course
Need for MV	15 (36.6)	1 (25.0)		6 (19.4)	0.899
Need for inotropes	16 (39.0)	1 (25.0)		10 (32.3)	1.000
Mortality	16 (39.0)	-		12 (38.7)	0.415
Mortality	Survivor		Non-survivor
Need for MV	10 (2.8)		12 (42.9)		0.041
Need for inotropes	13 (27.1)		14 (50.0)		0.044

Values are presented as number (%).

GBS: Guillain barre syndrome; MV: mechanical ventilator.

Figure 2.

Relation between C-reactive protein level and serum magnesium level.

Although the male death rate (54.2%) was higher than the female death rate (45.8%), that difference was not significant (P=0.389). The median length of PICU stay was higher in non-survivors than survivors (Table 3), and it was the same in the hypermagnesemia and normomagnesemia groups and lower in the hypomagnesemia group (Table 1). Table 3 shows that the pSOFA score was higher in non-survivors than survivors (P=0.002). The median magnesium level was slightly higher in patients who died than in those who survived (2.1 mg/dl; IQR, 1.9–2.5 vs. 2.0; IQR, 1.8–2.6, respectively), but the difference was not statistically significant (P=0.615).

Table 3.

Comparison between non-survivors and survivors regarding pSOFA score, magnesium, and PICU length of stay

Variable	Non-survivor	Survivor	P-value
pSOFA score	7.5 (5.0–11.5)	4.5 (2.0–8.5)	0.002
Magnesium (mg/dl)	2.1 (1.9–2.5)	2.0 (1.8–2.6)	0.615
PICU length of stay days	8.0 (3.5–13.5)	6.5 (4.0–11.0)	0.846

Values are presented as median (interquartile range).

pSOFA: pediatric Sequential Organ Failure Assessment; PICU: pediatric intensive care unit.

The ROC curve was analyzed to compare the predictive validity of magnesium with that of other factors relevant to mortality. Magnesium had an AUC of 0.535 (P=0.615), and CRP had an AUC of 0.449 (P=0.459). On the contrary, the pSOFA score highly predicted mortality (AUC=0.717, P<0.001), and in this study, a cutoff score of 5.5 had 64.3% sensitivity and 68.8% specificity (Figure 3). In the multivariate logistic regression for predicting mortality, the pSOFA score was the only significant predictor of mortality (Table 4).

Figure 3.

Receiver operating characteristic (ROC) curve of C-reactive protein (CRP), serum magnesium level, and pediatric Sequential Organ Failure Assessment (pSOFA) score for mortality detection in the studied patients

Table 4.

Multivariate logistic regression to detect independent predictors of mortality

		OR	95% CI	P-value
pSOFA score as a predictor of mortality		1.312	1.066–1.614	0.010

OR: odds ratio; pSOFA: pediatric Sequential Organ Failure Assessment.

DISCUSSION

Severe sepsis and septic shock are responsible for 25% of deaths in PICUs globally and account for nearly 8% of admissions to PICUs [12]. Magnesium is obligatory for human life and is involved in several biochemical and physiological processes in the body [13]. Critical illness influences normal body functions, and numerous physiological imbalances generally compromise the ability to maintain electrolyte levels; in crisis situations, magnesium is often overlooked and has not been explored extensively [6]. The goal of this study was to disclose that defect and evaluate how magnesium levels affect the outcomes of critically ill pediatric patients in a developing country because magnesium disorders have recently drawn attention from health practitioners as an innovative prognostic marker.

In this study, 144 patients were admitted to the PICU during the study period, and 76 (60.8%) of them were diagnosed with sepsis and enrolled. Therefore, sepsis is a major reason for PICU admission, which agrees with the World Health Organization’s declaration that sepsis should be a top health priority [1]. Sepsis was found related to the hypomagnesemia group (75%) which agrees with the following studies: Vijaywargiya et al. [6], Haque and Saleem [14], Dandinavar et al. [15], Erdoğan and Menevşe [16], Singhi et al. [17], confirming that magnesium plays a crucial role in sepsis due to its effects on the immune system, which are important in the pathogenesis of sepsis. In this study, respiratory problems were among the main causes of admission for critically ill septic children, which is in concordance with several studies: Erdoğan and Menevşe [16], and Bekhit et al. [18], and El Beleidy et al. [19].

Only 4 (5.3%) of those studied had hypomagnesemia, whereas 41 (53.9%) had normal magnesium levels, and 31 (40.8%) had hypermagnesemia. Similarly, Bekhit et al. [18] reported that only 3% of cases had hypomagnesemia, 57% had normal magnesium levels, and 40% had hypermagnesemia. As observed in another study by Vijaywargiya et al. [6], hypomagnesemia presented in only 7.8% of the studied cases upon admission, whereas hypermagnesemia was seen in 20.2% of patients. According to Zafar et al. [20], 50 (71.43%) of 70 critically ill patients had normomagnesemia, which is similar to our study (40.8% in the normomagnesemia group). In contrast to our findings, several studies have reported a higher percentage of patients with hypomagnesemia (Balcı et al. [21], Agarwal et al. [22], and Zafar et al. [20]). In addition, Deshmukh et al. [23] and Filyk [24] reported that 70% of patients admitted to the PICU had hypomagnesemia. Haque and Saleem [14] discovered hypomagnesemia in 44% of their PICU patients; El Beleidy et al. [19] found hypomagnesemia in 31% of their subjects; Erdoğan and Menevşe et al. [16] found hypomagnesemia in 29%, and Zafar et al. [20] found that 24.29% of their patients had hypomagnesemia. The current study found that the hypermagnesemia group (40.8%) was more prevalent than the hypomagnesemia group (5.3%), which agrees with research done by Naseem et al. [4], who reported more cases of hypermagnesemia (up to 21%) than hypomagnesemia (7%). The high number of patients with hypermagnesemia in the present study might be explained by the possibility of magnesium administration before PICU admission.

This study demonstrates that CRP was lower in patients with hypermagnesemia than in those with normal magnesium levels (P=0.016), which can be attributed to the role of magnesium in regulating inflammation. That finding agrees with the finding of Mazidi et al. [25] that magnesium supplementation significantly reduces serum CRP levels. Švagždienė et al. [26] declared that the serum magnesium level is inversely proportional to CRP. Another study, Rodríguez-Morán and Guerrero-Romero [27], found that magnesium deficiency is highly associated with raised CRP levels.

In the current study, a higher percentage (25%) of patients with low magnesium levels required mechanical support, relative to the hypermagnesemia group (23.5%), which can be attributed to the crucial role of magnesium in alleviating respiratory dysfunction. This is in concordance with the studies done by Vijaywargiya et al. [6] and Stephanos and George [28]. On the other hand, the need for inotropic support was higher in the hypermagnesemia group than in hypomagnesemia, which agrees with the study done by Celi et al. [29], who revealed that hypermagnesemia correlates with a need for vasopressors, perhaps because hypermagnesemia can cause cardiac dysfunction and heart failure. In this study, the overall mortality rate was 36.8%, which is similar to the rate reported by Abdelkader et al. [30]; however, it is greater than in studies from developed countries, which reported death rates of 6.2% and 2.7%, respectively [31,32].

According to the current data, the non-survival rate was higher in the normomagnesemia group than the hypomagnesemia group, where all four patients survived, which agrees with Haque and Saleem [14], who reported a higher number of deaths in children with normal magnesium levels than in children with hypomagnesemia. However, the median magnesium level was determined to be slightly elevated in those who died, compared with those who survived, and that could indicate that hypermagnesemia has a negative effect on critically ill septic patients, in agreement with Saglietti et al. [33], who found that hypermagnesemia had adverse effects on a pediatric population. That result is supported by the finding in our finding that patients who suffered from septic shock and multiple organ system failure were most often found in the hypermagnesemia group (Figure 1B). In addition, several studies have shown that hypermagnesemia is a strong predictor of poor outcomes [4,7-35].

In this study, the pSOFA score was an excellent predictor of mortality (AUC=0.717, P<0.001) when the cutoff level was 5.5, which agrees with Rosenzweig and Yuki [36]. Therefore, the pSOFA score should be acquired promptly upon PICU admission and used to direct patient management because more than a third of deaths in PICUs are of patients with sepsis [37]. Unfortunately, magnesium levels did not reveal a significant relationship with any of the outcomes, and the studies done by Vijaywargiya et al. [6] and Wang et al. [38] agree with that observation. Nonetheless, the current study raises a concern about hypermagnesemia.

This study has a few limitations. First, this observational study was performed in one tertiary center, which limits its generalizability. Second, it included a small number of patients. Thus, further multi-center research with a larger patient population is required to validate these findings about the utility of adding magnesium and other serum electrolyte levels to the pSOFA score to anticipate mortality. Sepsis is a serious health problem that threatens the lives of children. Serum magnesium levels cannot be used independently to predicts the outcomes of critically ill septic patients; however, hypermagnesemia has a negative impact on critically ill septic patients. All children admitted to the PICU should have their electrolyte levels checked properly and treated cautiously.

KEY MESSAGES

▪ Sepsis is a major cause of morbidity and mortality in children; it can be associated with electrolyte imbalance.

▪ As an initial predictor of mortality in critically ill septic children, serum magnesium can be used with other markers.

▪ Healthcare professionals should be cautious when giving magnesium supplementation, particularly to patients without hypomagnesemia.