Serum osmolarity as a predictor of mortality in ICU COVID-19 patients: a retrospective analysis

Introduction

Serum osmolarity is a key determinant of total body water distribution and is governed by the concentrations of serum electrolytes, glucose, and urea. Disturbances in osmolarity are strongly linked to fluid imbalances, such as dehydration and hypernatremia, which primarily result from alterations in water balance but may also arise from other mechanisms, ultimately contributing to adverse clinical outcomes, including increased mortality. Serum osmolarity is routinely used in the diagnostic evaluation of electrolyte and metabolic disorders, providing clinicians with essential information for patient management (Najem et al., 2024). In clinical practice, serum osmolality is most often estimated by automated calculation rather than direct measurement, as it is faster and more practical (Faria et al., 2017). Regulation of hydrolytic balance is routinely assessed in clinical laboratories to aid in the differential diagnosis of renal dysfunction and small molecule poisonings. Serum osmolarity is also useful for evaluating the effects of specific treatments and toxins on an individual’s fluid balance (Martín-Calderón et al., 2015). The first formula for calculating serum osmolarity was introduced by Dorwart and Chalmers in 1975, followed by a simplified version by Smithline and Gardner in 1976, providing clinicians with a practical tool to estimate osmolarity and guide the assessment of fluid and electrolyte disorders (Heavens et al., 2014).

The prognostic value of serum osmolarity has been investigated across various patient populations, with hyperosmolarity consistently associated with increased mortality (Dikme & Dikme, 2016). However, this association has not been thoroughly validated in critically ill patients in intensive care units (ICUs), particularly those with severe respiratory infections such as COVID-19. Experimental studies suggest that hyperosmolarity may aid recovery from lung injury by suppressing cytokine production, and clinical data indicate that hypernatremia is not consistently associated with mortality in patients with respiratory disease (Lucijanic et al., 2023). Nevertheless, the overall role of serum osmolarity as a prognostic marker in this population remains unclear, highlighting the need to evaluate its predictive value in critically ill including Coronavirus Disease 2019 (COVID-19) patients.

Thus, it is important to re-evaluate the association between serum osmolarity and mortality in patients with respiratory illness. In recent years, the world has faced several respiratory virus outbreaks, including COVID-19, which was first reported in Wuhan in December 2019 and declared a pandemic by the World Health Organization on March 11, 2020. As of January 5, 2025, COVID-19 has caused 7,083,246 deaths worldwide (World Health Organization, 2025). Given their higher mortality compared to other respiratory viral infections, early prognostic markers are crucial for managing severe cases (Camacho-Domínguez et al., 2024; Grasselli et al., 2020). While sodium, glucose, and blood urea nitrogen (BUN) are recognized prognostic indicators, the predictive value of combined serum osmolarity in COVID-19 remains uncertain. Therefore, this retrospective study aimed to examine the association between admission serum osmolarity and in-hospital mortality in critically ill COVID-19 patients admitted from the emergency department (ED) to the ICU of a tertiary care hospital between March 2020 and April 2022.

Materials & Methods

Results

A total of 267 patients with confirmed COVID-19 were included, with data obtained from the HIMS database. The mean age was 53.8 ± 12.3 years, and 159 patients (59.6%) were male. The study comprised 203 non-survivors and 64 survivors, resulting in a mortality rate of 76%. All deaths were directly attributable to COVID-19 complications, including severe acute respiratory distress syndrome (ARDS) and multi-organ failure. The high mortality reflects that this single tertiary referral center primarily treated the sickest patients during pandemic surges. Patient characteristics stratified by survival status are presented in Table 1.

Primary outcome

Median serum osmolarity and Na levels at admission were significantly higher in survivors compared to non-survivors (288.37 vs. 285.75 mOsm/L, p = 0.034, and 135 vs. 133 mEq/L, p = 0.004, respectively). The relationship between sodium (Na) level and the probability of mortality was shown in Fig. 1. These findings suggest that lower serum osmolarity and hyponatremia on admission are associated with increased in-hospital mortality in critically ill COVID-19 patients, highlighting the potential prognostic value of early osmolarity assessment. Notably, ROC analysis demonstrated that serum sodium (AUC = 0.620) slightly outperformed calculated osmolarity (AUC = 0.588) in predicting mortality.

ROC analysis

ROC curve analysis indicated moderate predictive performance for serum osmolarity, Na, and neutrophil count, with AUC values of 0.588, 0.620, and 0.586, respectively. The optimal cut-off value for Na was 135 mEq/L, yielding an AUC of 0.593 (95% CI [0.511–0.675]). At this threshold, sensitivity for predicting mortality was 62.1%, specificity 54.7%, positive predictive value 81.3%, and negative predictive value 31.3% (Fig. 2, Table 3).

Table 3:

The ROC analysis for mortality.

		AUC		CI 95%		p
Serum Osmolarity		0.588		0.511	0.665	0.034
Na		0.620		0.540	0.701	0.004
Na, cut-off: 135 mmol/L		0.593		0.511	0.675	0.025
Neutrophyl Count (109/L)		0.586		0.506	0.666	0.038
		Survivors (n)	Non-Survivors (n)			%
Na (mmol/L)	<135	29	126	Sensitivity		62.1
	≥135	35	77	PPV		81.3
ROC analysis				Specificity		54.7
				NPV		31.3

DOI: 10.7717/peerj.20590/table-3

Notes:

ROC

Receiver operating characteristic

PPV

Positive predictive value

NPV

Negative predictive value

AUC

Area under the curve

CI

Confidence interval

Na

Natrium

Discussion

In this retrospective study of critically ill COVID-19 patients admitted to the ICU, we found that lower serum osmolarity and hyponatremia on admission were significantly associated with increased in-hospital mortality. Additionally, higher neutrophil and lymphocyte counts were observed in survivors, whereas inflammatory markers such as CRP, ferritin, and D-dimer did not predict mortality. Importantly, vasopressor use, and ETI emerged as independent predictors of mortality, consistent with the severity of critical illness in this study. In the multivariate model, Na, vasopressor use, and ETI remained significant independent predictors, with odds ratios of 0.89, 3.73, and 5.20, respectively, and the overall model demonstrated good discrimination (c-statistic 0.713).

Serum osmolarity plays a central role in intracellular and extracellular water distribution, and disturbances in osmolarity can lead to clinically adverse outcomes. Delayed recognition of osmolar abnormalities or inadequate fluid management may exacerbate these outcomes in critically ill patients. In our study of critically ill COVID-19 patients admitted from the ED to the ICU, lower admission serum osmolarity and hyponatremia were significantly associated with increased in-hospital mortality. While we did not directly assess patients’ volume status, hypo-osmolarity appeared to reflect more severe systemic derangements. Our ROC analysis revealed that serum Na demonstrated slightly better discriminative ability than calculated serum osmolarity (AUC 0.620 vs. 0.588), indicating that Na may serve as a stronger individual prognostic factor. Nonetheless, osmolarity, which integrates Na, glucose, and BUN, remains a valuable composite marker reflecting total osmotic balance.

Previous studies have demonstrated the predictive value of serum osmolarity in various patient populations, including patients with sepsis, cardiovascular disease, and non-COVID-19 respiratory illnesses (Dikme & Dikme, 2016; Martín-Calderón et al., 2015). Shen et al. (2017) reported that hyperosmolarity is associated with increased mortality in critically ill patients with cardiac, cerebral, vascular, and gastrointestinal diagnoses. Nicholson, Bennett & Silke (2012) showed that both hypo- and hyperosmolarity on admission were associated with increased mortality in acutely ill medical patients. In COVID-19, Lucijanic et al. (2023) found that both hypo- and hyperosmolar patients had higher mortality compared to normo-osmolar patients, with hyperosmolar patients showing the greatest risk. Our results extend these findings by demonstrating that admission hypo-osmolarity is a significant predictor of mortality among critically ill COVID-19 patients, highlighting its potential as an early prognostic marker. Hypo-osmolarity may identify a high-risk subgroup of ICU patients who could benefit from closer monitoring and early interventions to optimize fluid and electrolyte balance.

Sodium disturbances are a major determinant of osmolarity, and their relationship with mortality is complex. Hypernatremia has consistently been linked to higher mortality, whereas hyponatremia has shown mixed results (Hirsch et al., 2021; Tzoulis et al., 2021; Hu et al., 2021). Hyponatremia is more frequent than hypernatremia, but its prognostic value is limited except in hypovolemic states (Tzoulis et al., 2021). In COVID-19, Hirsch et al. (2021) reported that the association between hyponatremia and mortality was no longer significant after correcting sodium for serum glucose. In our study, hyponatremia was the principal driver of hypo-osmolarity and was independently associated with increased mortality, supporting its value as a prognostic marker.

The logistic regression model yielded a c-statistic of 0.713, with 70.4% sensitivity and 65.8% specificity, indicating acceptable discriminatory performance. These findings reinforce that admission Na levels, vasopressor use, and ETI can serve as reliable indicators of disease severity and mortality risk in critically ill COVID-19 patients. In addition to biochemical parameters, clinical severity markers remained strong predictors of mortality. Vasopressors and ETI during hospitalization were independently associated with death, reflecting hemodynamic instability and respiratory failure in critically ill patients. While inflammatory markers such as CRP, ferritin, and D-dimer were not associated with mortality, median PCT levels were slightly higher in non-survivors but remained within the viral infection range, limiting their clinical relevance. These findings emphasize that combining biochemical markers like osmolarity and Na with clinical severity indicators may improve early risk stratification in the ICU.

Several limitations should be acknowledged. First, the retrospective single-center design may limit generalizability, particularly as our center served as a tertiary referral hospital treating the sickest patients during pandemic surges, which explains the high observed mortality rate. Second, only admission laboratory values were evaluated, and dynamic changes in osmolarity or electrolytes during ICU stay were not analyzed. Third, the study design precludes causal inference, and residual confounding cannot be excluded despite multivariate adjustment. Finally, our sample size, while sufficient to detect meaningful associations, was relatively modest, and external validation in larger multicenter cohorts is warranted.

Conclusions

This study demonstrates that lower admission serum osmolarity and hyponatremia are independently associated with increased in-hospital mortality among critically ill COVID-19 patients. In the multivariate model, serum Na, vasopressor use, and ETI emerged as significant independent predictors of mortality, with acceptable model discrimination (c-statistic 0.713, sensitivity 70.4%, specificity 65.8%). Among biochemical parameters, Na showed slightly superior prognostic performance compared to calculated osmolarity; however, osmolarity remains an integrated and clinically meaningful indicator of total osmotic and fluid balance.

These findings suggest that early identification and correction of hypo-osmolar or hyponatremic states could contribute to improved management and risk stratification of critically ill COVID-19 patients. Calculated serum osmolarity represents a simple, inexpensive, and readily available biomarker that can be incorporated into initial ICU assessments to guide clinical decision-making. Future prospective, multicenter studies are warranted to validate these results and explore whether targeted interventions to normalize osmolarity and Na levels may improve survival outcomes in this high-risk population.

Supplemental Information