Serum Sodium Trajectories and All-Cause Mortality in Critically Ill Patients with Sepsis-Associated Delirium: A Retrospective Analysis Based on the MIMIC-IV Database

Serum Sodium Trajectories and All-Cause Mortality in Critically Ill Patients with Sepsis-Associated Delirium: A Retrospective Analysis Based on the MIMIC-IV Database | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Serum Sodium Trajectories and All-Cause Mortality in Critically Ill Patients with Sepsis-Associated Delirium: A Retrospective Analysis Based on the MIMIC-IV Database Shuyang Dai, Bingjie Li, Zongshan Zhang, Gaoli Zhang, Poshi Xu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9036323/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 10 You are reading this latest preprint version Abstract Objective: To investigate the association between dynamic serum sodium trajectories during the first 5 days of intensive care unit (ICU) admission and all-cause mortality in critically ill patients with sepsis-associated delirium (SAD), providing evidence-based insights for prognostic assessment and clinical management of this population. Methods: This retrospective cohort study extracted clinical data from adult SAD patients admitted to ICU between 2008 and 2022 from the MIMIC-IV version 3.1 database. Group-based trajectory modeling (GBTM) was employed to classify serum sodium trajectories within 5 days of ICU admission. Kaplan-Meier survival curves were used to compare survival differences among trajectory groups, while Cox proportional hazards models were applied to analyze the associations between serum sodium trajectories and 30-day and 365-day all-cause mortality. Subgroup analysis and sensitivity analysis were conducted to validate the robustness of these associations. Results: A total of 1,447 adult SAD patients were finally enrolled in the study after applying exclusion criteria. GBTM identified three distinct serum sodium trajectories: normal-stable (n=868, 59.99%), persistent-hyponatremia (n=263, 18.18%), and hypernatremia-fluctuating (n=316, 21.84%). Kaplan-Meier analysis revealed significantly higher 30-day mortality (14.17% vs. 25.63% P<0.001) and 365-day mortality (16.13% vs. 27.85%, P<0.001) in the hypernatremia-fluctuating group compared with the normal-stable group. After adjusting for demographics, disease severity scores, comorbidities, and therapeutic interventions in multivariable Cox models, the hypernatremia-fluctuating trajectory remained an independent risk factor for 30-day (HR=1.744, 95% CI: 1.296-2.347, P<0.001) and 365-day mortality (HR=1.694, 95% CI: 1.277-2.247, P<0.001). Subgroup analyses indicated that the association between the Hypernatremic-Fluctuating trajectory and mortality was generally consistent across different subgroups defined by age, BMI, sex, and comorbidities.. Conclusions: Three distinct serum sodium trajectories were identified in critically ill SAD patients: normal-stable, persistent-hyponatremia, and hypernatremia-fluctuating. The hypernatremia-fluctuating trajectory represents an independent risk factor for both short-term and long-term all-cause mortality. Dynamic monitoring of serum sodium trajectories provides superior prognostic value compared with single-point measurements. Implementing serial serum sodium assessments during the first 5 days of ICU admission and identifying hypernatremia-fluctuating patterns may facilitate early risk stratification and individualized sodium homeostasis management, ultimately improving clinical outcomes in SAD patients. sepsis-associated delirium serum sodium trajectory group-based trajectory modeling all-cause mortality critically ill patients Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Sepsis-associated delirium (SAD) represents one of the most common neurological complications of sepsis, affecting between 9% and 71% of critically ill septic patients, with incidence reaching 70%-87% among those requiring mechanical ventilation[1, 2]. Defined as an acute alteration in mental status characterized by inattention, disorganized thinking, and fluctuating levels of consciousness, SAD has been consistently associated with increased short-term mortality, prolonged ICU length of stay, and long-term cognitive impairment among survivors[3, 4]. Despite its substantial clinical significance, the pathophysiological mechanisms underlying SAD remain incompletely understood, with neuroinflammation, blood-brain barrier disruption, and neurotransmitter imbalance currently recognized as key contributing factors[5, 6]. Electrolyte disturbances, particularly dysnatremia, are highly prevalent in critically ill patients and increasingly recognized as important independent determinants of adverse outcomes[7, 8]. As the predominant extracellular cation, sodium plays a fundamental role in maintaining osmotic balance, regulating cell volume, and determining neuronal excitability. Both hyponatremia and hypernatremia are common in sepsis, with reported incidence ranging from 15% to 30% depending on the study population and diagnostic criteria employed[9, 10]. These electrolyte abnormalities may hold particular pathophysiological relevance in SAD, as rapid shifts in serum osmolality can exacerbate cerebral edema or precipitate osmotic demyelination, thereby worsening delirium and neurological outcomes[11, 12]. Notably, the blood-brain barrier of SAD patients is already damaged by neuroinflammation, making their brain tissue more sensitive to sodium fluctuations, yet the dynamic association between serum sodium trajectories and SAD prognosis has not been clarified. Accumulating evidence suggests that dynamic patterns of serum sodium variation may carry greater prognostic significance than static single-point measurements. In a landmark study, Chewcharat et al.[13]applied group-based trajectory modeling (GBTM) to identify five distinct serum sodium trajectories in hospitalized patients, demonstrating that any deviation from the normal-stable pattern was associated with increased risks of in-hospital and one-year mortality. Subsequent investigations have extended these findings across diverse critical care populations, including patients with acute kidney injury[14], sepsis with lactic acidosis[15], and those requiring renal replacement therapy, consistently identifying fluctuating trajectories as harbingers of poor prognosis. Notably, Li et al.[15]recently demonstrated that sepsis patients with lactic acidosis who maintained stable normonatremia exhibited the lowest 30-day mortality, whereas those with fluctuating sodium trajectories experienced the worst outcomes. However, Li et al.'s study did not include delirium patients, and the prognostic impact of hyponatremia may be different from that in SAD population. In addition, studies on serum sodium trajectories in patients with traumatic brain injury have also confirmed the adverse prognostic value of sodium fluctuation[11], but the brain injury mechanism of traumatic brain injury is mechanical damage, which is different from the inflammatory brain injury of SAD, and it is unclear whether the characteristics and prognostic value of serum sodium trajectories are specific in SAD population. To date, no study has specifically characterized the serum sodium trajectory subtypes in SAD patients nor verified their independent prognostic value after adjusting for comprehensive clinical covariates. Given the high incidence of both hyponatremia and delirium in sepsis, alongside their potentially shared pathophysiological mechanisms involving neuroinflammation and blood-brain barrier dysfunction, we hypothesized that distinct serum sodium trajectories would be identifiable in SAD patients and that unstable trajectories would correlate with increased mortality. Utilizing the MIMIC-IV database, this study aims to: (1) identify distinct serum sodium trajectory subgroups during the first 5 days following ICU admission in SAD patients; (2) evaluate the associations between these trajectories and all-cause mortality at 30 and 365 days; and (3) assess the consistency of these associations across clinically relevant subgroups, thereby informing evidence-based prognostic assessment and individualized electrolyte management strategies for this vulnerable population. Methods Data Source and Study Population This retrospective cohort study utilized data from the Medical Information Mart for Intensive Care IV (MIMIC-IV, version 3.1) database, a publicly available, de-identified critical care database containing comprehensive clinical data from patients admitted to Beth Israel Deaconess Medical Center in Boston, Massachusetts, between 2008 and 2022[16]. The database encompasses detailed information on demographics, vital signs, laboratory measurements, medications, procedures, and outcomes. One author (S.D.) completed the required training and obtained certification from the Collaborative Institutional Training Initiative (CITI) program (Certificate No. 14012091) to access the database. The study cohort comprised adult patients (≥18 years) admitted to the ICU with documented diagnoses of both sepsis and delirium. Delirium diagnosis was required to occur within 48 hours of ICU admission to exclude the interference of late-onset delirium on serum sodium trajectory. Sepsis was identified using International Classification of Diseases (ICD)-9 and ICD-10 codes (ICD-9: 995.91, 995.92; ICD-10: A40.x, A41.x), supplemented by a Sequential Organ Failure Assessment (SOFA) score ≥2, consistent with the Sepsis-3 criteria[17]. Delirium was defined using the Confusion Assessment Method for the ICU (CAM-ICU), with the diagnostic criteria extracted from the MIMIC-IV database's delirium assessment module as previously validated[18, 19]. The following exclusion criteria were applied: (1) ICU length of stay <72 hours (reason: patients with short ICU stay have insufficient serum sodium detection data to complete trajectory modeling, and the prognosis bias of such patients is large); (2) missing serum sodium measurements for any 24-hour period during the first 5 days of ICU admission; (3) receipt of continuous renal replacement therapy (CRRT) during the observation period; (4) multiple ICU admissions for sepsis (only the first admission was included); and (5) missing key covariates. Data Collection Data extraction was performed using PostgreSQL (version 15) and Navicat Premium (version 17) through structured query language. Extracted variables included: (1) Demographics: age, sex, race, body mass index (BMI); (2) Comorbidities: hypertension, type 2 diabetes mellitus, heart failure, chronic obstructive pulmonary disease (COPD), chronic kidney disease (CKD), pneumonia, malignancy, acute myocardial infarction, ischemic cardiomyopathy, ischemic stroke; (3) Laboratory parameters: white blood cell count (WBC), hemoglobin, platelet count, serum chloride, potassium, lactate, glucose, anion gap, and serum sodium (measured at least every 24 hours during the first 96 hours); (4) Vital signs: systolic blood pressure (SBP), diastolic blood pressure (DBP), heart rate, respiration rate, body temperature, and oxygen saturation; (5) Disease severity scores: SOFA score, Simplified Acute Physiology Score II (SAPS II), Acute Physiology Score III (APS III), Oxford Acute Severity of Illness Score (OASIS), and Charlson Comorbidity Index (CCI); (6) Therapeutic interventions: mechanical ventilation, vasoactive medications; and (7) Outcomes: ICU mortality, hospital mortality, and 30-day survival status. For patients with multiple serum sodium measurements within a 24-hour period, the first recorded value was used. Variables with >20% missing data were excluded, while those with ≤20% missing data were imputed using multiple imputation methods[20]. The variance inflation factor (VIF) was used to assess multicollinearity among covariates in the Cox proportional hazards model, with a VIF value < 5 indicating no significant multicollinearity. Clinical Outcomes The primary outcome was all-cause mortality within 30 days of ICU admission. The secondary outcome was all-cause mortality within 365 days of ICU admission. Serum Sodium Trajectory Classification In this study, we examined serum sodium trajectories during the initial five 24-hour periods following ICU admission as the exposure of interest. Based on baseline serum sodium tertiles, patients were categorized into T1, T2, and T3 groups. Simultaneously, group-based trajectory modeling (GBTM) was employed to identify latent trajectories of serum sodium levels over time. The analysis was implemented using the "traj" plugin in STATA 17.0 software (StataCorp LLC, College Station, TX, USA)[21, 22]. This model, based on the finite mixture modeling framework, assumes that the population comprises several latent subgroups with similar developmental trajectories. During model fitting, models with 1 to 6 groups were sequentially compared for goodness-of-fit. Optimal model selection was based on the following criteria: (1) minimum Bayesian Information Criterion (BIC) value; (2) average posterior probability (AvePP) ≥0.70 for each group; (3) odds of correct classification (OCC) ≥5.0; and (4) group size ≥5% of the total sample[23]. The time variable was defined as days 1 through 5 following ICU admission, with serum sodium concentration (mmol/L) serving as the continuous dependent variable. Cubic polynomial functions were used to model nonlinear trends. Following model convergence, each patient was assigned to their most likely trajectory group based on maximum posterior probability for subsequent between-group comparisons. Supplementary Table 1 shows the detailed fitting indexes of GBTM models with 2-5 classes, including log-likelihood, AIC, BIC, entropy value, group percentage. Supplementary Table 2 presents the average posterior probabilities(AvePP) of the 3-class model. Compared with models with fewer classes, the 3-class model demonstrated superior fit, evidenced by lower Akaike Information Criterion (AIC) and BIC values and entropy >0.7, indicating clearer classification. From a clinical interpretability standpoint, both persistent-hyponatremia (Trajectory 2, 18.18%) and hypernatremia-fluctuating (Trajectory 3, 21.84%) represented clinically meaningful high-risk subgroups. Although 4-class and 5-class models showed marginally better AIC values, the proportion of high-risk patients did not differ substantively. Adhering to the principle of parsimony, the 3-class model achieved optimal balance between complexity and interpretability by avoiding over-segmentation while maintaining clinical relevance. Therefore, serum sodium trajectories were categorized into three classes in this study. Statistical Analysis Continuous variables were expressed as mean ± standard deviation (SD) or median (interquartile range, IQR) based on normality testing (Kolmogorov-Smirnov test). Categorical variables were reported as frequencies and percentages. Between-group comparisons were performed using analysis of variance (ANOVA) or Kruskal-Wallis tests for continuous variables, and chi-square or Fisher's exact tests for categorical variables. This study examined the association between initial serum sodium levels at ICU admission (categorical variable) and subsequent 5-day serum sodium trajectories, and their impact on mortality risk. Hazard ratios (HRs) and 95% confidence intervals (CIs) for 30-day and 365-day mortality were calculated. Additionally, Kaplan-Meier survival analysis with log-rank tests was conducted to compare mortality risk across different serum sodium patterns. Univariable Cox regression (Model 1) was used to screen covariates associated with 30-day mortality (Supplementary Table 3), with variables showing P<0.05 subsequently included in multivariable models. Model 2 adjusted for sex, age, race, BMI, and respiratory rate. Model 3 further adjusted for disease severity scores (SOFA, APS III, SAPS II, OASIS, CCI), comorbidities (AKI, pneumonia, CKD, malignancy, heart failure, ischemic heart disease), vasoactive medications, and laboratory parameters (hemoglobin, anion gap, chloride). Subgroup analyses were conducted for age, gender, BMI, AKI, vasoactive drugs and COPD. Two sensitivity analyses were performed to verify the robustness of the study results. Given that the proportion of missing data for all covariates was less than 5% (Supplementary Table 4), we first imputed missing values using mean imputation and repeated the primary analysis. Subsequently, for the analyses of 30-day and 365-day mortality, we re-analyzed the data by including all patients diagnosed with SAD during their ICU stay. Data processing and analysis were conducted using DecisionLinnc 1.0 software and R (version 4.4.2). Statistical significance was set at two-tailed P<0.05. Results Baseline Characteristics The MIMIC-IV database provided data for 1,447 SAD patients, with the selection flowchart shown in Figure 1. Table 1 presents baseline characteristics of SAD patients classified by three serum sodium trajectory groups. The median patient age was 67.0 years, with male predominance (59.30%). A total of 711 patients (49.14%) had AKI, 260 (17.97%) had COPD, 430 (29.72%) had T2DM, and 469 (32.41%) had heart failure. Compared with the normal-stable group, the hypernatremia-fluctuating group had a higher proportion of AKI (57.59% vs. 46.54%), pneumonia (56.33% vs. 41.82%) and higher CCI scores, indicating more severe underlying diseases and organ dysfunction. Figure 2 illustrates the serum sodium trajectories for the three groups: Class 1 (normal-stable, 59.99%, n=868), Class 2 (persistent-hyponatremia, 18.18%, n=263,), and Class 3 (hypernatremia-fluctuating, 21.84%, n=316 ). Serum Sodium Trajectories and Mortality Figure 3 presents the Kaplan-Meier survival curves comparing all-cause mortality across the three serum sodium trajectory categories over 30-day (Panel A) and 365-day (Panel B) follow-up periods. The survival probability for Class 3 (hypernatremic-fluctuating, green dashed line) was significantly lower than that of both Class 1 (Normal-Stable, blue solid line) and Class 2 (persistent-hyponatremic, red dotted line), with log-rank tests indicating highly significant differences between Class 1 vs. Class 3 (P<0.001 at both time points) and Class 2 vs. Class 3 (P=0.010 at both time points). In contrast, no statistically significant difference in survival was observed between Class 1 and Class 2 (P=0.279 at 30 days; P=0.307 at 365 days), suggesting comparable mortality risks for patients with normal-stable versus persistent-hyponatremic trajectories. The "Number at risk" tables below each panel further illustrate the progressive decline in surviving participants within each group over time, with Class 3 consistently exhibiting the steepest reduction in at-risk population, corroborating its association with elevated mortality. Table 2 summarizes the association between serum sodium levels—categorized both by baseline tertiles and by trajectory classes—and 30-day and 365-day all-cause mortality among patients with SAD, across three progressively adjusted Cox proportional hazards models. In unadjusted analyses (Model 1), Class 3 (Hypernatremic-Fluctuating) was significantly associated with increased risk of both 30-day (HR=1.941, 95% CI: 1.466–2.57, P<0.001) and 365-day mortality (HR=1.867, 95% CI: 1.430–2.438, P<0.001), compared to Class 1 (Normal-Stable). These associations remained robust after sequential adjustment for sex, age, race, BMI, and respiratory rate in Model 2 and further for disease severity scores (SOFA, APS III, SAPS II, OASIS, CCI), comorbidities (AKI, pneumonia, CKD, malignancy, heart failure, ischemic heart disease), vasoactive medications, and laboratory parameters (hemoglobin, anion gap, chloride) in Model 3 (30-day HR=1.744, 95% CI: 1.296–2.347; 365-day HR=1.694, 95% CI: 1.277–2.247; both P<0.001). The multicollinearity test showed that all covariates in the model had VIF < 5, with no significant multicollinearity. In contrast, neither baseline serum sodium tertiles nor Class 2 (persistent-hyponatremic) showed statistically significant associations with mortality in any model. These findings indicate that dynamic fluctuations in serum sodium (Class 3), rather than static baseline levels or persistent-hyponatremia alone, are independently predictive of short- and long-term mortality in this cohort. Subgroup Analysis and Sensitivity Analysis Figure 4 presents subgroup analyses evaluating the consistency of the association between serum sodium trajectory classes and mortality across key clinical and demographic strata in patients with SAD, for both 30-day (Panel A) and 365-day (Panel B) outcomes. Across all prespecified subgroups—including age (0.05; range: 0.149–0.565 for 30-day; 0.145–0.462 for 365-day), indicating that the adverse prognostic impact of the hypernatremic-fluctuating trajectory is robust and largely unaffected by these baseline characteristics. Notably, the risk of mortality in Class 3 was more pronounced in young patients (<65 years, 30-day HR=2.912) and patients without vasoactive drug use (30-day HR=3.548), suggesting that these subgroups may require more intensive sodium management. These findings support the generalizability of Class 3 as an independent risk marker for mortality across diverse patient profiles within the SAD population. Sensitivity analyses demonstrated the robustness of the results based on imputed data (Supplementary Table 5). Furthermore, including all patients diagnosed with SAD during their ICU stay did not alter the association patterns between serum sodium trajectory classes and 30-day or 365-day mortality (Supplementary Table 6). Discussion This retrospective cohort analysis based on the MIMIC-IV database represents the first study to identify three clinically distinct serum sodium dynamic trajectories in critically ill patients with sepsis-associated delirium (SAD) and to confirm that the hypernatremia-fluctuating trajectory constitutes an independent risk factor for all-cause mortality at both 30 and 365 days. After comprehensive adjustment for demographics, disease severity, comorbidities, and therapeutic interventions, patients with the hypernatremia-fluctuating trajectory exhibited significantly elevated risks of short-term (HR=1.744, 95% CI: 1.296–2.347, P<0.001) and long-term (HR=1.694, 95% CI: 1.277–2.247, P<0.001) mortality compared to those with the normal-stable trajectory. This association remained consistent across all prespecified clinical subgroups, including age, sex, body mass index (BMI), acute kidney injury (AKI) status, vasoactive drug use, and chronic obstructive pulmonary disease (COPD), with no significant interactions observed (all P for interaction >0.05). These findings underscore the robust and universal clinical value of dynamic serum sodium monitoring for prognostic assessment in SAD patients and provide evidence-based guidance for electrolyte management in this population. Clinical Significance of Serum Sodium Trajectories and Comparison with Previous Studies Traditional investigations have predominantly focused on relationships between single-point serum sodium measurements and outcomes. The landmark study by Chewcharat et al.[13] employed GBTM to identify five serum sodium trajectories in hospitalized patients, demonstrating that any abnormal trajectory was associated with increased risks of in-hospital and one-year mortality compared with stable normonatremia. Our study extends these findings specifically to the SAD population: unadjusted analysis showed that the 30-day and 365-day mortality of the persistent-hyponatremia group (Class 2) were 17.11% and 19.01%, slightly higher than those of the normal-stable group (Class 1, 14.17% and 16.13%), but this association lost statistical significance after comprehensive adjustment for disease severity, comorbidities, and therapeutic interventions (30-day HR=0.831, 95% CI: 0.577-1.197, P=0.319; 365-day HR=0.823, 95% CI: 0.583-1.162, P=0.269). Conversely, the adverse prognostic effect of the hypernatremia-fluctuating trajectory (Class 3) persisted, suggesting that fluctuating patterns of hypernatremia may reflect more complex pathophysiological disturbances. This finding resonates with recent investigations in sepsis patients with lactic acidosis[15], which similarly utilized the MIMIC-IV database and identified lowest 30-day mortality among patients maintaining stable normonatremia (approximately 138 mmol/L), with fluctuating trajectories showing the worst outcomes. Notably, our study extends this observation to the SAD population and confirms its robustness over follow-up extending to one year, suggesting that sodium fluctuations may exert lasting deleterious effects on neurological function. Pathophysiological Mechanisms The pathophysiology of SAD encompasses multiple mechanisms including neuroinflammation, blood-brain barrier (BBB) disruption, and neurotransmitter imbalance[24–26]. Disruption of sodium homeostasis may exacerbate these pathological processes through several pathways: First, hypernatremia induces elevated plasma osmolality, causing cerebral cell dehydration and volume regulatory responses[27, 28]. When serum sodium fluctuates rapidly, intracellular organic osmolytes (such as inositol and taurine) cannot be reaccumulated promptly, potentially leading to osmotic demyelination syndrome (ODS)[27, 28]. Although traditionally associated with rapid correction of hyponatremia, accumulating evidence indicates that rapid sodium shifts in either direction can cause neurological injury[28]. In SAD patients, where the blood-brain barrier is already compromised by systemic inflammation, dramatic fluctuations in serum sodium more readily precipitate or worsen cerebral edema, neuronal injury, and cognitive dysfunction[6, 25]. Moreover, the damaged BBB in SAD patients leads to increased permeability of sodium ions into the brain interstitium, which further aggravates neuronal depolarization and abnormal excitability, forming a vicious circle of sodium fluctuation and delirium worsening Second, hypernatremia often indicates volume depletion and tissue hypoperfusion, which is particularly hazardous in septic patients[29]. A large-scale cohort study by Ng et al.[29] demonstrated that hypernatremia was associated with 27% increased ICU mortality and 52% increased hospital mortality. In our study, patients with hypernatremia-fluctuating trajectories showed higher vasoactive medication utilization (73.90% vs. 79.38%), potentially reflecting more severe hemodynamic instability. Furthermore, hypernatremia interacts with systemic inflammatory responses: inflammatory cytokines such as IL-6 and TNF-α can affect hypothalamic osmoregulatory centers[30], while hyperosmolar states themselves can activate microglia, amplifying neuroinflammation[31]. In SAD patients, the activated microglia can further secrete inflammatory factors, which not only worsen delirium but also inhibit renal water reabsorption, leading to further fluctuations in serum sodium. Third, serum sodium fluctuations may serve as markers of disease severity[15]. In our study, patients with hypernatremia-fluctuating trajectories exhibited higher SOFA and APS III scores, along with increased prevalence of acute kidney injury and heart failure. Dramatic variations in serum sodium likely reflect impaired renal concentrating and diluting capacity, neuroendocrine regulatory dysfunction, and complexity of therapeutic interventions[29]. Compared with baseline sodium tertile analysis, trajectory modeling captures this dynamic process, thereby maintaining independent prognostic predictive value even after multivariable adjustment. Strengths and Limitations This study possesses several strengths: First, the large sample size from the MIMIC-IV database, encompassing 1,447 SAD patients with comprehensive clinical data, ensures adequate statistical power and generalizability. Second, this represents the first application of group-based trajectory modeling to characterize serum sodium dynamics in SAD patients, systematically identifying trajectory types with prognostic significance and addressing the limitation of previous studies focusing solely on static sodium measurements. Third, multivariable Cox regression models comprehensively adjusted for potential confounders, with subgroup and sensitivity analyses validating result robustness. Several limitations should be acknowledged: First, as a retrospective observational study, causal relationships cannot be established, and residual confounding from unmeasured disease severity indicators may influence results[32]; Second, the single-center nature of the MIMIC-IV database limits generalizability, although the database encompasses racial diversity[16]; Third, individual variation in sodium measurement frequency and timing may introduce measurement error; while we required at least one measurement per 24-hour period, complete standardization of monitoring protocols was not feasible; Fourth, the absence of data regarding delirium severity, duration, and long-term cognitive function follow-up restricts comprehensive assessment of neurological outcomes; Fifth, classification of sodium trajectories relies on statistical modeling, and simplified criteria for clinical application may be needed; Sixth, we did not consider the impact of clinical interventions (e.g., fluid resuscitation, diuretic use) on serum sodium trajectories, which may be an important confounding facto; Seventh, we only analyzed the trajectory of serum sodium alone, and the combined prognostic value of sodium with other electrolytes (e.g., potassium, chloride) remains to be explored. Clinical Implications and Future Directions Our findings carry important implications for clinical management of critically ill SAD patients: Our findings carry important implications for clinical management of critically ill SAD patients: Clinicians should abandon single-point serum sodium measurement and implement dynamic monitoring of serum sodium at least once every 24 hours during the first 5 days of ICU admission for SAD patients, with the target of maintaining serum sodium within the normal range (135-145 mmol/L) and avoiding rapid fluctuations (≥5 mmol/L within 24 hours). Implementing high-frequency sodium testing during the first 5 days of ICU admission to promptly identify hypernatremia-fluctuating patterns. For patients with hypernatremia-fluctuating trajectories, early risk stratification and individualized interventions to maintain sodium homeostasis—such as optimized fluid management (isotonic crystalloid as the first choice), judicious use of diuretics/sodium supplementation, and active treatment of underlying complications including AKI and heart failure—should be considered to reduce mortality risk. Additionally, heightened intensity of sodium monitoring and management may be warranted for high-risk subgroups including younger and overweight/obese patients and patients without vasoactive drug use, as our subanalysis showed a more pronounced mortality risk in these subgroups. Future research should pursue several directions: First, multicenter prospective cohort studies are needed to validate the serum sodium trajectory types identified herein and their associations with SAD patient outcomes, while exploring optimal timing and strategies for intervention. Second, mechanistic investigations should elucidate molecular pathways linking abnormal sodium fluctuations to SAD prognosis, clarifying interactions among neuroinflammation, blood-brain barrier function, and sodium homeostasis. Third, associations between serum sodium trajectories and long-term cognitive function, quality of life, and other outcomes in SAD patients require evaluation to inform long-term prognostic management. Fourth, integration of serum sodium trajectories with other clinical indicators (such as inflammatory markers and organ function parameters) to construct prognostic prediction models for SAD patients may enhance accuracy of clinical risk stratification. Conclusions In summary, this study identified three serum sodium trajectories—normal-stable, persistent-hyponatremia, and hypernatremia-fluctuating—in critically ill SAD patients. The hypernatremia-fluctuating trajectory emerged as an independent risk factor for all-cause mortality at 30 and 365 days, with dynamic serum sodium trajectories demonstrating significantly superior prognostic value compared with single-point baseline measurements. Dynamic monitoring of serum sodium trajectories during the first 5 days of ICU admission, with prompt identification of hypernatremia-fluctuating characteristics, may enable early risk stratification and individualized sodium homeostasis management, holding promise for improving clinical outcomes in this vulnerable population. Abbreviations AKI: Acute kidney injury; APS III: Acute Physiology Score III; BMI: Body mass index; BIC: Bayesian Information Criterion; CAM-ICU: Confusion Assessment Method for the ICU; CCI: Charlson Comorbidity Index; CKD: Chronic kidney disease; COPD: Chronic obstructive pulmonary disease; CRRT: Continuous renal replacement therapy; CI: Confidence interval; DBP: Diastolic blood pressure; GBTM: Group-based trajectory modeling; HR: Hazard ratio; IQR: Interquartile range; ICU: Intensive care unit; ODS: Osmotic demyelination syndrome; OASIS: Oxford Acute Severity of Illness Score; OCC: Odds of correct classification; SAD: Sepsis-associated delirium; SAPS II: Simplified Acute Physiology Score II; SBP: Systolic blood pressure; SOFA: Sequential Organ Failure Assessment; SpO2: Peripheral oxygen saturation; SD: Standard deviation; T2DM: Type 2 diabetes mellitus; VIF: Variance inflation factor; WBC: White blood cell count. Declarations Consent for publication All the authors gave their written consent to publication. Conflict of Interest Statement: The authors declare no competing interests (financial or non-financial) relevant to this study. Funding Statement: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Ethics Approval Statement: This study uses de-identified, publicly available data from the MIMIC-IV database, approved by the Institutional Review Board (IRB) of Beth Israel Deaconess Medical Center (IRB No.: 2019P000002). No additional ethical review or informed consent was required. Author Contribution Statement: SD: Study conception and design, data extraction, statistical analysis, manuscript drafting; BL: Data quality control, laboratory validation; ZZ: Literature review, manuscript revision, subgroup analysis; GZ: Clinical data curation, reference formatting; PX: Study supervision, final model validation, manuscript critical revision and approval. All authors read and approved the final manuscript. Acknowledgements We sincerely thank the MIMIC-IV database team for providing the clinical data required for this study, and the Biostatistics Research Center of Fuwai Central China Cardiovascular Hospital for the guidance on statistical modeling and analysis. Data availability All relevant data supporting the findings of this study are available from https://physionet.org/content/mimiciv/3.1/. Access to the MIMIC-IV database requires completion of the National Institutes of Health (NIH) training course and CITI program certification References Stollings JL, Kotfis K, Chanques G, Pun BT, Pandharipande PP, Ely EW. Delirium in critical illness: clinical manifestations, outcomes, and management. Intensive Care Med. 2021;47:1089–103. https://doi.org/10.1007/s00134-021-06503-1. Ho M-H, Lee JJ, Lai PCK, Li PWC. Prevalence of delirium among critically ill patients who received extracorporeal membrane oxygenation therapy: A systematic review and proportional meta-analysis. Intensive Crit Care Nurs. 2023;79:103498. https://doi.org/10.1016/j.iccn.2023.103498. Goldberg TE, Chen C, Wang Y, Jung E, Swanson A, Ing C, et al. Association of Delirium With Long-term Cognitive Decline: A Meta-analysis. JAMA Neurol. 2020;77:1373–81. https://doi.org/10.1001/jamaneurol.2020.2273. Leighton SP, Herron JW, Jackson E, Sheridan M, Deligianni F, Cavanagh J. Delirium and the risk of developing dementia: a cohort study of 12 949 patients. J Neurol Neurosurg Psychiatry. 2022;93:822–7. https://doi.org/10.1136/jnnp-2022-328903. Gao Q, Hernandes MS. Sepsis-Associated Encephalopathy and Blood-Brain Barrier Dysfunction. Inflammation. 2021;44:2143–50. https://doi.org/10.1007/s10753-021-01501-3. Hong Y, Chen P, Gao J, Lin Y, Chen L, Shang X. Sepsis-associated encephalopathy: From pathophysiology to clinical management. Int Immunopharmacol. 2023;124 Pt A:110800. https://doi.org/10.1016/j.intimp.2023.110800. Darmon M, Diconne E, Souweine B, Ruckly S, Adrie C, Azoulay E, et al. Prognostic consequences of borderline dysnatremia: pay attention to minimal serum sodium change. Crit Care. 2013;17:R12. https://doi.org/10.1186/cc11937. Sakr Y, Rother S, Ferreira AMP, Ewald C, Dünisch P, Riedemmann N, et al. Fluctuations in serum sodium level are associated with an increased risk of death in surgical ICU patients. Crit Care Med. 2013;41:133–42. https://doi.org/10.1097/CCM.0b013e318265f576. Castello LM, Gavelli F, Baldrighi M, Salmi L, Mearelli F, Fiotti N, et al. Hypernatremia and moderate-to-severe hyponatremia are independent predictors of mortality in septic patients at emergency department presentation: A sub-group analysis of the need-speed trial. Eur J Intern Med. 2021;83:21–7. https://doi.org/10.1016/j.ejim.2020.10.003. Alpar S, Tatlıparmak AC. U-shaped relationship between sodium levels and short-term mortality in Sepsis. Am J Emerg Med. 2025;97:227–32. https://doi.org/10.1016/j.ajem.2025.08.003. Harrois A, Anstey JR, van der Jagt M, Taccone FS, Udy AA, Citerio G, et al. Variability in Serum Sodium Concentration and Prognostic Significance in Severe Traumatic Brain Injury: A Multicenter Observational Study. Neurocrit Care. 2021;34:899–907. https://doi.org/10.1007/s12028-020-01118-8. Wang X, Ma H, Chen W, Wen D, You C, Tao C, et al. The impact of serum sodium variability on surgical patients with aneurysmal subarachnoid hemorrhage. Neurosurg Rev. 2025;48:55. https://doi.org/10.1007/s10143-025-03212-x. Chewcharat A, Thongprayoon C, Cheungpasitporn W, Mao MA, Thirunavukkarasu S, Kashani KB. Trajectories of Serum Sodium on In-Hospital and 1-Year Survival among Hospitalized Patients. Clin J Am Soc Nephrol. 2020;15:600–7. https://doi.org/10.2215/CJN.12281019. S H, X L, B C, Y Z, Y L, T H. Association between serum sodium trajectory and mortality in patients with acute kidney injury: a retrospective cohort study. BMC nephrology. 2024;25. https://doi.org/10.1186/s12882-024-03586-y. Li H, Zhou Q, Nan Y, Liu C, Zhang Y. Group-based Trajectory Modeling of Serum Sodium and Survival in Sepsis Patients with Lactic Acidosis: Results from MIMIC-IV Database. Tohoku J Exp Med. 2025;265:123–34. https://doi.org/10.1620/tjem.2024.J091. Johnson AEW, Bulgarelli L, Shen L, Gayles A, Shammout A, Horng S, et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci Data. 2023;10:1. https://doi.org/10.1038/s41597-022-01899-x. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315:801–10. https://doi.org/10.1001/jama.2016.0287. Guenther U, Popp J, Koecher L, Muders T, Wrigge H, Ely EW, et al. Validity and reliability of the CAM-ICU Flowsheet to diagnose delirium in surgical ICU patients. J Crit Care. 2010;25:144–51. https://doi.org/10.1016/j.jcrc.2009.08.005. Diao Y, Yu X, Zhang Q, Chen X. The predictive value of confusion assessment method-intensive care unit and intensive care delirium screening checklist for delirium in critically ill patients in the intensive care unit: A systematic review and meta-analysis. Nurs Crit Care. 2024;29:1224–35. https://doi.org/10.1111/nicc.13064. Pedersen AB, Mikkelsen EM, Cronin-Fenton D, Kristensen NR, Pham TM, Pedersen L, et al. Missing data and multiple imputation in clinical epidemiological research. Clin Epidemiol. 2017;9:157–66. https://doi.org/10.2147/CLEP.S129785. Jones BL, Nagin DS. A note on a Stata plugin for estimating group-based trajectory models. Sociol Methods Res. 2013. https://doi.org/doi:10.1177/0049124113503141. Nagin DS. Group-Based Modeling of Development. 2005. Nagin DS, Odgers CL. Group-based trajectory modeling in clinical research. Annu Rev Clin Psychol. 2010;6:109–38. https://doi.org/10.1146/annurev.clinpsy.121208.131413. Kodali MC, Chen H, Liao F-F. Temporal unsnarling of brain’s acute neuroinflammatory transcriptional profiles reveals panendothelitis as the earliest event preceding microgliosis. Mol Psychiatry. 2021;26:3905–19. https://doi.org/10.1038/s41380-020-00955-5. Taylor J, Parker M, Casey CP, Tanabe S, Kunkel D, Rivera C, et al. Postoperative delirium and changes in the blood-brain barrier, neuroinflammation, and cerebrospinal fluid lactate: a prospective cohort study. Br J Anaesth. 2022;129:219–30. https://doi.org/10.1016/j.bja.2022.01.005. Mz X, Cx L, Lg Z, Y Y, Y W. Postoperative delirium, neuroinflammation, and influencing factors of postoperative delirium: A review. Medicine. 2023;102. https://doi.org/10.1097/MD.0000000000032991. Garg P, Aggarwal A, Malhotra R, Dhall S. Osmotic Demyelination Syndrome - Evolution of Extrapontine Before Pontine Myelinolysis on Magnetic Resonance Imaging. J Neurosci Rural Pract. 2019;10:126–35. https://doi.org/10.4103/jnrp.jnrp_240_18. See XY, Chang Y-C, Peng C-Y, Wang S-S, Chi K-Y, Chiang C-H, et al. Rate of Sodium Correction and Osmotic Demyelination Syndrome in Severe Hyponatremia: A Meta-Analysis. J Crit Care Med (Targu Mures). 2024;10:209–12. https://doi.org/10.2478/jccm-2024-0030. Ng PY, Cheung RYT, Ip A, Chan WM, Sin WC, Yap DY-H. A retrospective cohort study on the clinical outcomes of patients admitted to intensive care units with dysnatremia. Sci Rep. 2023;13:21236. https://doi.org/10.1038/s41598-023-48399-5. Smith RJ, Lachner C, Singh VP, Trivedi S, Khatua B, Cartin-Ceba R. Cytokine profiles in intensive care unit delirium. Acute Crit Care. 2022;37:415–28. https://doi.org/10.4266/acc.2021.01508. Piva S, McCreadie VA, Latronico N. Neuroinflammation in sepsis: sepsis associated delirium. Cardiovasc Hematol Disord Drug Targets. 2015;15:10–8. https://doi.org/10.2174/1871529x15666150108112452. Chen Y, Zong C, Zou L, Zhang Z, Yang T, Zong J, et al. A novel clinical prediction model for in-hospital mortality in sepsis patients complicated by ARDS: A MIMIC IV database and external validation study. Heliyon. 2024;10:e33337. https://doi.org/10.1016/j.heliyon.2024.e33337. Tables Table 1. Clinical Characteristics and Outcomes by Serum Sodium Trajectory in SAD Patients Variables Total(n=1,447) Class 1(n=868) Class 2(n=263) Class 3(n=316) P Age (years) 67.0 (56.0-79.0) 66.0 (55.0-78.0) 68.0 (58.0-78.5) 69.5 (57.0-81.0) 0.038 Gender, n(%) Female 589 (40.70) 344 (39.63) 107 (40.68) 138 (43.67) 0.457 Male 858 (59.30) 524 (60.37) 156 (59.32) 178 (56.33) Race (%) Black 133 (9.19) 70 (8.06) 22 (8.37) 41 (12.97) 0.129 White 1,123 (77.61) 680 (78.34) 208 (79.09) 235 (74.37) Others 191 (13.20) 118 (13.59) 33 (12.55) 40 (12.66) BMI 27.658 (24.049-32.665) 27.993 (24.42-33.207) 26.352 (23.016-31.887) 27.337 (23.903-31.029) 0.015 SOFA 7 (4-9) 7 (4-9) 7 (4-10) 7 (4-10) 0.952 APSIII 48 (37-63) 47 (36-61) 50 (37.5-65) 48 (37-64) 0.138 SAPS II 41 (32-50) 40 (31-50) 40 (32-50) 42 (33-51) 0.441 OASIS 36 (31-42) 36 (31-41.25) 36 (29-41) 36 (31-42) 0.100 CCI 5 (3-7) 5 (3-7) 6 (3-8) 6 (3-8) <0.001 Hypertension (%) No 856 (59.16) 508 (58.53) 150 (57.03) 198 (62.66) 0.327 Yes 591 (40.84) 360 (41.47) 113 (42.97) 118 (37.34) AKI(%) No 736 (50.86) 464 (53.46) 138 (52.47) 134 (42.41) 0.003 Yes 711 (49.14) 404 (46.54) 125 (47.53) 182 (57.59) Pneumonia (%) No 795 (54.94) 505 (58.18) 152 (57.79) 138 (43.67) <0.001 Yes 652 (45.06) 363 (41.82) 111 (42.21) 178 (56.33) Stroke (%) No 1,309 (90.46) 779 (89.75) 246 (93.54) 284 (89.87) 0.172 Yes 138 (9.54) 89 (10.25) 17 (6.46) 32 (10.13) CKD(%) No 1,147 (79.27) 701 (80.76) 208 (79.09) 238 (75.32) 0.123 Yes 300 (20.73) 167 (19.24) 55 (20.91) 78 (24.68) Cancer(%) No 1,254 (86.66) 766 (88.25) 221 (84.03) 267 (84.49) 0.093 Yes 193 (13.34) 102 (11.75) 42 (15.97) 49 (15.51) T2DM (%) No 1,017 (70.28) 618 (71.20) 172 (65.40) 227 (71.84) 0.156 Yes 430 (29.72) 250 (28.80) 91 (34.60) 89 (28.16) Heart failure (%) No 978 (67.59) 595 (68.55) 169 (64.26) 214 (67.72) 0.428 Yes 469 (32.41) 273 (31.45) 94 (35.74) 102 (32.28) Ischemic cardiomyopathy (%) No 934 (64.55) 574 (66.13) 156 (59.32) 204 (64.56) 0.129 Yes 513 (35.45) 294 (33.87) 107 (40.68) 112 (35.44) COPD (%) No 1,187 (82.03) 713 (82.14) 213 (80.99) 261 (82.59) 0.874 Yes 260 (17.97) 155 (17.86) 50 (19.01) 55 (17.41) Hemoglobin (g/dL) 10.4 (8.7-12.3) 10.5 (8.7-12.4) 10 (8.4-11.65) 10.6 (9-12.6) 0.001 Platelets (×103/µL) 181 (127-249) 186 (130-251) 181 (121.5-262) 174 (123-235) 0.190 WBC (×103/µL) 12.6 (8.6-17.2) 12.8 (8.4-17.725) 12.3 (8.8-16.65) 12.1 (8.875-16.6) 0.544 Anion gap 14 (12-17) 14 (12-17) 14 (12-18) 15 (13-17) 0.102 Chloride(mEq/L) 104 (100-108) 105 (101-108) 100 (94-104) 107 (103-111) <0.001 FBG (mg/dL) 134 (110-175) 136 (111-177) 130 (104-172.5) 137 (110-175) 0.856 Potassium (mEq/L) 4.1 (3.7-4.6) 4.1 (3.8-4.6) 4.2 (3.7-4.7) 4 (3.6-4.5) 0.002 Baseline sodium (mEq/L) 139 (136-142) 139 (136-141) 134 (131-137) 143 (140-146) <0.001 SBP(mmHg) 119 (103-138) 119 (102-138) 119 (100.5-138.5) 121 (106.75-141) 0.137 Heart rate(beats/min) 89 (77.5-105) 89 (78-105) 89 (77-104) 89.5 (77.75-105) 0.461 DBP(mmHg) 68 (57-81) 67 (57-79) 66 (54-81.5) 71.5 (60-85) 0.158 Respiration rate(breaths/min) 19 (16-23) 19 (16-24) 18 (15-23) 20 (16-24) 0.167 Temperature(f) 98.4 (97.8-99) 98.4 (97.8-99.1) 98.3 (97.7-98.8) 98.5 (98-99.2) 0.317 SpO2(%) 99 (96-100) 99 (96-100) 99 (96-100) 99 (96-100) 0.885 Invasive ventilation (%) No 51 (3.52) 27 (3.11) 9 (3.42) 15 (4.75) 0.400 Yes 1396 (96.48) 841 (96.89) 254 (96.58) 301 (95.25) Vasoactive drugs(%) No 351 (24.26) 201 (23.16) 54 (20.53) 96 (30.38) 0.011 Yes 1,096 (75.74) 667 (76.84) 209 (79.47) 220 (69.62) 30-day mortality(%) 249 (17.21) 123 (14.17) 45 (17.11) 81 (25.63) <0.001 365-day mortality(%) 278 (19.21) 140 (16.13) 50 (19.01) 88 (27.85) <0.001 Data are presented as median (interquartile range, IQR) or n (%). Abbreviations: BMI, body mass index; SOFA, Sequential Organ Failure Assessment; APS III, Acute Physiology Score III; SAPS II, Simplified Acute Physiology Score II; OASIS, Oxford Acute Severity of Illness Score; CCI, Charlson Comorbidity Index; AKI, acute kidney injury; CKD, chronic kidney disease; T2DM, type 2 diabetes mellitus; COPD, chronic obstructive pulmonary disease; WBC, white blood cell count; FBG, fasting blood glucose; SBP, systolic blood pressure; DBP, diastolic blood pressure; SpO2, peripheral oxygen saturation. Table 2 . Association of Serum Sodium Levels with 30-Day and 365-Day Mortality Among SAD Patients Outcome Model 1 Model 2 Model 3 HR (95% CI) P HR (95% CI) P HR (95% CI) P 30-day mortality Basic serum sodium level tertiles T 1 ref ref ref T 2 0.747 (0.557–1.003) 0.052 0.760 (0.566–1.020) 0.068 0.998 (0.721–1.383) 0.992 T 3 0.809 (0.596–1.099) 0.176 0.758 (0.557–1.034) 0.08 0.998 (0.679–1.465) 0.99 Serum sodium Trajectories Class 1 ref ref ref Class 2 1.208(0.858-1.699) 0.279 1.162 (0.825–1.637) 0.389 0.831 (0.577–1.197) 0.319 Class 3 1.941(1.466-2.57) <0.001 1.735 (1.307–2.305) <0.001 1.744 (1.296–2.347) <0.001 365-day mortality Basic serum sodium level tertiles T 1 ref ref ref T 2 0.827 (0.628–1.090) 0.177 0.842 (0.639–1.110) 0.223 1.123 (0.827–1.526) 0.458 T 3 0.817 (0.609–1.096) 0.178 0.771 (0.573–1.038) 0.086 1.035 (0.715–1.496) 0.857 Serum sodium Trajectories Class 1 ref ref ref Class 2 1.183 (0.856–1.634) 0.308 1.146 (0.829–1.585) 0.409 0.823 (0.583–1.162) 0.269 Class 3 1.867 (1.430–2.438) <0.001 1.684 (1.286–2.205) <0.001 1.694 (1.277–2.247) <0.001 T1: sodium<138.0, T2: 138.0≤sodium<142.0, T3: sodium≥142.0. Model 1: Unadjusted crude model. Model 2: Adjusted for sex, age, race, BMI, and respiratory rate. Model 3: Model 2 plus disease severity scores (SOFA, APS III, SAPS II, OASIS, CCI), comorbidities (AKI, pneumonia, CKD, malignancy, heart failure, ischemic heart disease), vasoactive medications, and laboratory parameters (hemoglobin, anion gap, chloride). Additional Declarations No competing interests reported. Supplementary Files SupplementaryMaterials.doc tables3.10.doc Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 19 Apr, 2026 Reviews received at journal 18 Apr, 2026 Reviewers agreed at journal 17 Apr, 2026 Reviewers agreed at journal 13 Apr, 2026 Reviewers agreed at journal 10 Apr, 2026 Reviewers invited by journal 08 Apr, 2026 Editor invited by journal 18 Mar, 2026 Editor assigned by journal 12 Mar, 2026 Submission checks completed at journal 12 Mar, 2026 First submitted to journal 05 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9036323","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":623830333,"identity":"6c7afef7-7017-409e-bd96-1d7e63079310","order_by":0,"name":"Shuyang Dai","email":"","orcid":"","institution":"Fuwai Central China Cardiovascular Hospital","correspondingAuthor":false,"prefix":"","firstName":"Shuyang","middleName":"","lastName":"Dai","suffix":""},{"id":623830334,"identity":"57199b7a-c903-4f45-822a-4e627869dd7e","order_by":1,"name":"Bingjie Li","email":"","orcid":"","institution":"Fuwai Central China Cardiovascular Hospital","correspondingAuthor":false,"prefix":"","firstName":"Bingjie","middleName":"","lastName":"Li","suffix":""},{"id":623830335,"identity":"3d3e3713-b949-4268-8453-a4ebf9ac6387","order_by":2,"name":"Zongshan Zhang","email":"","orcid":"","institution":"Fuwai Central China Cardiovascular Hospital","correspondingAuthor":false,"prefix":"","firstName":"Zongshan","middleName":"","lastName":"Zhang","suffix":""},{"id":623830336,"identity":"c5af2613-8afa-4cdb-99d2-1c6f6846d759","order_by":3,"name":"Gaoli Zhang","email":"","orcid":"","institution":"Fuwai Central China Cardiovascular Hospital","correspondingAuthor":false,"prefix":"","firstName":"Gaoli","middleName":"","lastName":"Zhang","suffix":""},{"id":623830337,"identity":"329157b4-4416-4a62-84c9-b91331983501","order_by":4,"name":"Poshi Xu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzUlEQVRIiWNgGAWjYBAC++PNBx98/GPDzM/eQKyeM8eSDWc2pLFL9hwgVsuNHDNh3obD/AYzEojUwTgjwYxx5o40aQPJxxtvMNTYRBPUwszzIO3BxzM2xubSacUWDMfSchsIaWFjTzhuOIMtLdlydo6ZBGPDYcJaeBgS26R52A7Xb7h5hkgtEhzJbNK8bYeZDW7wEKnFgOcYs+GMM2nMkj1AvyQQ4xcD9v6PDz5UgKLy8MYbH2psCGtB0S6RQIpyiBZSdYyCUTAKRsHIAADcg0JhMwA4FgAAAABJRU5ErkJggg==","orcid":"","institution":"Fuwai Central China Cardiovascular Hospital","correspondingAuthor":true,"prefix":"","firstName":"Poshi","middleName":"","lastName":"Xu","suffix":""}],"badges":[],"createdAt":"2026-03-05 05:53:46","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9036323/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9036323/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107246726,"identity":"8aca41dc-bacb-41f6-a1b4-6971e2492881","added_by":"auto","created_at":"2026-04-19 08:10:35","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":94486,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFlowchart of the selection of patients. \u003c/strong\u003eMIMIC-IV: Medical Information Mart for Intensive Care IV; ICU: intensive care unit; BMI, body mass index.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9036323/v1/4090be7252392e9ba1fa2854.png"},{"id":107246906,"identity":"84cc01b0-6e35-4c7a-8b53-918e7b09ea55","added_by":"auto","created_at":"2026-04-19 08:10:49","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":58864,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSerum sodium trajectories over the first 5 days of ICU admission in three identified groups of SAD patients. Class 1, normal-stable trajectory; Class 2, persistent-hyponatremia trajectory; Class 3, hypernatremia-fluctuating trajectory.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-9036323/v1/7f0e3ba0b5ec240dbc3c6557.png"},{"id":107246901,"identity":"37772ad0-0ac1-4d0d-9943-156a0e0b77a9","added_by":"auto","created_at":"2026-04-19 08:10:48","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":70997,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eKaplan-Meier survival curves for all-cause mortality according to serum sodium trajectory groups in SAD patients. (A) 30-day all-cause mortality; (B) 365-day all-cause mortality. Log-rank test was used for intergroup comparison.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-9036323/v1/930cf9715571e1ac3a2bd872.png"},{"id":107246841,"identity":"4566dd9c-5f55-4ccc-895f-0a8de2c6f570","added_by":"auto","created_at":"2026-04-19 08:10:36","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":215789,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSubgroup analysis of the association between serum sodium trajectories and all-cause mortality in SAD patients. (A) 30-day all-cause mortality; (B) 365-day all-cause mortality. P for interaction was calculated to assess the homogeneity of HRs across subgroups. Abbreviations: AKI, acute kidney injury; BMI, body mass index; COPD, chronic obstructive pulmonary disease.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-9036323/v1/a6e8c6f9286120b91f1b948f.png"},{"id":107246952,"identity":"c39b408f-f42d-4049-b5da-b91855f5e5ca","added_by":"auto","created_at":"2026-04-19 08:11:01","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1260838,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9036323/v1/67cae9ac-de80-4d00-9d54-749201d4c51c.pdf"},{"id":107246856,"identity":"fcc87184-c080-4cf4-93c9-0eeaf7d99472","added_by":"auto","created_at":"2026-04-19 08:10:37","extension":"doc","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":232960,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterials.doc","url":"https://assets-eu.researchsquare.com/files/rs-9036323/v1/2437ec09a0dfa583d9c19295.doc"},{"id":107246827,"identity":"5cf4fa45-6dca-42e3-8422-9cbbd36ee960","added_by":"auto","created_at":"2026-04-19 08:10:36","extension":"doc","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":192000,"visible":true,"origin":"","legend":"","description":"","filename":"tables3.10.doc","url":"https://assets-eu.researchsquare.com/files/rs-9036323/v1/7b34cc75631e6a4368fd0b98.doc"}],"financialInterests":"No competing interests reported.","formattedTitle":"Serum Sodium Trajectories and All-Cause Mortality in Critically Ill Patients with Sepsis-Associated Delirium: A Retrospective Analysis Based on the MIMIC-IV Database","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSepsis-associated delirium (SAD) represents one of the most common neurological complications of sepsis, affecting between 9% and 71% of critically ill septic patients, with incidence reaching 70%-87% among those requiring mechanical ventilation[1, 2]. Defined as an acute alteration in mental status characterized by inattention, disorganized thinking, and fluctuating levels of consciousness, SAD has been consistently associated with increased short-term mortality, prolonged ICU length of stay, and long-term cognitive impairment among survivors[3, 4]. Despite its substantial clinical significance, the pathophysiological mechanisms underlying SAD remain incompletely understood, with neuroinflammation, blood-brain barrier disruption, and neurotransmitter imbalance currently recognized as key contributing factors[5, 6].\u003c/p\u003e\n\u003cp\u003eElectrolyte disturbances, particularly dysnatremia, are highly prevalent in critically ill patients and increasingly recognized as important independent determinants of adverse outcomes[7, 8]. As the predominant extracellular cation, sodium plays a fundamental role in maintaining osmotic balance, regulating cell volume, and determining neuronal excitability. Both hyponatremia and hypernatremia are common in sepsis, with reported incidence ranging from 15% to 30% depending on the study population and diagnostic criteria employed[9, 10]. These electrolyte abnormalities may hold particular pathophysiological relevance in SAD, as rapid shifts in serum osmolality can exacerbate cerebral edema or precipitate osmotic demyelination, thereby worsening delirium and neurological outcomes[11, 12].\u0026nbsp;Notably, the blood-brain barrier of SAD patients is already damaged by neuroinflammation, making their brain tissue more sensitive to sodium fluctuations, yet the dynamic association between serum sodium trajectories and SAD prognosis has not been clarified.\u003c/p\u003e\n\u003cp\u003eAccumulating evidence suggests that dynamic patterns of serum sodium variation may carry greater prognostic significance than static single-point measurements. In a landmark study, Chewcharat et al.[13]applied group-based trajectory modeling (GBTM) to identify five distinct serum sodium trajectories in hospitalized patients, demonstrating that any deviation from the normal-stable pattern was associated with increased risks of in-hospital and one-year mortality. Subsequent investigations have extended these findings across diverse critical care populations, including patients with acute kidney injury[14], sepsis with lactic acidosis[15], and those requiring renal replacement therapy, consistently identifying fluctuating trajectories as harbingers of poor prognosis. Notably, Li et al.[15]recently demonstrated that sepsis patients with lactic acidosis who maintained stable normonatremia exhibited the lowest 30-day mortality, whereas those with fluctuating sodium trajectories experienced the worst outcomes. However, Li et al.\u0026apos;s study did not include delirium patients, and the prognostic impact of hyponatremia may be different from that in SAD population. In addition, studies on serum sodium trajectories in patients with traumatic brain injury have also confirmed the adverse prognostic value of sodium fluctuation[11], but the brain injury mechanism of traumatic brain injury is mechanical damage, which is different from the inflammatory brain injury of SAD, and it is unclear whether the characteristics and prognostic value of serum sodium trajectories are specific in SAD population.\u0026nbsp;To date, no study has specifically characterized the serum sodium trajectory subtypes in SAD patients nor verified their independent prognostic value after adjusting for comprehensive clinical covariates.\u003c/p\u003e\n\u003cp\u003eGiven the high incidence of both hyponatremia and delirium in sepsis, alongside their potentially shared pathophysiological mechanisms involving neuroinflammation and blood-brain barrier dysfunction, we hypothesized that distinct serum sodium trajectories would be identifiable in SAD patients and that unstable trajectories would correlate with increased mortality. Utilizing the MIMIC-IV database, this study aims to: (1) identify distinct serum sodium trajectory subgroups during the first 5 days following ICU admission in SAD patients; (2) evaluate the associations between these trajectories and all-cause mortality at 30 and 365 days; and (3) assess the consistency of these associations across clinically relevant subgroups, thereby informing evidence-based prognostic assessment and individualized electrolyte management strategies for this vulnerable population.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eData Source and Study Population\u003c/p\u003e\n\u003cp\u003eThis retrospective cohort study utilized data from the Medical Information Mart for Intensive Care IV (MIMIC-IV, version 3.1) database, a publicly available, de-identified critical care database containing comprehensive clinical data from patients admitted to Beth Israel Deaconess Medical Center in Boston, Massachusetts, between 2008 and 2022[16]. The database encompasses detailed information on demographics, vital signs, laboratory measurements, medications, procedures, and outcomes. One author (S.D.) completed the required training and obtained certification from the Collaborative Institutional Training Initiative (CITI) program (Certificate No. 14012091) to access the database.\u003c/p\u003e\n\u003cp\u003eThe study cohort comprised adult patients (\u0026ge;18 years) admitted to the ICU with documented diagnoses of both sepsis and delirium.\u0026nbsp;Delirium diagnosis was required to occur within 48 hours of ICU admission to exclude the interference of late-onset delirium on serum sodium trajectory. Sepsis was identified using International Classification of Diseases (ICD)-9 and ICD-10 codes (ICD-9: 995.91, 995.92; ICD-10: A40.x, A41.x), supplemented by a Sequential Organ Failure Assessment (SOFA) score \u0026ge;2, consistent with the Sepsis-3 criteria[17]. Delirium was defined using the Confusion Assessment Method for the ICU (CAM-ICU), with the diagnostic criteria extracted from the MIMIC-IV database\u0026apos;s delirium assessment module as previously validated[18, 19]. The following exclusion criteria were applied: (1) ICU length of stay \u0026lt;72 hours\u0026nbsp;(reason: patients with short ICU stay have insufficient serum sodium detection data to complete trajectory modeling, and the prognosis bias of such patients is large); (2) missing serum sodium measurements for any 24-hour period during the first 5 days of ICU admission; (3) receipt of continuous renal replacement therapy (CRRT) during the observation period; (4) multiple ICU admissions for sepsis (only the first admission was included); and (5) missing key covariates.\u003c/p\u003e\n\u003cp\u003eData Collection\u003c/p\u003e\n\u003cp\u003eData extraction was performed using PostgreSQL (version 15) and Navicat Premium (version 17) through structured query language. Extracted variables included: (1) Demographics: age, sex, race, body mass index (BMI); (2) Comorbidities: hypertension, type 2 diabetes mellitus, heart failure, chronic obstructive pulmonary disease (COPD), chronic kidney disease (CKD), pneumonia, malignancy, acute myocardial infarction, ischemic cardiomyopathy, ischemic stroke; (3) Laboratory parameters: white blood cell count (WBC), hemoglobin, platelet count, serum chloride, potassium, lactate, glucose, anion gap, and serum sodium (measured at least every\u0026nbsp;24\u0026nbsp;hours during the first\u0026nbsp;96\u0026nbsp;hours); (4) Vital signs: systolic blood pressure (SBP), diastolic blood pressure (DBP), heart rate,\u0026nbsp;respiration rate, body temperature, and oxygen saturation; (5) Disease severity scores: SOFA score, Simplified Acute Physiology Score II (SAPS II), Acute Physiology Score III (APS III), Oxford Acute Severity of Illness Score (OASIS), and Charlson Comorbidity Index (CCI); (6) Therapeutic interventions: mechanical ventilation, vasoactive medications; and (7) Outcomes: ICU mortality, hospital mortality, and 30-day survival status.\u003c/p\u003e\n\u003cp\u003eFor patients with multiple serum sodium measurements within a 24-hour period, the first recorded value was used. Variables with \u0026gt;20% missing data were excluded, while those with \u0026le;20% missing data were imputed using multiple imputation methods[20].\u0026nbsp;The variance inflation factor (VIF) was used to assess multicollinearity among covariates in the Cox proportional hazards model, with a VIF value \u0026lt; 5 indicating no significant multicollinearity.\u003c/p\u003e\n\u003cp\u003eClinical Outcomes\u003c/p\u003e\n\u003cp\u003eThe primary outcome was all-cause mortality within 30 days of ICU admission. The secondary outcome was all-cause mortality within 365 days of ICU admission.\u003c/p\u003e\n\u003cp\u003eSerum Sodium Trajectory Classification\u003c/p\u003e\n\u003cp\u003eIn this study, we examined serum sodium trajectories during the initial five 24-hour periods following ICU admission as the exposure of interest. Based on baseline serum sodium tertiles, patients were categorized into T1, T2, and T3 groups. Simultaneously, group-based trajectory modeling (GBTM) was employed to identify latent trajectories of serum sodium levels over time. The analysis was implemented using the \u0026quot;traj\u0026quot; plugin in STATA 17.0 software (StataCorp LLC, College Station, TX, USA)[21, 22]. This model, based on the finite mixture modeling framework, assumes that the population comprises several latent subgroups with similar developmental trajectories.\u003c/p\u003e\n\u003cp\u003eDuring model fitting, models with 1 to 6 groups were sequentially compared for goodness-of-fit. Optimal model selection was based on the following criteria: (1) minimum Bayesian Information Criterion (BIC) value; (2) average posterior probability (AvePP) \u0026ge;0.70 for each group; (3) odds of correct classification (OCC) \u0026ge;5.0; and (4) group size \u0026ge;5% of the total sample[23]. The time variable was defined as days 1 through 5 following ICU admission, with serum sodium concentration (mmol/L) serving as the continuous dependent variable. Cubic polynomial functions were used to model nonlinear trends. Following model convergence, each patient was assigned to their most likely trajectory group based on maximum posterior probability for subsequent between-group comparisons.\u003c/p\u003e\n\u003cp\u003eSupplementary Table 1\u0026nbsp;shows the detailed fitting indexes of GBTM models with 2-5 classes, including log-likelihood, AIC, BIC, entropy value, group percentage.\u0026nbsp;Supplementary Table 2\u0026nbsp;presents the\u0026nbsp;average posterior probabilities(AvePP)\u0026nbsp;of the\u0026nbsp;3-class model.\u0026nbsp;Compared with models with fewer classes, the 3-class model demonstrated superior fit, evidenced by lower Akaike Information Criterion (AIC) and BIC values and entropy \u0026gt;0.7, indicating clearer classification. From a clinical interpretability standpoint, both persistent-hyponatremia (Trajectory 2,\u0026nbsp;18.18%) and hypernatremia-fluctuating (Trajectory 3,\u0026nbsp;21.84%) represented clinically meaningful high-risk subgroups. Although 4-class and 5-class models showed marginally better AIC values, the proportion of high-risk patients did not differ substantively. Adhering to the principle of parsimony, the 3-class model achieved optimal balance between complexity and interpretability by avoiding over-segmentation while maintaining clinical relevance. Therefore, serum sodium trajectories were categorized into three classes in this study.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eStatistical Analysis\u003c/p\u003e\n\u003cp\u003eContinuous variables were expressed as mean \u0026plusmn; standard deviation (SD) or median (interquartile range, IQR) based on normality testing (Kolmogorov-Smirnov test). Categorical variables were reported as frequencies and percentages. Between-group comparisons were performed using analysis of variance (ANOVA) or Kruskal-Wallis tests for continuous variables, and chi-square or Fisher\u0026apos;s exact tests for categorical variables.\u003c/p\u003e\n\u003cp\u003eThis study examined the association between initial serum sodium levels at ICU admission (categorical variable) and subsequent 5-day serum sodium trajectories, and their impact on mortality risk. Hazard ratios (HRs) and 95% confidence intervals (CIs) for 30-day and 365-day mortality were calculated. Additionally, Kaplan-Meier survival analysis with log-rank tests was conducted to compare mortality risk across different serum sodium patterns. Univariable Cox regression (Model 1) was used to screen covariates associated with 30-day mortality (Supplementary Table 3), with variables showing P\u0026lt;0.05 subsequently included in multivariable models. Model 2 adjusted for sex, age, race, BMI, and respiratory rate. Model 3 further adjusted for disease severity scores (SOFA, APS III, SAPS II, OASIS, CCI), comorbidities (AKI, pneumonia, CKD, malignancy, heart failure, ischemic heart disease), vasoactive medications, and laboratory parameters (hemoglobin, anion gap, chloride). Subgroup analyses were conducted for age, gender, BMI, AKI, vasoactive drugs and COPD. Two sensitivity analyses were performed to verify the robustness of the study results. Given that the proportion of missing data for all covariates was less than 5% (Supplementary Table 4), we first imputed missing values using mean imputation and repeated the primary analysis. Subsequently, for the analyses of 30-day and 365-day mortality, we re-analyzed the data by including all patients diagnosed with SAD during their ICU stay. Data processing and analysis were conducted using DecisionLinnc 1.0 software and R (version 4.4.2). Statistical significance was set at two-tailed P\u0026lt;0.05.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eBaseline Characteristics\u003c/p\u003e\n\u003cp\u003eThe MIMIC-IV database provided data for 1,447 SAD patients, with the selection flowchart shown in Figure 1. Table 1 presents baseline characteristics of SAD patients classified by three serum sodium trajectory groups. The median patient age was 67.0 years, with male predominance (59.30%). A total of 711 patients (49.14%) had AKI, 260 (17.97%) had COPD, 430 (29.72%) had T2DM, and 469 (32.41%) had heart failure. Compared with the normal-stable group, the hypernatremia-fluctuating group had a higher proportion of AKI (57.59% vs. 46.54%), pneumonia (56.33% vs. 41.82%) and higher CCI scores, indicating more severe underlying diseases and organ dysfunction. Figure 2 illustrates the serum sodium trajectories for the three groups: Class 1 (normal-stable, 59.99%, n=868), Class 2 (persistent-hyponatremia, 18.18%, n=263,), and Class 3 (hypernatremia-fluctuating, 21.84%, n=316 ).\u003c/p\u003e\n\u003cp\u003eSerum Sodium Trajectories and Mortality\u003c/p\u003e\n\u003cp\u003eFigure 3 presents the Kaplan-Meier survival curves comparing all-cause mortality across the three serum sodium trajectory categories over 30-day (Panel A) and 365-day (Panel B) follow-up periods. The survival probability for Class 3 (hypernatremic-fluctuating, green dashed line) was significantly lower than that of both Class 1 (Normal-Stable, blue solid line) and Class 2 (persistent-hyponatremic, red dotted line), with log-rank tests indicating highly significant differences between Class 1 vs. Class 3 (P\u0026lt;0.001 at both time points) and Class 2 vs. Class 3 (P=0.010 at both time points). In contrast, no statistically significant difference in survival was observed between Class 1 and Class 2 (P=0.279 at 30 days; P=0.307 at 365 days), suggesting comparable mortality risks for patients with normal-stable versus persistent-hyponatremic trajectories. The \u0026quot;Number at risk\u0026quot; tables below each panel further illustrate the progressive decline in surviving participants within each group over time, with Class 3 consistently exhibiting the steepest reduction in at-risk population, corroborating its association with elevated mortality.\u003c/p\u003e\n\u003cp\u003eTable 2 summarizes the association between serum sodium levels\u0026mdash;categorized both by baseline tertiles and by trajectory classes\u0026mdash;and 30-day and 365-day all-cause mortality among patients with SAD, across three progressively adjusted Cox proportional hazards models. In unadjusted analyses (Model 1), Class 3 (Hypernatremic-Fluctuating) was significantly associated with increased risk of both 30-day (HR=1.941, 95% CI: 1.466\u0026ndash;2.57, P\u0026lt;0.001) and 365-day mortality (HR=1.867, 95% CI: 1.430\u0026ndash;2.438, P\u0026lt;0.001), compared to Class 1 (Normal-Stable). These associations remained robust after sequential adjustment for sex, age, race, BMI, and respiratory rate in Model 2 and further for disease severity scores (SOFA, APS III, SAPS II, OASIS, CCI), comorbidities (AKI, pneumonia, CKD, malignancy, heart failure, ischemic heart disease), vasoactive medications, and laboratory parameters (hemoglobin, anion gap, chloride) in Model 3 (30-day HR=1.744, 95% CI: 1.296\u0026ndash;2.347; 365-day HR=1.694, 95% CI: 1.277\u0026ndash;2.247; both P\u0026lt;0.001). The multicollinearity test showed that all covariates in the model had VIF \u0026lt; 5, with no significant multicollinearity. In contrast, neither baseline serum sodium tertiles nor Class 2 (persistent-hyponatremic) showed statistically significant associations with mortality in any model. These findings indicate that dynamic fluctuations in serum sodium (Class 3), rather than static baseline levels or persistent-hyponatremia alone, are independently predictive of short- and long-term mortality in this cohort.\u003c/p\u003e\n\u003cp\u003eSubgroup Analysis and Sensitivity Analysis\u003c/p\u003e\n\u003cp\u003eFigure 4 presents subgroup analyses evaluating the consistency of the association between serum sodium trajectory classes and mortality across key clinical and demographic strata in patients with SAD, for both 30-day (Panel A) and 365-day (Panel B) outcomes. Across all prespecified subgroups\u0026mdash;including age (0.05; range: 0.149\u0026ndash;0.565 for 30-day; 0.145\u0026ndash;0.462 for 365-day), indicating that the adverse prognostic impact of the hypernatremic-fluctuating trajectory is robust and largely unaffected by these baseline characteristics. Notably, the risk of mortality in Class 3 was more pronounced in young patients (\u0026lt;65 years, 30-day HR=2.912) and patients without vasoactive drug use (30-day HR=3.548), suggesting that these subgroups may require more intensive sodium management. These findings support the generalizability of Class 3 as an independent risk marker for mortality across diverse patient profiles within the SAD population. Sensitivity analyses demonstrated the robustness of the results based on imputed data (Supplementary Table 5). Furthermore, including all patients diagnosed with SAD during their ICU stay did not alter the association patterns between serum sodium trajectory classes and 30-day or 365-day mortality (Supplementary Table 6).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis retrospective cohort analysis based on the MIMIC-IV database represents the first study to identify three clinically distinct serum sodium dynamic trajectories in critically ill patients with sepsis-associated delirium (SAD) and to confirm that the hypernatremia-fluctuating trajectory constitutes an independent risk factor for all-cause mortality at both 30 and 365 days. After comprehensive adjustment for demographics, disease severity, comorbidities, and therapeutic interventions, patients with the hypernatremia-fluctuating trajectory exhibited significantly elevated risks of short-term (HR=1.744, 95% CI: 1.296\u0026ndash;2.347, P\u0026lt;0.001) and long-term (HR=1.694, 95% CI: 1.277\u0026ndash;2.247, P\u0026lt;0.001) mortality compared to those with the normal-stable trajectory. This association remained consistent across all prespecified clinical subgroups, including age, sex, body mass index (BMI), acute kidney injury (AKI) status, vasoactive drug use, and chronic obstructive pulmonary disease (COPD), with no significant interactions observed (all P for interaction \u0026gt;0.05). These findings underscore the robust and universal clinical value of dynamic serum sodium monitoring for prognostic assessment in SAD patients and provide evidence-based guidance for electrolyte management in this population.\u003c/p\u003e\n\u003cp\u003eClinical Significance of Serum Sodium Trajectories and Comparison with Previous Studies\u003c/p\u003e\n\u003cp\u003eTraditional investigations have predominantly focused on relationships between single-point serum sodium measurements and outcomes. The landmark study by Chewcharat et al.[13]\u0026nbsp;employed GBTM to identify five serum sodium trajectories in hospitalized patients, demonstrating that any abnormal trajectory was associated with increased risks of in-hospital and one-year mortality compared with stable normonatremia. Our study extends these findings specifically to the SAD population: unadjusted analysis showed that the 30-day and 365-day mortality of the persistent-hyponatremia group (Class 2) were 17.11% and 19.01%, slightly higher than those of the normal-stable group (Class 1, 14.17% and 16.13%), but this association lost statistical significance after comprehensive adjustment for disease severity, comorbidities, and therapeutic interventions (30-day HR=0.831, 95% CI: 0.577-1.197, P=0.319; 365-day HR=0.823, 95% CI: 0.583-1.162, P=0.269). Conversely, the adverse prognostic effect of the hypernatremia-fluctuating trajectory (Class 3) persisted, suggesting that fluctuating patterns of hypernatremia may reflect more complex pathophysiological disturbances. This finding resonates with recent investigations in sepsis patients with lactic acidosis[15], which similarly utilized the MIMIC-IV database and identified lowest 30-day mortality among patients maintaining stable normonatremia (approximately 138 mmol/L), with fluctuating trajectories showing the worst outcomes. Notably, our study extends this observation to the SAD population and confirms its robustness over follow-up extending to one year, suggesting that sodium fluctuations may exert lasting deleterious effects on neurological function.\u003c/p\u003e\n\u003cp\u003ePathophysiological Mechanisms\u003c/p\u003e\n\u003cp\u003eThe pathophysiology of SAD encompasses multiple mechanisms including neuroinflammation, blood-brain barrier (BBB) disruption, and neurotransmitter imbalance[24\u0026ndash;26]. Disruption of sodium homeostasis may exacerbate these pathological processes through several pathways:\u003c/p\u003e\n\u003cp\u003eFirst, hypernatremia induces elevated plasma osmolality, causing cerebral cell dehydration and volume regulatory responses[27, 28]. When serum sodium fluctuates rapidly, intracellular organic osmolytes (such as inositol and taurine) cannot be reaccumulated promptly, potentially leading to osmotic demyelination syndrome (ODS)[27, 28]. Although traditionally associated with rapid correction of hyponatremia, accumulating evidence indicates that rapid sodium shifts in either direction can cause neurological injury[28]. In SAD patients, where the blood-brain barrier is already compromised by systemic inflammation, dramatic fluctuations in serum sodium more readily precipitate or worsen cerebral edema, neuronal injury, and cognitive dysfunction[6, 25].\u0026nbsp;Moreover, the damaged BBB in SAD patients leads to increased permeability of sodium ions into the brain interstitium, which further aggravates neuronal depolarization and abnormal excitability, forming a vicious circle of sodium fluctuation and delirium worsening\u003c/p\u003e\n\u003cp\u003eSecond, hypernatremia often indicates volume depletion and tissue hypoperfusion, which is particularly hazardous in septic patients[29]. A large-scale cohort study by Ng et al.[29]\u0026nbsp;demonstrated that hypernatremia was associated with 27% increased ICU mortality and 52% increased hospital mortality. In our study, patients with hypernatremia-fluctuating trajectories showed higher vasoactive medication utilization (73.90% vs. 79.38%), potentially reflecting more severe hemodynamic instability. Furthermore, hypernatremia interacts with systemic inflammatory responses: inflammatory cytokines such as IL-6 and TNF-\u0026alpha; can affect hypothalamic osmoregulatory centers[30], while hyperosmolar states themselves can activate microglia, amplifying neuroinflammation[31].\u0026nbsp;In SAD patients, the activated microglia can further secrete inflammatory factors, which not only worsen delirium but also inhibit renal water reabsorption, leading to further fluctuations in serum sodium.\u003c/p\u003e\n\u003cp\u003eThird, serum sodium fluctuations may serve as markers of disease severity[15]. In our study, patients with hypernatremia-fluctuating trajectories exhibited higher SOFA and APS III scores, along with increased prevalence of acute kidney injury and heart failure. Dramatic variations in serum sodium likely reflect impaired renal concentrating and diluting capacity, neuroendocrine regulatory dysfunction, and complexity of therapeutic interventions[29]. Compared with baseline sodium tertile analysis, trajectory modeling captures this dynamic process, thereby maintaining independent prognostic predictive value even after multivariable adjustment.\u003c/p\u003e\n\u003cp\u003eStrengths and Limitations\u003c/p\u003e\n\u003cp\u003eThis study possesses several strengths: First, the large sample size from the MIMIC-IV database, encompassing\u0026nbsp;1,447\u0026nbsp;SAD patients with comprehensive clinical data, ensures adequate statistical power and generalizability. Second, this represents the first application of group-based trajectory modeling to characterize serum sodium dynamics in SAD patients, systematically identifying trajectory types with prognostic significance and addressing the limitation of previous studies focusing solely on static sodium measurements. Third, multivariable Cox regression models comprehensively adjusted for potential confounders, with subgroup and sensitivity analyses validating result robustness.\u003c/p\u003e\n\u003cp\u003eSeveral limitations should be acknowledged: First, as a retrospective observational study, causal relationships cannot be established, and residual confounding from unmeasured disease severity indicators may influence results[32];\u0026nbsp;Second, the single-center nature of the MIMIC-IV database limits generalizability, although the database encompasses racial diversity[16];\u0026nbsp;Third, individual variation in sodium measurement frequency and timing may introduce measurement error; while we required at least one measurement per 24-hour period, complete standardization of monitoring protocols was not feasible;\u0026nbsp;Fourth, the absence of data regarding delirium severity, duration, and long-term cognitive function follow-up restricts comprehensive assessment of neurological outcomes;\u0026nbsp;Fifth, classification of sodium trajectories relies on statistical modeling, and simplified criteria for clinical application may be needed; Sixth, we did not consider the impact of clinical interventions (e.g., fluid resuscitation, diuretic use) on serum sodium trajectories, which may be an important confounding facto; Seventh, we only analyzed the trajectory of serum sodium alone, and the combined prognostic value of sodium with other electrolytes (e.g., potassium, chloride) remains to be explored.\u003c/p\u003e\n\u003cp\u003eClinical Implications and Future Directions\u003c/p\u003e\n\u003cp\u003eOur findings carry important implications for clinical management of critically ill SAD patients: Our findings carry important implications for clinical management of critically ill SAD patients: Clinicians should abandon single-point serum sodium measurement and implement dynamic monitoring of serum sodium at least once every 24 hours during the first 5 days of ICU admission for SAD patients, with the target of maintaining serum sodium within the normal range (135-145 mmol/L) and avoiding rapid fluctuations (\u0026ge;5 mmol/L within 24 hours). Implementing high-frequency sodium testing during the first 5 days of ICU admission to promptly identify hypernatremia-fluctuating patterns. For patients with hypernatremia-fluctuating trajectories, early risk stratification and individualized interventions to maintain sodium homeostasis\u0026mdash;such as optimized fluid management (isotonic crystalloid as the first choice), judicious use of diuretics/sodium supplementation, and active treatment of underlying complications including AKI and heart failure\u0026mdash;should be considered to reduce mortality risk. Additionally, heightened intensity of sodium monitoring and management may be warranted for high-risk subgroups including younger and overweight/obese patients and patients without vasoactive drug use, as our subanalysis showed a more pronounced mortality risk in these subgroups.\u003c/p\u003e\n\u003cp\u003eFuture research should pursue several directions: First, multicenter prospective cohort studies are needed to validate the serum sodium trajectory types identified herein and their associations with SAD patient outcomes, while exploring optimal timing and strategies for intervention. Second, mechanistic investigations should elucidate molecular pathways linking abnormal sodium fluctuations to SAD prognosis, clarifying interactions among neuroinflammation, blood-brain barrier function, and sodium homeostasis. Third, associations between serum sodium trajectories and long-term cognitive function, quality of life, and other outcomes in SAD patients require evaluation to inform long-term prognostic management. Fourth, integration of serum sodium trajectories with other clinical indicators (such as inflammatory markers and organ function parameters) to construct prognostic prediction models for SAD patients may enhance accuracy of clinical risk stratification.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn summary, this study identified three serum sodium trajectories\u0026mdash;normal-stable, persistent-hyponatremia, and hypernatremia-fluctuating\u0026mdash;in critically ill SAD patients. The hypernatremia-fluctuating trajectory emerged as an independent risk factor for all-cause mortality at 30 and 365 days, with dynamic serum sodium trajectories demonstrating significantly superior prognostic value compared with single-point baseline measurements. Dynamic monitoring of serum sodium trajectories during the first 5 days of ICU admission, with prompt identification of hypernatremia-fluctuating characteristics, may enable early risk stratification and individualized sodium homeostasis management, holding promise for improving clinical outcomes in this vulnerable population.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eAKI: Acute kidney injury;\u003c/p\u003e\n\u003cp\u003eAPS III: Acute Physiology Score III;\u003c/p\u003e\n\u003cp\u003eBMI: Body mass index;\u003c/p\u003e\n\u003cp\u003eBIC: Bayesian Information Criterion;\u003c/p\u003e\n\u003cp\u003eCAM-ICU: Confusion Assessment Method for the ICU;\u003c/p\u003e\n\u003cp\u003eCCI: Charlson Comorbidity Index;\u003c/p\u003e\n\u003cp\u003eCKD: Chronic kidney disease;\u003c/p\u003e\n\u003cp\u003eCOPD: Chronic obstructive pulmonary disease;\u003c/p\u003e\n\u003cp\u003eCRRT: Continuous renal replacement therapy;\u003c/p\u003e\n\u003cp\u003eCI: Confidence interval;\u003c/p\u003e\n\u003cp\u003eDBP: Diastolic blood pressure;\u003c/p\u003e\n\u003cp\u003eGBTM: Group-based trajectory modeling;\u003c/p\u003e\n\u003cp\u003eHR: Hazard ratio;\u003c/p\u003e\n\u003cp\u003eIQR: Interquartile range;\u003c/p\u003e\n\u003cp\u003eICU: Intensive care unit;\u003c/p\u003e\n\u003cp\u003eODS: Osmotic demyelination syndrome;\u003c/p\u003e\n\u003cp\u003eOASIS: Oxford Acute Severity of Illness Score;\u003c/p\u003e\n\u003cp\u003eOCC: Odds of correct classification;\u003c/p\u003e\n\u003cp\u003eSAD: Sepsis-associated delirium;\u003c/p\u003e\n\u003cp\u003eSAPS II: Simplified Acute Physiology Score II;\u003c/p\u003e\n\u003cp\u003eSBP: Systolic blood pressure;\u003c/p\u003e\n\u003cp\u003eSOFA: Sequential Organ Failure Assessment;\u003c/p\u003e\n\u003cp\u003eSpO2: Peripheral oxygen saturation;\u003c/p\u003e\n\u003cp\u003eSD: Standard deviation;\u003c/p\u003e\n\u003cp\u003eT2DM: Type 2 diabetes mellitus;\u003c/p\u003e\n\u003cp\u003eVIF: Variance inflation factor;\u003c/p\u003e\n\u003cp\u003eWBC: White blood cell count.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll the authors gave their written consent to publication.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eConflict of Interest Statement:\u0026nbsp;\u003c/strong\u003eThe authors declare no competing interests (financial or non-financial) relevant to this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding Statement:\u003c/strong\u003e This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Approval Statement:\u0026nbsp;\u003c/strong\u003eThis study uses de-identified, publicly available data from the MIMIC-IV database, approved by the Institutional Review Board (IRB) of Beth Israel Deaconess Medical Center (IRB No.: 2019P000002). No additional ethical review or informed consent was required.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contribution Statement:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSD: Study conception and design, data extraction, statistical analysis, manuscript drafting; BL: Data quality control, laboratory validation; ZZ: Literature review, manuscript revision, subgroup analysis; GZ: Clinical data curation, reference formatting; PX: Study supervision, final model validation, manuscript critical revision and approval. All authors read and approved the final manuscript.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe sincerely thank the MIMIC-IV database team for providing the clinical data required for this study, and the Biostatistics Research Center of Fuwai Central China Cardiovascular Hospital for the guidance on statistical modeling and analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll relevant data supporting the findings of this study are available from https://physionet.org/content/mimiciv/3.1/. Access to the MIMIC-IV database requires completion of the National Institutes of Health (NIH) training course and CITI program certification\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eStollings JL, Kotfis K, Chanques G, Pun BT, Pandharipande PP, Ely EW. Delirium in critical illness: clinical manifestations, outcomes, and management. Intensive Care Med. 2021;47:1089\u0026ndash;103. https://doi.org/10.1007/s00134-021-06503-1.\u003c/li\u003e\n\u003cli\u003eHo M-H, Lee JJ, Lai PCK, Li PWC. Prevalence of delirium among critically ill patients who received extracorporeal membrane oxygenation therapy: A systematic review and proportional meta-analysis. Intensive Crit Care Nurs. 2023;79:103498. https://doi.org/10.1016/j.iccn.2023.103498.\u003c/li\u003e\n\u003cli\u003eGoldberg TE, Chen C, Wang Y, Jung E, Swanson A, Ing C, et al. Association of Delirium With Long-term Cognitive Decline: A Meta-analysis. JAMA Neurol. 2020;77:1373\u0026ndash;81. https://doi.org/10.1001/jamaneurol.2020.2273.\u003c/li\u003e\n\u003cli\u003eLeighton SP, Herron JW, Jackson E, Sheridan M, Deligianni F, Cavanagh J. Delirium and the risk of developing dementia: a cohort study of 12 949 patients. J Neurol Neurosurg Psychiatry. 2022;93:822\u0026ndash;7. https://doi.org/10.1136/jnnp-2022-328903.\u003c/li\u003e\n\u003cli\u003eGao Q, Hernandes MS. Sepsis-Associated Encephalopathy and Blood-Brain Barrier Dysfunction. Inflammation. 2021;44:2143\u0026ndash;50. https://doi.org/10.1007/s10753-021-01501-3.\u003c/li\u003e\n\u003cli\u003eHong Y, Chen P, Gao J, Lin Y, Chen L, Shang X. Sepsis-associated encephalopathy: From pathophysiology to clinical management. Int Immunopharmacol. 2023;124 Pt A:110800. https://doi.org/10.1016/j.intimp.2023.110800.\u003c/li\u003e\n\u003cli\u003eDarmon M, Diconne E, Souweine B, Ruckly S, Adrie C, Azoulay E, et al. Prognostic consequences of borderline dysnatremia: pay attention to minimal serum sodium change. Crit Care. 2013;17:R12. https://doi.org/10.1186/cc11937.\u003c/li\u003e\n\u003cli\u003eSakr Y, Rother S, Ferreira AMP, Ewald C, D\u0026uuml;nisch P, Riedemmann N, et al. Fluctuations in serum sodium level are associated with an increased risk of death in surgical ICU patients. Crit Care Med. 2013;41:133\u0026ndash;42. https://doi.org/10.1097/CCM.0b013e318265f576.\u003c/li\u003e\n\u003cli\u003eCastello LM, Gavelli F, Baldrighi M, Salmi L, Mearelli F, Fiotti N, et al. Hypernatremia and moderate-to-severe hyponatremia are independent predictors of mortality in septic patients at emergency department presentation: A sub-group analysis of the need-speed trial. Eur J Intern Med. 2021;83:21\u0026ndash;7. https://doi.org/10.1016/j.ejim.2020.10.003.\u003c/li\u003e\n\u003cli\u003eAlpar S, Tatlıparmak AC. U-shaped relationship between sodium levels and short-term mortality in Sepsis. Am J Emerg Med. 2025;97:227\u0026ndash;32. https://doi.org/10.1016/j.ajem.2025.08.003.\u003c/li\u003e\n\u003cli\u003eHarrois A, Anstey JR, van der Jagt M, Taccone FS, Udy AA, Citerio G, et al. Variability in Serum Sodium Concentration and Prognostic Significance in Severe Traumatic Brain Injury: A Multicenter Observational Study. Neurocrit Care. 2021;34:899\u0026ndash;907. https://doi.org/10.1007/s12028-020-01118-8.\u003c/li\u003e\n\u003cli\u003eWang X, Ma H, Chen W, Wen D, You C, Tao C, et al. The impact of serum sodium variability on surgical patients with aneurysmal subarachnoid hemorrhage. Neurosurg Rev. 2025;48:55. https://doi.org/10.1007/s10143-025-03212-x.\u003c/li\u003e\n\u003cli\u003eChewcharat A, Thongprayoon C, Cheungpasitporn W, Mao MA, Thirunavukkarasu S, Kashani KB. Trajectories of Serum Sodium on In-Hospital and 1-Year Survival among Hospitalized Patients. Clin J Am Soc Nephrol. 2020;15:600\u0026ndash;7. https://doi.org/10.2215/CJN.12281019.\u003c/li\u003e\n\u003cli\u003eS H, X L, B C, Y Z, Y L, T H. Association between serum sodium trajectory and mortality in patients with acute kidney injury: a retrospective cohort study. BMC nephrology. 2024;25. https://doi.org/10.1186/s12882-024-03586-y.\u003c/li\u003e\n\u003cli\u003eLi H, Zhou Q, Nan Y, Liu C, Zhang Y. Group-based Trajectory Modeling of Serum Sodium and Survival in Sepsis Patients with Lactic Acidosis: Results from MIMIC-IV Database. Tohoku J Exp Med. 2025;265:123\u0026ndash;34. https://doi.org/10.1620/tjem.2024.J091.\u003c/li\u003e\n\u003cli\u003eJohnson AEW, Bulgarelli L, Shen L, Gayles A, Shammout A, Horng S, et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci Data. 2023;10:1. https://doi.org/10.1038/s41597-022-01899-x.\u003c/li\u003e\n\u003cli\u003eSinger M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315:801\u0026ndash;10. https://doi.org/10.1001/jama.2016.0287.\u003c/li\u003e\n\u003cli\u003eGuenther U, Popp J, Koecher L, Muders T, Wrigge H, Ely EW, et al. Validity and reliability of the CAM-ICU Flowsheet to diagnose delirium in surgical ICU patients. J Crit Care. 2010;25:144\u0026ndash;51. https://doi.org/10.1016/j.jcrc.2009.08.005.\u003c/li\u003e\n\u003cli\u003eDiao Y, Yu X, Zhang Q, Chen X. The predictive value of confusion assessment method-intensive care unit and intensive care delirium screening checklist for delirium in critically ill patients in the intensive care unit: A systematic review and meta-analysis. Nurs Crit Care. 2024;29:1224\u0026ndash;35. https://doi.org/10.1111/nicc.13064.\u003c/li\u003e\n\u003cli\u003ePedersen AB, Mikkelsen EM, Cronin-Fenton D, Kristensen NR, Pham TM, Pedersen L, et al. Missing data and multiple imputation in clinical epidemiological research. Clin Epidemiol. 2017;9:157\u0026ndash;66. https://doi.org/10.2147/CLEP.S129785.\u003c/li\u003e\n\u003cli\u003eJones BL, Nagin DS. A note on a Stata plugin for estimating group-based trajectory models. Sociol Methods Res. 2013. https://doi.org/doi:10.1177/0049124113503141.\u003c/li\u003e\n\u003cli\u003eNagin DS. Group-Based Modeling of Development. 2005.\u003c/li\u003e\n\u003cli\u003eNagin DS, Odgers CL. Group-based trajectory modeling in clinical research. Annu Rev Clin Psychol. 2010;6:109\u0026ndash;38. https://doi.org/10.1146/annurev.clinpsy.121208.131413.\u003c/li\u003e\n\u003cli\u003eKodali MC, Chen H, Liao F-F. Temporal unsnarling of brain\u0026rsquo;s acute neuroinflammatory transcriptional profiles reveals panendothelitis as the earliest event preceding microgliosis. Mol Psychiatry. 2021;26:3905\u0026ndash;19. https://doi.org/10.1038/s41380-020-00955-5.\u003c/li\u003e\n\u003cli\u003eTaylor J, Parker M, Casey CP, Tanabe S, Kunkel D, Rivera C, et al. Postoperative delirium and changes in the blood-brain barrier, neuroinflammation, and cerebrospinal fluid lactate: a prospective cohort study. Br J Anaesth. 2022;129:219\u0026ndash;30. https://doi.org/10.1016/j.bja.2022.01.005.\u003c/li\u003e\n\u003cli\u003eMz X, Cx L, Lg Z, Y Y, Y W. Postoperative delirium, neuroinflammation, and influencing factors of postoperative delirium: A review. Medicine. 2023;102. https://doi.org/10.1097/MD.0000000000032991.\u003c/li\u003e\n\u003cli\u003eGarg P, Aggarwal A, Malhotra R, Dhall S. Osmotic Demyelination Syndrome - Evolution of Extrapontine Before Pontine Myelinolysis on Magnetic Resonance Imaging. J Neurosci Rural Pract. 2019;10:126\u0026ndash;35. https://doi.org/10.4103/jnrp.jnrp_240_18.\u003c/li\u003e\n\u003cli\u003eSee XY, Chang Y-C, Peng C-Y, Wang S-S, Chi K-Y, Chiang C-H, et al. Rate of Sodium Correction and Osmotic Demyelination Syndrome in Severe Hyponatremia: A Meta-Analysis. J Crit Care Med (Targu Mures). 2024;10:209\u0026ndash;12. https://doi.org/10.2478/jccm-2024-0030.\u003c/li\u003e\n\u003cli\u003eNg PY, Cheung RYT, Ip A, Chan WM, Sin WC, Yap DY-H. A retrospective cohort study on the clinical outcomes of patients admitted to intensive care units with dysnatremia. Sci Rep. 2023;13:21236. https://doi.org/10.1038/s41598-023-48399-5.\u003c/li\u003e\n\u003cli\u003eSmith RJ, Lachner C, Singh VP, Trivedi S, Khatua B, Cartin-Ceba R. Cytokine profiles in intensive care unit delirium. Acute Crit Care. 2022;37:415\u0026ndash;28. https://doi.org/10.4266/acc.2021.01508.\u003c/li\u003e\n\u003cli\u003ePiva S, McCreadie VA, Latronico N. Neuroinflammation in sepsis: sepsis associated delirium. Cardiovasc Hematol Disord Drug Targets. 2015;15:10\u0026ndash;8. https://doi.org/10.2174/1871529x15666150108112452.\u003c/li\u003e\n\u003cli\u003eChen Y, Zong C, Zou L, Zhang Z, Yang T, Zong J, et al. A novel clinical prediction model for in-hospital mortality in sepsis patients complicated by ARDS: A MIMIC IV database and external validation study. Heliyon. 2024;10:e33337. https://doi.org/10.1016/j.heliyon.2024.e33337.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cstrong\u003eTable 1. Clinical Characteristics and Outcomes by Serum Sodium Trajectory in SAD Patients\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 119px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariables\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTotal(n=1,447)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eClass 1(n=868)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eClass 2(n=263)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eClass 3(n=316)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eP\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eAge (years)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e67.0 (56.0-79.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e66.0 (55.0-78.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e68.0 (58.0-78.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e69.5 (57.0-81.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.038\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eGender, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e589 (40.70)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e344 (39.63)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e107 (40.68)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e138 (43.67)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.457\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e858 (59.30)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e524 (60.37)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e156 (59.32)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e178 (56.33)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eRace (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eBlack\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e133 (9.19)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e70 (8.06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e22 (8.37)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e41 (12.97)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.129\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eWhite\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e1,123 (77.61)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e680 (78.34)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e208 (79.09)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e235 (74.37)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eOthers\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e191 (13.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e118 (13.59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e33 (12.55)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e40 (12.66)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eBMI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e27.658 (24.049-32.665)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e27.993 (24.42-33.207)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e26.352 (23.016-31.887)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e27.337 (23.903-31.029)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.015\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eSOFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e7 (4-9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e7 (4-9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e7 (4-10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e7 (4-10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.952\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eAPSIII\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e48 (37-63)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e47 (36-61)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e50 (37.5-65)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e48 (37-64)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.138\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eSAPS II\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e41 (32-50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e40 (31-50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e40 (32-50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e42 (33-51)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.441\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eOASIS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e36 (31-42)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e36 (31-41.25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e36 (29-41)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e36 (31-42)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.100\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eCCI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e5 (3-7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e5 (3-7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e6 (3-8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e6 (3-8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eHypertension (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e856 (59.16)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e508 (58.53)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e150 (57.03)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e198 (62.66)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.327\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e591 (40.84)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e360 (41.47)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e113 (42.97)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e118 (37.34)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eAKI(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e736 (50.86)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e464 (53.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e138 (52.47)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e134 (42.41)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e711 (49.14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e404 (46.54)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e125 (47.53)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e182 (57.59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003ePneumonia (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e795 (54.94)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e505 (58.18)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e152 (57.79)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e138 (43.67)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e652 (45.06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e363 (41.82)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e111 (42.21)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e178 (56.33)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eStroke (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e1,309 (90.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e779 (89.75)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e246 (93.54)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e284 (89.87)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.172\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e138 (9.54)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e89 (10.25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e17 (6.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e32 (10.13)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eCKD(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e1,147 (79.27)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e701 (80.76)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e208 (79.09)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e238 (75.32)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.123\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e300 (20.73)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e167 (19.24)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e55 (20.91)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e78 (24.68)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eCancer(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e1,254 (86.66)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e766 (88.25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e221 (84.03)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e267 (84.49)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.093\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e193 (13.34)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e102 (11.75)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e42 (15.97)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e49 (15.51)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eT2DM (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e1,017 (70.28)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e618 (71.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e172 (65.40)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e227 (71.84)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.156\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e430 (29.72)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e250 (28.80)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e91 (34.60)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e89 (28.16)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eHeart failure (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e978 (67.59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e595 (68.55)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e169 (64.26)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e214 (67.72)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.428\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e469 (32.41)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e273 (31.45)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e94 (35.74)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e102 (32.28)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eIschemic cardiomyopathy (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e934 (64.55)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e574 (66.13)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e156 (59.32)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e204 (64.56)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.129\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e513 (35.45)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e294 (33.87)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e107 (40.68)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e112 (35.44)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eCOPD (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e1,187 (82.03)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e713 (82.14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e213 (80.99)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e261 (82.59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.874\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e260 (17.97)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e155 (17.86)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e50 (19.01)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e55 (17.41)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eHemoglobin (g/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e10.4 (8.7-12.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e10.5 (8.7-12.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e10 (8.4-11.65)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e10.6 (9-12.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003ePlatelets (\u0026times;103/\u0026micro;L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e181 (127-249)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e186 (130-251)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e181 (121.5-262)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e174 (123-235)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.190\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eWBC (\u0026times;103/\u0026micro;L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e12.6 (8.6-17.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e12.8 (8.4-17.725)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e12.3 (8.8-16.65)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e12.1 (8.875-16.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.544\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eAnion gap\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e14 (12-17)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e14 (12-17)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e14 (12-18)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e15 (13-17)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.102\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eChloride(mEq/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e104 (100-108)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e105 (101-108)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e100 (94-104)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e107 (103-111)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eFBG (mg/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e134 (110-175)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e136 (111-177)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e130 (104-172.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e137 (110-175)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.856\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003ePotassium (mEq/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e4.1 (3.7-4.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e4.1 (3.8-4.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e4.2 (3.7-4.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e4 (3.6-4.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eBaseline sodium (mEq/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e139 (136-142)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e139 (136-141)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e134 (131-137)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e143 (140-146)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eSBP(mmHg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e119 (103-138)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e119 (102-138)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e119 (100.5-138.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e121 (106.75-141)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.137\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eHeart rate(beats/min)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e89 (77.5-105)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e89 (78-105)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e89 (77-104)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e89.5 (77.75-105)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.461\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eDBP(mmHg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e68 (57-81)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e67 (57-79)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e66 (54-81.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e71.5 (60-85)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.158\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eRespiration rate(breaths/min)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e19 (16-23)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e19 (16-24)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e18 (15-23)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e20 (16-24)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.167\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eTemperature(f)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e98.4 (97.8-99)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e98.4 (97.8-99.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e98.3 (97.7-98.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e98.5 (98-99.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.317\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eSpO2(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e99 (96-100)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e99 (96-100)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e99 (96-100)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e99 (96-100)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.885\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eInvasive ventilation (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e51 (3.52)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e27 (3.11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e9 (3.42)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e15 (4.75)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.400\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e1396 (96.48)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e841 (96.89)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e254 (96.58)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e301 (95.25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eVasoactive drugs(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e351 (24.26)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e201 (23.16)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e54 (20.53)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e96 (30.38)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e0.011\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e1,096 (75.74)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e667 (76.84)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e209 (79.47)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e220 (69.62)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003e30-day mortality(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e249 (17.21)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e123 (14.17)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e45 (17.11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e81 (25.63)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 119px;\"\u003e\n \u003cp\u003e365-day mortality(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 99px;\"\u003e\n \u003cp\u003e278 (19.21)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e140 (16.13)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 97px;\"\u003e\n \u003cp\u003e50 (19.01)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 101px;\"\u003e\n \u003cp\u003e88 (27.85)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eData are presented as median (interquartile range, IQR) or n (%). Abbreviations: BMI, body mass index; SOFA, Sequential Organ Failure Assessment; APS III, Acute Physiology Score III; SAPS II, Simplified Acute Physiology Score II; OASIS, Oxford Acute Severity of Illness Score; CCI, Charlson Comorbidity Index; AKI, acute kidney injury; CKD, chronic kidney disease; T2DM, type 2 diabetes mellitus; COPD, chronic obstructive pulmonary disease; WBC, white blood cell count; FBG, fasting blood glucose; SBP, systolic blood pressure; DBP, diastolic blood pressure; SpO2, peripheral oxygen saturation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003cstrong\u003e. Association of Serum Sodium Levels with 30-Day and 365-Day Mortality Among SAD Patients\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"590\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eOutcome\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 174px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 164px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;2\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 174px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eHR (95% CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eP\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eHR (95% CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eP\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eHR (95% CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eP\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"7\" style=\"width: 590px;\"\u003e\n \u003cp\u003e30-day mortality\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"7\" style=\"width: 590px;\"\u003e\n \u003cp\u003eBasic serum sodium level\u0026nbsp;tertiles\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eT 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eT 2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003e0.747 (0.557\u0026ndash;1.003)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e0.052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003e0.760 (0.566\u0026ndash;1.020)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e0.068\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003e0.998 (0.721\u0026ndash;1.383)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e0.992\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eT 3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003e0.809 (0.596\u0026ndash;1.099)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e0.176\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003e0.758 (0.557\u0026ndash;1.034)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e0.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003e0.998 (0.679\u0026ndash;1.465)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e0.99\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"7\" style=\"width: 590px;\"\u003e\n \u003cp\u003eSerum sodium Trajectories\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eClass 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eClass 2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003e1.208(0.858-1.699)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e0.279\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003e1.162 (0.825\u0026ndash;1.637)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e0.389\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003e0.831 (0.577\u0026ndash;1.197)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e0.319\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eClass 3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003e1.941(1.466-2.57)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003e1.735 (1.307\u0026ndash;2.305)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003e1.744 (1.296\u0026ndash;2.347)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"7\" style=\"width: 590px;\"\u003e\n \u003cp\u003e365-day mortality\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"7\" style=\"width: 590px;\"\u003e\n \u003cp\u003eBasic serum sodium level\u0026nbsp;tertiles\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eT 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eT 2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003e0.827 (0.628\u0026ndash;1.090)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e0.177\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003e0.842 (0.639\u0026ndash;1.110)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e0.223\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003e1.123 (0.827\u0026ndash;1.526)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e0.458\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eT 3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003e0.817 (0.609\u0026ndash;1.096)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e0.178\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003e0.771 (0.573\u0026ndash;1.038)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e0.086\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003e1.035 (0.715\u0026ndash;1.496)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e0.857\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"7\" style=\"width: 590px;\"\u003e\n \u003cp\u003eSerum sodium Trajectories\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eClass 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003eref\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eClass 2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003e1.183 (0.856\u0026ndash;1.634)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e0.308\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003e1.146 (0.829\u0026ndash;1.585)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e0.409\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003e0.823 (0.583\u0026ndash;1.162)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e0.269\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003eClass 3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 116px;\"\u003e\n \u003cp\u003e1.867 (1.430\u0026ndash;2.438)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003e1.684 (1.286\u0026ndash;2.205)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 121px;\"\u003e\n \u003cp\u003e1.694 (1.277\u0026ndash;2.247)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eT1: sodium\u0026lt;138.0, T2: 138.0\u0026le;sodium\u0026lt;142.0, T3: sodium\u0026ge;142.0. Model 1: Unadjusted crude model. Model 2: Adjusted for sex, age, race, BMI, and respiratory rate. Model 3: Model 2 plus disease severity scores (SOFA, APS III, SAPS II, OASIS, CCI), comorbidities (AKI, pneumonia, CKD, malignancy, heart failure, ischemic heart disease), vasoactive medications, and laboratory parameters (hemoglobin, anion gap, chloride).\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-informatics-and-decision-making","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"midm","sideBox":"Learn more about [BMC Medical Informatics and Decision Making](http://bmcmedinformdecismak.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/midm/default.aspx","title":"BMC Medical Informatics and Decision Making","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"sepsis-associated delirium, serum sodium trajectory, group-based trajectory modeling, all-cause mortality, critically ill patients","lastPublishedDoi":"10.21203/rs.3.rs-9036323/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9036323/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eObjective:\u003c/strong\u003e To investigate the association between dynamic serum sodium trajectories during the first 5 days of intensive care unit (ICU) admission and all-cause mortality in critically ill patients with sepsis-associated delirium (SAD), providing evidence-based insights for prognostic assessment and clinical management of this population.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003e This retrospective cohort study extracted clinical data from adult SAD patients admitted to ICU between 2008 and 2022 from the MIMIC-IV version 3.1 database. Group-based trajectory modeling (GBTM) was employed to classify serum sodium trajectories within 5 days of ICU admission. Kaplan-Meier survival curves were used to compare survival differences among trajectory groups, while Cox proportional hazards models were applied to analyze the associations between serum sodium trajectories and 30-day and 365-day all-cause mortality. Subgroup analysis and sensitivity analysis were conducted to validate the robustness of these associations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e A total of 1,447 adult SAD patients were finally enrolled in the study after applying exclusion criteria. GBTM identified three distinct serum sodium trajectories: normal-stable (n=868, 59.99%), persistent-hyponatremia (n=263, 18.18%), and hypernatremia-fluctuating (n=316, 21.84%). Kaplan-Meier analysis revealed significantly higher 30-day mortality (14.17% vs. 25.63% P\u0026lt;0.001) and 365-day mortality (16.13% vs. 27.85%, P\u0026lt;0.001) in the hypernatremia-fluctuating group compared with the normal-stable group. After adjusting for demographics, disease severity scores, comorbidities, and therapeutic interventions in multivariable Cox models, the hypernatremia-fluctuating trajectory remained an independent risk factor for 30-day (HR=1.744, 95% CI: 1.296-2.347, P\u0026lt;0.001) and 365-day mortality (HR=1.694, 95% CI: 1.277-2.247, P\u0026lt;0.001). Subgroup analyses indicated that the association between the Hypernatremic-Fluctuating trajectory and mortality was generally consistent across different subgroups defined by age, BMI, sex, and comorbidities..\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions: \u003c/strong\u003eThree distinct serum sodium trajectories were identified in critically ill SAD patients: normal-stable, persistent-hyponatremia, and hypernatremia-fluctuating. The hypernatremia-fluctuating trajectory represents an independent risk factor for both short-term and long-term all-cause mortality. Dynamic monitoring of serum sodium trajectories provides superior prognostic value compared with single-point measurements. Implementing serial serum sodium assessments during the first 5 days of ICU admission and identifying hypernatremia-fluctuating patterns may facilitate early risk stratification and individualized sodium homeostasis management, ultimately improving clinical outcomes in SAD patients.\u003c/p\u003e","manuscriptTitle":"Serum Sodium Trajectories and All-Cause Mortality in Critically Ill Patients with Sepsis-Associated Delirium: A Retrospective Analysis Based on the MIMIC-IV Database","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-19 08:09:30","doi":"10.21203/rs.3.rs-9036323/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-04-19T14:48:22+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-18T07:35:14+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"165601106130103767770407479578769517678","date":"2026-04-17T09:09:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"198562467776756222853014150829332593140","date":"2026-04-13T06:53:41+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"100602981805963110016278366726554496142","date":"2026-04-11T00:16:58+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-08T21:22:54+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-18T20:39:58+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-12T08:55:18+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-12T08:54:35+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Medical Informatics and Decision Making","date":"2026-03-05T05:50:11+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-informatics-and-decision-making","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"midm","sideBox":"Learn more about [BMC Medical Informatics and Decision Making](http://bmcmedinformdecismak.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/midm/default.aspx","title":"BMC Medical Informatics and Decision Making","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"37c6356e-2be5-4449-b35e-6ab2bf48dfac","owner":[],"postedDate":"April 19th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-19T08:09:30+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-19 08:09:30","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9036323","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9036323","identity":"rs-9036323","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}