Abstract
Background
Electrolyte disturbances are prevalent complications in critically ill patients with sepsis, significantly impacting patient prognosis. However, the specific association between serum potassium levels and mortality risk in this population remains poorly understood. This study aimed to investigate the association between serum potassium levels during hospitalization and the risk of 28-day and 90-day mortality in critically ill patients with sepsis.
Methods
Data were obtained from the Medical Information Mart for Intensive Care (MIMIC-IV) database, and patients with severe sepsis requiring ICU admission were stratified into quartiles based on their mean serum potassium levels. Outcomes assessed included 28-day and 90-day mortality. A multivariate Cox proportional hazards model was used to investigate the association between serum potassium levels and mortality, with restricted cubic splines to identify potential nonlinear correlations. A dichotomous Cox proportional hazards model was applied to analyze the association further, and Kaplan-Meier analysis assessed the mortality risk across different potassium ranges.
Results
A total of 25,203 patients were included, with 28-day and 90-day mortality rates of 27.84% and 40.48%, respectively. Multivariate analysis showed a significant association between serum potassium levels and mortality. Restricted cubic splines identified an inflection point at 4.4 mmol/L, with potassium levels above this threshold associated with higher mortality (28-day mortality: HR 2.96, 95% CI = 2.43–3.60; 90-day mortality: HR 2.19, 95% CI = 1.81–2.64). Kaplan-Meier analysis confirmed a significantly higher risk of death for patients with serum potassium levels above 4.4 mmol/L compared to those within the 3.5–4.4 mmol/L range (P<0.001).
Conclusion
In critically ill patients with sepsis, serum potassium levels exceeding 4.4 mmol/L are associated with an increased risk of death. Maintaining the average serum potassium level within the range of 3.5–4.4 mmol/L appears to be safe and may contribute to better outcomes in this patient population.
Introduction
Sepsis, characterized by life-threatening organ dysfunction resulting from a dysregulated host response to infection, stands as a primary cause of mortality among intensive care unit (ICU) patients, with approximately 20–30% succumbing to the condition within hospital walls [1, 2]. Despite extensive research efforts, including initiatives like the Sepsis Survival Campaign and the implementation of national core measures, sepsis mortality rates persist stubbornly high [3, 4]. Recognized as a syndrome rather than a distinct disease entity, sepsis involves complex abnormalities across multiple body systems and functions, encompassing cardiovascular, neuronal, metabolic, coagulation, and immune responses [5]. Consequently, predicting the clinical outcomes of ICU patients grappling with sepsis has emerged as a crucial research endeavor.
Electrolyte disturbances are prevalent among critically ill ICU patients and not only influence treatment strategies and ICU length of stay but also significantly impact patient prognosis. While researchers have extensively studied the effects of elevated sodium, increased phosphorus, and decreased calcium on clinical outcomes in patients with sepsis, serum potassium has received comparatively less attention [6–8]. Potassium, a vital electrolyte in the human body, plays a crucial role in maintaining the normal function of the cardiovascular system, skeletal muscle, visceral organs, and nervous system [9, 10]. Imbalances in serum potassium can lead to severe complications in critically ill patients, including arrhythmias and muscle weakness, with potential consequences such as cardiac arrest and respiratory failure [11]. In this study, we aimed to assess the predictive capacity of mean serum potassium levels for mortality in critically ill patients with sepsis and explore its association with sepsis-related mortality.
Methods
Study population
This study utilized health-related data obtained from the MIMIC-IV (version 2.2) repository, a comprehensive database managed by the MIT Computational Physiology Laboratory [12]. Researcher Guang Zhao adhered to all access protocols and conducted the data extraction. Patients diagnosed with sepsis according to the sepsis 3.0 diagnostic criteria were included [13]. Specific exclusion criteria were applied, including: (1) individuals below 18 years old at their initial admission; (2) patients with multiple sepsis admissions, with data extracted solely from the first admission; and (3) patients lacking adequate data, particularly serum potassium levels. A total of 25,203 patients were included in the analysis and stratified into four groups based on quartiles of serum potassium levels.
Ethics statement
This study adhered to the regulations outlined in the Helsinki Declaration. Approval for accessing the MIMIC-IV database was granted by the review committee of Beth Israel Deaconess Medical Center and Massachusetts Institute of Technology. As the data is openly accessible within the MIMIC-IV database, the need for ethical approval and informed consent was waived in this study.
Data collection
Various demographic and health-related factors were collected from the MIMIC-IV database and categorized into two primary domains: (1) Baseline characteristic variables, including age, race, sex, BMI, marital status, and comorbidities such as hypertension, diabetes, heart failure, renal failure, respiratory failure, and malignant tumor. (2) Clinical parameters, encompassing hemoglobin (Hb), platelet count (PLT), red blood cell count (RBC), hematocrit (HCT), white blood cell count (WBC), blood glucose, potassium (K+), sodium (Na+), serum calcium (Ca2+), albumin, lactate, and severity of illness scores such as Sequential Organ Failure Assessment (SOFA) score, Logistic Organ Dysfunction System (LODS) score, Simplified Acute Physiology Score II (SAPS-II), Oxford Acute Severity of Illness Score (OASIS), and Acute Physiology Score III (APSIII) [14–16]. Follow-up commenced on the day of admission and concluded on the day of death. With the exception of serum potassium, all laboratory variables and disease severity scores were extracted from data collected within 24 hours of patient admission.
Outcomes
The primary endpoint of the study was 28-day mortality, with the secondary endpoint being 90-day mortality.
Statistical analysis
Participants were stratified into four groups based on quartiles of mean potassium levels (Q1-Q4). Continuous variables were expressed as mean ± standard deviation (SD) or median (interquartile range, IQR) and compared using t-tests or Kruskal-Wallis rank sum tests, as appropriate. Categorical variables were presented as frequencies and percentages and compared using Fisher’s exact test when necessary. We assessed 28-day mortality and 90-day mortality for each potassium quartile during the follow-up period. A multivariate Cox proportional hazards regression model was employed to adjust for confounding factors and assess the independent predictive value of potassium. Model 1 remained unadjusted, while Model 2 included adjustments for age, sex, BMI, race, and marital status. Model 3 further incorporated adjustments for hypertension, diabetes, heart failure, renal failure, respiratory failure, and cancer based on Model 2. To explore the nonlinear relationship between potassium ion and mortality, we utilized restricted cubic spline Cox proportional hazards regression models and smooth curve fitting (penalized spline). The threshold was estimated by evaluating all possible values and selecting the most probable threshold point. We investigated the association between potassium and 28-day and 90-day mortality risk using two-piecewise Cox proportional hazards models on either side of the inflection point. Additionally, we conducted stratified analyses by sex, age (<70 years or ≥70 years), race, marital status, and coexisting conditions to assess the consistency of potassium’s prognostic value on outcomes. Furthermore, we employed Kaplan-Meier curves to analyze the risk of death in relation to serum potassium levels. Statistical significance was defined as two-sided p-values less than 0.05. All statistical analyses were performed using IBM SPSS version 24.0 and Empowerstats software version 6.0.
Results
Baseline characteristics
Table 1 displays the baseline characteristics of the cohort study participants (n = 25,203), categorized by serum potassium quartile. The mean age of participants was 67.3 years, with 57.79% being male. When compared to participants in the lowest quartile, those with higher potassium levels were more likely to be male, have diabetes, heart failure, and renal failure, and exhibited higher rates of in-hospital and 28-day mortality. Additionally, significant differences were observed in biochemical parameters between the groups. Participants in the highest quartile demonstrated significantly elevated platelet (PLT), blood glucose, and calcium (Ca2+) levels, as well as higher Sequential Organ Failure Assessment (SOFA) score, Simplified Acute Physiology Score II (SAPS II), and Acute Physiology Score III (APS III) scores compared to those in the first quartile. However, levels of albumin, lactate, and mean arterial pressure (MAP) were notably lower than those in the first quartile.
Table 1. Baseline characteristics according to the serum potassium quartiles.
| Quartiles of potassium | P value | |||||
|---|---|---|---|---|---|---|
| overall | Q1 | Q2 | Q3 | Q4 | ||
| (0.9–3.7) | (3.8–4) | (4.1–4.4) | (4.5–16.7) | |||
| N | 25203 | 6223 | 5014 | 7469 | 6497 | |
| Age, years | 67.30 | 65.30 | 66.70 | 68.10 | 68.70 | <0.001 |
| (56.10–78.20) | (53.30–77.20) | (55.23–78.50) | (57.30–78.40) | (58.40–78.80) | ||
| BMI, kg/m2 | 27.95 | 27.43 | 27.76 | 27.92 | 28.56 | <0.001 |
| (24.26–32.62) | (23.51–32.50) | (24.19–32.26) | (24.43–32.28) | (24.85–33.21) | ||
| Gender | <0.001 | |||||
| female | 10639 (42.21%) | 3187 (51.21%) | 2270 (45.27%) | 2887 (38.65%) | 2295 (35.32%) | |
| male | 14564 (57.79%) | 3036 (48.79%) | 2744 (54.73%) | 4582 (61.35%) | 4202 (64.68%) | |
| Race | <0.001 | |||||
| white | 16832 (66.79%) | 4059 (65.23%) | 3333 (66.47%) | 5087 (68.11%) | 4353 (67.00%) | |
| black | 2169 | 566 | 378 | 575 | 650 | |
| (8.61%) | (9.10%) | (7.54%) | (7.70%) | (10.00%) | ||
| others | 6202 (24.61%) | 1598 (25.68%) | 1303 (25.99%) | 1807 (24.19%) | 1494 (23.00%) | |
| Marital status | <0.001 | |||||
| single | 6561 (28.74%) | 1754 (31.35%) | 1400 (31.08%) | 1792 (26.32%) | 1615 (27.26%) | |
| married | 11422 (50.03%) | 2604 (46.54%) | 2146 (47.64%) | 3598 (52.85%) | 3074 (51.89%) | |
| divorced | 1785 | 480 | 358 | 513 | 434 | |
| (7.82%) | (8.58%) | (7.95%) | (7.54%) | (7.33%) | ||
| widowed | 3064 (13.42%) | 757 | 601 | 905 | 801 | |
| (13.53%) | (13.34%) | (13.29%) | (13.52%) | |||
| Hb (g/dL) | 11.7 | 11.7 | 11.7 | 11.8 | 11.6 | <0.001 |
| (10.0–13.3) | (10.0–13.2) | (10.0–13.3) | (10.2–13.4) | (10.0–13.2) | ||
| PLT (K/uL) | 211 (153–280) | 204 (145–274) | 210 (152–276) | 212 (156–277) | 216 (156–290) | <0.001 |
| RBC (m/uL) | 3.84 (3.27–4.38) | 3.83 (3.29–4.35) | 3.86 (3.28–4.40) | 3.87 (3.30–4.39) | 3.79 (3.24–4.36) | <0.001 |
| HCT (%) | 35.4 (30.7–39.8) | 34.9 (30.3–39.2) | 35.2 (30.5–39.7) | 35.9 (31.2–40.0) | 35.4 (30.6–40.0) | <0.001 |
| WBC (K/uL) | 9.6 (6.9–13.6) | 9.7 (6.8–13.8) | 9.8 (6.9–13.8) | 9.5 (7.0–13.3) | 9.6 (7.0–13.5) | 0.294 |
| Glucose (mg/dL) | 120 (100–154) | 119 (100–150) | 121 (101–155) | 119 (100–152) | 120 (99–159) | 0.121 |
| Na+ (mmol/L) | 139 (136–141) | 139 (136–141) | 139 (136–141) | 139 (136–141) | 138 (135–140) | <0.001 |
| Ca2+(mmol/L) | 1.11 (1.05–1.16) | 1.08 (1.02–1.14) | 1.10 (1.05–1.16) | 1.12 (1.07–1.17) | 1.13 (1.07–1.18) | <0.001 |
| Albumin (g/dL) | 3.4 (2.8–3.9) | 3.3 (2.8–3.8) | 3.3 (2.8–3.8) | 3.5 (2.9–3.9) | 3.4 (2.9–3.9) | <0.001 |
| Lactate (mmol/L) | 1.7 (1.2–2.5) | 1.8 (1.3–2.65) | 1.7 (1.2–2.5) | 1.6 (1.2–2.3) | 1.6 (1.2–2.5) | <0.001 |
| MAP (mmHg) | 77 (67–90) | 80 (69–92) | 79 (69–92) | 76 (67–87) | 75 (65–88) | <0.001 |
| SOFA | 3 (2–4) | 3 (2–4) | 3 (2–4) | 3 (2–4) | 3 (2–5) | <0.001 |
| LODS | 5 (3–7) | 5 (3–7) | 5 (3–7) | 5 (3–7) | 5 (3–8) | <0.001 |
| SAPSII | 37 (29–46) | 35 (27–44) | 36 (28–44) | 36 (29–46) | 40 (32–51) | <0.001 |
| OASIS | 33 (27–39) | 33 (27–39) | 33 (27–39) | 32 (27–38) | 33 (28–39) | <0.001 |
| APSIII | 45 (33–60) | 44 (33–59) | 44 (32–59) | 43 (32–58) | 47 (35–64) | <0.001 |
| Commorbidities | ||||||
| Hypertension | <0.001 | |||||
| no | 13881 (55.08%) | 3324 (53.41%) | 2745 (54.75%) | 4028 (53.93%) | 3784 (58.24%) | |
| yes | 11322 (44.92%) | 2899 (46.59%) | 2269 (45.25%) | 3441 (46.07%) | 2713 (41.76%) | |
| Diabetes | <0.001 | |||||
| no | 17609 (69.87%) | 4777 (76.76%) | 3636 (72.52%) | 5175 (69.29%) | 4021 (61.89%) | |
| yes | 7594 (30.13%) | 1446 (23.24%) | 1378 (27.48%) | 2294 (30.71%) | 2476 (38.11%) | |
| Heart Failure | <0.001 | |||||
| no | 19665 (78.03%) | 5158 (82.89%) | 3976 (79.30%) | 5657 (75.74%) | 4874 (75.02%) | |
| yes | 5538 (21.97%) | 1065 (17.11%) | 1038 (20.70%) | 1812 (24.26%) | 1623 (24.98%) | |
| Kidney Failure | <0.001 | |||||
| no | 12627 (50.10%) | 3332 (53.54%) | 2660 (53.05%) | 3907 (52.31%) | 2728 (41.99%) | |
| yes | 12576 (49.90%) | 2891 (46.46%) | 2354 (46.95%) | 3562 (47.69%) | 3769 (58.01%) | |
| Respiratory Failure | <0.001 | |||||
| no | 15533 (61.63%) | 3791 (60.92%) | 2995 (59.73%) | 4727 (63.29%) | 4020 (61.87%) | |
| yes | 9670 (38.37%) | 2432 (39.08%) | 2019 (40.27%) | 2742 (36.71%) | 2477 (38.13%) | |
| Malignant tumor | 0.089 | |||||
| no | 17520 (69.52%) | 4349 (69.89%) | 3537 (70.54%) | 5188 (69.46%) | 4446 (68.43%) | |
| yes | 7683 (30.48%) | 1874 (30.11%) | 1477 (29.46%) | 2281 (30.54%) | 2051 (31.57%) | |
| Events | ||||||
| Death in hospital | <0.001 | |||||
| no | 22717 (90.14%) | 5678 (91.24%) | 4547 (90.69%) | 6845 (91.65%) | 5647 (86.92%) | |
| yes | 2486 (9.86%) | 545 (8.76%) | 467 (9.31%) | 624 (8.35%) | 850 (13.08%) | |
| 28-day mortality | <0.001 | |||||
| no | 7777 (72.16%) | 1937 (72.68%) | 1523 (73.72%) | 2220 (75.38%) | 2097 (67.62%) | |
| yes | 3000 (27.84%) | 728 (27.32%) | 543 (26.28%) | 725 (24.62%) | 1004 (32.38%) | |
| 90-day mortality | 0.024 | |||||
| no | 6414 (59.52%) | 1598 (59.96%) | 1229 (59.49%) | 1805 (61.29%) | 1782 (57.47%) | |
| yes | 4363 (40.48%) | 1067 (40.04%) | 837 (40.51%) | 1140 (38.71%) | 1319 (42.53%) | |
BMI, body mass index; Hb, hemoglobin; PLT, platelet; RBC, red blood cell; HCT, hematocrit; WBC, white blood cell; SOFA, sequential organ failure assessment; LODS, logistic organ dysfunction system score; OASIS, oxford acute severity of illness score; SAPSII, simplified acute physiology score ii
Relationships of potassium with mortality
Multivariable Cox regression analysis was utilized to construct three models (Table 2). In Model 3, following adjustments for age, sex, BMI, race, marital status, hypertension, diabetes, heart failure, renal failure, respiratory failure, and malignancy, serum potassium levels were evaluated in quartiles relative to Q1 (reference). We observed hazard ratios (HRs) for potassium and 28-day mortality for Q2, Q3, and Q4 to be 0.73 (95% CI: 0.59, 0.91, P = 0.0044), 0.78 (95% CI: 0.64, 0.95, P = 0.0141) and 1.10 (95% CI:0.92,1.32, P = 0.3074), respectively. For 90-day mortality, the HRs were 0.70 (95% CI: 0.57, 0.85, P = 0.0003) and 0.72 (95% CI: 0.60, 0.86, P = 0.0003) for Q2 and Q3, respectively. Notably, there was a nonlinear relationship between serum potassium levels and both 28-day and 90-day mortality.
Table 2. Hazard ratios (95% CI) for mortality based on potassium quartiles.
| Varible | Crude | Mode I | Model II | |||
|---|---|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| 28-day mortality | ||||||
| Q1 | 1 (ref) | 1 (ref) | 1 (ref) | |||
| Q2 | 0.70 (0.62, 0.81) | <0.0001 | 0.72 (0.58, 0.88) | 0.002 | 0.73 (0.59, 0.91) | 0.0044 |
| Q3 | 0.68 (0.60, 0.77) | <0.0001 | 0.75 (0.62, 0.91) | 0.0035 | 0.78 (0.64, 0.95) | 0.0141 |
| Q4 | 1.06 (0.95, 1.19) | 0.2919 | 1.03 (0.87, 1.23) | 0.7264 | 1.10 (0.92, 1.32) | 0.3074 |
| 90-day mortality | ||||||
| Q1 | 1 (ref) | 1 (ref) | 1 (ref) | |||
| Q2 | 0.70 (0.62, 0.79) | <0.0001 | 0.69 (0.57, 0.84) | 0.0001 | 0.70 (0.57, 0.85) | 0.0003 |
| Q3 | 0.69 (0.61, 0.77) | <0.0001 | 0.70 (0.59, 0.84) | 0.0001 | 0.72 (0.60, 0.86) | 0.0003 |
| Q4 | 1.05 (0.95, 1.17) | 0.3511 | 1.04 (0.89, 1.23) | 0.6003 | 1.08 (0.91, 1.27) | 0.3908 |
Adjust model I adjust for: age, sex, BMI, race, and marital status
Adjust model II adjust for: age, sex, BMI, race, marital status, hypertension, diabetes, heart failure, renal failure, respiratory failure, and malignancy
The detection of nonlinear relationships
Cox proportional hazards regression models employing restricted cubic splines and smooth curve-fitting (penalized splines) were utilized to delve deeper into the association, building upon multivariate Cox regression analyses revealing a nonlinear relationship between serum potassium levels and both 28-day and 90-day mortality. Adjusted smoothed plots displayed a discernible upward trend, illustrating the association between serum potassium levels and both 28-day (Fig 1A) and 90-day mortality (Fig 1B). Furthermore, standard Cox proportional-hazards regression models and two-segment Cox proportional-hazards regression models were fitted to analyze the association between baseline potassium concentrations and mortality. Employing a two-segment Cox proportional hazards regression model, an inflection point of 4.4 was determined for both 28-day and in-hospital mortality (both P-values log-likelihood ratio <0.001) (Table 3). Upon adjustment for various covariates, the risk of both 28-day and 90-day mortality increased significantly by 196% per unit increase in potassium concentration (HR 2.96, 95% CI = 2.43–3.60, P < 0.01) and 119% (HR 2.19, 95% CI = 1.81–2.64, P < 0.0001), respectively.
Fig 1. Association of serum potassium levels with mortality.
Association of serum potassium levels with 28-day mortality (A) and 90-day mortality (B) in septic patients, with each hazard ratio calculated using a potassium level of 4.4 as a reference for both (A) and (B). The analysis is adjusted for age, sex, BMI, race, marital status, hypertension, diabetes, heart failure, renal failure, respiratory failure, and malignancy. Solid lines represent the hazard ratio estimates, while the red areas depict their corresponding 95% confidence intervals.
Table 3. Threshold effect analysis of potassium on 28-day mortality and 90-day mortality in patients with sepsis.
| Adjusted | |
|---|---|
| HR (95%CI) / P value | |
| 28-day mortality | |
| Fitting by two-piecewise Cox proportional risk model | |
| inflection point | 4.4 |
| K < 4.4 | 0.95 (0.77, 1.17) 0.6138 |
| K > = 4.4 | 2.96 (2.43, 3.60) <0.0001 |
| P for Log-likelihood ratio | <0.001 |
| in-hospital mortality | |
| Fitting by two-piecewise Cox proportional risk model | |
| inflection point | 4.4 |
| K < 4.4 | 0.91 (0.76, 1.08) 0.2796 |
| K > = 4.4 | 2.19 (1.81, 2.64) <0.0001 |
| P for Log-likelihood ratio | <0.001 |
Adjust model adjust for: age, sex, BMI, race, marital status, hypertension, diabetes, heart failure, renal failure, respiratory failure, and malignancy.
Stratified analyses
To further evaluate the relationship between serum potassium levels and the risk of 28-day mortality (Fig 2A) and 90-day mortality (Fig 2B) across various subgroups, we conducted stratified analyses based on age, sex, race, marital status, diabetes, hypertension, heart failure, renal failure, respiratory failure, and malignancy. Our findings revealed that both 28-day and 90-day mortality risks were notably higher in male patients compared to their female counterparts. Similarly, individuals with renal failure exhibited significantly elevated mortality risks compared to those without renal failure. Otherwise, serum potassium demonstrated consistent associations across most subpopulations.
Fig 2. Stratified analyses of the associations between potassium and mortality.
Stratified analysis of serum potassium levels and their association with 28-day mortality (A) and 90-day mortality (B). Serum potassium exhibited consistent associations across subgroups for both 28-day and 90-day mortality, except for the male and renal failure subgroups.
Impact of serum potassium within the physiological range on the mortality risk
In this study, we observed a gradual increase in the risk of death associated with serum potassium levels exceeding 4.4mmol/L. To explore potential differences in mortality risk within the normal range of serum potassium, we categorized potassium levels into two groups: 3.5–4.4mmol/L and 4.4–5.5mmol/L. Kaplan-Meier analyses were conducted to compare the risks of death at 28 days (Fig 3A) and 90 days (Fig 3B) between these groups. Our findings revealed that patients with higher serum potassium levels had a significantly elevated risk of death compared to those with lower serum potassium levels (P<0.001).
Fig 3. Kaplan-Meier curves for 28-day and 90-day cumulative mortality risk.
The cumulative risk of 28-day (A) and 90-day mortality (B) in critically ill patients with sepsis stratified by serum potassium levels of 3.5–4.4mmol/L and 4.4–5.5mmol/L.
Discussion
This study marks the inaugural investigation into the association between mean serum potassium levels and mortality among patients with severe sepsis. Through threshold effect analysis, we identified that mortality risk in septic patients exhibited a curvilinear increase when mean serum potassium exceeded 4.4mmol/L. This underscores the robust predictive value of serum potassium for both 28-day and 90-day mortality outcomes. Additionally, maintaining serum potassium levels within the range of 3.5–4.4mmol/L appears to be a safe and prudent approach.
During ICU stays, critically ill patients frequently experience electrolyte disturbances, significantly correlating with elevated mortality rates [17, 18]. Electrolytes play pivotal roles in various physiological functions, with precise intracellular and extracellular concentrations crucial for normal metabolic processes and organ function [19, 20]. The electrolyte profiles of critically ill patients manifest considerable heterogeneity due to disease status and specific interventions like fluid resuscitation and drug therapy, complicating electrolyte disorder correction and treatment strategies [21–23]. Potassium ions notably play a critical role, with deviations in blood potassium levels disrupting normal physiological function, impacting patient prognosis, and potentially precipitating organ failure in severe cases. Hypokalemia can induce acidosis, renal dysfunction, muscle cramps, and arrhythmias, while hyperkalemia can precipitate muscle paralysis, cardiac arrest, or life-threatening arrhythmias [24, 25]. Studies consistently demonstrate that both hypokalemia and hyperkalemia heighten mortality risk, with these electrolyte abnormalities closely associated with elevated risks of fatal complications [26]. The observed association between serum potassium levels and mortality risk in critically ill patients with sepsis underscores the importance of monitoring and managing electrolyte imbalances in this population.
Sepsis, a prevalent critical condition in the ICU, requires close attention to its clinical manifestations, particularly hypotension and shock, which signify more severe stages of the illness. The elevation of blood potassium levels correlates with diminished blood pressure. Intracellular potassium ion efflux precipitates persistent vasodilation, further exacerbating hypotension [27]. Additionally, potassium efflux induces continuous hyperpolarization of cell membranes, resulting in impaired vascular reactivity in septic patients [28]. This elucidates why the administration of vasopressors does not significantly ameliorate microcirculatory disorders [29]. Potassium ion imbalance also escalates the risk of arrhythmias and potentially heart failure, thereby amplifying mortality risks in critically ill septic patients [30]. Research has substantiated the involvement of potassium ions in macrophage transmembrane signal transduction, eliciting cytokine production, and fostering inflammatory responses [31]. Consequently, vigilant monitoring of potassium level fluctuations not only mitigates complications in septic patients but also curtails inflammation progression, thereby improving patient prognoses.
In our investigation, we observed a considerable elevation in mortality risk associated with potassium levels exceeding 4.4mmol per liter. Consequently, we sought to explore whether this heightened mortality risk persisted among septic patients with serum potassium levels within the physiological normal range (3.5 to 5.5mmol/L). Utilizing Kaplan-Meier analysis, we discerned a noteworthy reduction in mortality risk when patients maintained lower normal potassium levels. This finding aligns with a substantial retrospective observational study, which identified potassium levels exceeding 4.5 mmol/L as a robust independent predictor of 30-day all-cause mortality, with mortality risk escalating alongside potassium levels [32, 33]. Moreover, Hessels et al. [26]suggested that maintaining serum potassium concentrations between 3.5 and 4.0 mmol/L was safe for critically ill patients. Building upon these insights, we advocate for maintaining potassium levels within the safe range of 3.5–4.4mmol/L in critically ill patients with sepsis.
This study possesses several limitations. Firstly, as an observational study conducted at a single center, it cannot establish causality. Despite employing multivariate adjustment to control for confounding variables, the potential for residual confounding impacting prognosis remains. Moreover, the absence of personal patient histories, such as smoking and alcohol consumption, represents another limitation.
Conclusion
In conclusion, our findings suggest that mean serum potassium levels are a valuable predictor of 28-day and 90-day mortality risk in critically ill patients with sepsis, with a safe target range identified between 3.5 and 4.4 mmol/L. Therefore, assessing mean serum potassium levels during hospitalization may provide insight into risk assessment and prognosis within this patient population. Future research endeavors should aim to investigate the effectiveness of interventions aimed at modulating serum potassium levels to improve clinical outcomes in these individuals.
Acknowledgments
The authors extend their gratitude to the participants and staff of the MIMIC-IV dataset for their invaluable contributions to this study.
Data Availability
The data analyzed in this study were sourced from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database, accessible via https://mimic.mit.edu/.
Funding Statement
The author(s) received no specific funding for this work.
References
- 1.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(8):801–10. Epub 2016/02/24. doi: 10.1001/jama.2016.0287 ; PubMed Central PMCID: PMC4968574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med. 2001;29(7):1303–10. Epub 2001/07/11. doi: 10.1097/00003246-200107000-00002 . [DOI] [PubMed] [Google Scholar]
- 3.Coopersmith CM, De Backer D, Deutschman CS, Ferrer R, Lat I, Machado FR, et al. Surviving sepsis campaign: research priorities for sepsis and septic shock. Intensive Care Med. 2018;44(9):1400–26. Epub 2018/07/05. doi: 10.1007/s00134-018-5175-z ; PubMed Central PMCID: PMC7095388. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Shirazy M, Omar I, Abduljabbar D, Bousselmi K, Alkhaja M, Chaari A, et al. Prevalence and Prognostic Impact of Hypernatremia in Sepsis and Septic Shock Patients in the Intensive Care Unit: A Single Centre Experience. J Crit Care Med (Targu Mures). 2020;6(1):52–8. Epub 2020/02/28. doi: 10.2478/jccm-2020-0001 ; PubMed Central PMCID: PMC7029404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Cohen M, Banerjee D. Biomarkers in Sepsis: A Current Review of New Technologies. J Intensive Care Med. 2024;39(5):399–405. Epub 2023/08/18. doi: 10.1177/08850666231194535 . [DOI] [PubMed] [Google Scholar]
- 6.Lin CY, Chen YM, Tsai YH, Hung KY, Fang YT, Chang YP, et al. Association of Hypernatremia with Immune Profiles and Clinical Outcomes in Adult Intensive Care Unit Patients with Sepsis. Biomedicines. 2022;10(9). Epub 2022/09/24. doi: 10.3390/biomedicines10092285 ; PubMed Central PMCID: PMC9496274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wang H, Zhang L, Liao W, Huang J, Xu J, Yang J, et al. Hyperphosphatemia rather than hypophosphatemia indicates a poor prognosis in patients with sepsis. Clin Biochem. 2021;91:9–15. Epub 2021/02/19. doi: 10.1016/j.clinbiochem.2021.01.016 . [DOI] [PubMed] [Google Scholar]
- 8.Sood A, Singh G, Singh TG, Gupta K. Pathological role of the calcium-sensing receptor in sepsis-induced hypotensive shock: Therapeutic possibilities and unanswered questions. Drug Dev Res. 2022;83(6):1241–5. Epub 2022/06/12. doi: 10.1002/ddr.21959 . [DOI] [PubMed] [Google Scholar]
- 9.Ravn Jacobsen M, Jabbari R, Glinge C, Kjær Stampe N, Butt JH, Blanche P, et al. Potassium Disturbances and Risk of Ventricular Fibrillation Among Patients With ST-Segment-Elevation Myocardial Infarction. J Am Heart Assoc. 2020;9(4):e014160. Epub 2020/02/19. doi: 10.1161/JAHA.119.014160 ; PubMed Central PMCID: PMC7070188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fan Y, Wu M, Li X, Zhao J, Shi J, Ding L, et al. Potassium levels and the risk of all-cause and cardiovascular mortality among patients with cardiovascular diseases: a meta-analysis of cohort studies. Nutr J. 2024;23(1):8. Epub 2024/01/10. doi: 10.1186/s12937-023-00888-z ; PubMed Central PMCID: PMC10777575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Tongyoo S, Viarasilpa T, Permpikul C. Serum potassium levels and outcomes in critically ill patients in the medical intensive care unit. J Int Med Res. 2018;46(3):1254–62. Epub 2018/01/16. doi: 10.1177/0300060517744427 ; PubMed Central PMCID: PMC5972260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.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):1. Epub 2023/01/04. doi: 10.1038/s41597-022-01899-x ; PubMed Central PMCID: PMC9810617. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tong R, Ding X, Liu F, Li H, Liu H, Song H, et al. Classification of subtypes and identification of dysregulated genes in sepsis. Front Cell Infect Microbiol. 2023;13:1226159. Epub 2023/09/06. doi: 10.3389/fcimb.2023.1226159 ; PubMed Central PMCID: PMC10475835. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Schertz AR, Lenoir KM, Bertoni AG, Levine BJ, Mongraw-Chaffin M, Thomas KW. Sepsis Prediction Model for Determining Sepsis vs SIRS, qSOFA, and SOFA. JAMA Netw Open. 2023;6(8):e2329729. Epub 2023/08/25. doi: 10.1001/jamanetworkopen.2023.29729 ; PubMed Central PMCID: PMC10457723. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wang L, Zhang Z, Hu T. Effectiveness of LODS, OASIS, and SAPS II to predict in-hospital mortality for intensive care patients with ST elevation myocardial infarction. Sci Rep. 2021;11(1):23887. Epub 2021/12/15. doi: 10.1038/s41598-021-03397-3 ; PubMed Central PMCID: PMC8668882. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Zheng X, Hu T, Liu T, Wang W. Simplified acute physiology score III is excellent for predicting in-hospital mortality in coronary care unit patients with acute myocardial infarction: A retrospective study. Front Cardiovasc Med. 2022;9:989561. Epub 2022/12/27. doi: 10.3389/fcvm.2022.989561 ; PubMed Central PMCID: PMC9775274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Laupland KB, Tabah A, Jacobs N, Ramanan M. Determinants of serum magnesium abnormalities and outcome among admissions to the intensive care unit. Anaesth Crit Care Pain Med. 2020;39(6):793–7. Epub 2020/10/16. doi: 10.1016/j.accpm.2020.07.020 . [DOI] [PubMed] [Google Scholar]
- 18.Tucker AM, Johnson TN. Acid-base disorders: A primer for clinicians. Nutr Clin Pract. 2022;37(5):980–9. Epub 2022/06/27. doi: 10.1002/ncp.10881 . [DOI] [PubMed] [Google Scholar]
- 19.Dervishi A. A deep learning backcasting approach to the electrolyte, metabolite, and acid-base parameters that predict risk in ICU patients. PLoS One. 2020;15(12):e0242878. Epub 2020/12/18. doi: 10.1371/journal.pone.0242878 ; PubMed Central PMCID: PMC7746262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Sim JK, Ko RE, Na SJ, Suh GY, Jeon K. Intensive care unit-acquired hyponatremia in critically ill medical patients. J Transl Med. 2020;18(1):268. Epub 2020/07/04. doi: 10.1186/s12967-020-02443-4 ; PubMed Central PMCID: PMC7333267. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Jaber S, Paugam C, Futier E, Lefrant JY, Lasocki S, Lescot T, et al. Sodium bicarbonate therapy for patients with severe metabolic acidaemia in the intensive care unit (BICAR-ICU): a multicentre, open-label, randomised controlled, phase 3 trial. Lancet. 2018;392(10141):31–40. Epub 2018/06/19. doi: 10.1016/S0140-6736(18)31080-8 . [DOI] [PubMed] [Google Scholar]
- 22.Formiga F, Chivite D, Corbella X, Conde-Martel A, Arévalo-Lorido JC, Trullàs JC, et al. Influence of potassium levels on one-year outcomes in elderly patients with acute heart failure. Eur J Intern Med. 2019;60:24–30. Epub 2019/02/07. doi: 10.1016/j.ejim.2018.10.016 . [DOI] [PubMed] [Google Scholar]
- 23.Pfortmueller CA, Dabrowski W, Wise R, van Regenmortel N, Malbrain M. Fluid accumulation syndrome in sepsis and septic shock: pathophysiology, relevance and treatment-a comprehensive review. Ann Intensive Care. 2024;14(1):115. Epub 2024/07/21. doi: 10.1186/s13613-024-01336-9 ; PubMed Central PMCID: PMC11264678. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kim MJ, Valerio C, Knobloch GK. Potassium Disorders: Hypokalemia and Hyperkalemia. Am Fam Physician. 2023;107(1):59–70. Epub 2023/01/24. . [PubMed] [Google Scholar]
- 25.McLean RM, Wang NX. Potassium. Adv Food Nutr Res. 2021;96:89–121. Epub 2021/06/12. doi: 10.1016/bs.afnr.2021.02.013 . [DOI] [PubMed] [Google Scholar]
- 26.Hessels L, Hoekstra M, Mijzen LJ, Vogelzang M, Dieperink W, Lansink AO, et al. The relationship between serum potassium, potassium variability and in-hospital mortality in critically ill patients and a before-after analysis on the impact of computer-assisted potassium control. Crit Care. 2015;19(1):4. Epub 2015/01/07. doi: 10.1186/s13054-014-0720-9 ; PubMed Central PMCID: PMC4312438. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hosoda H. Effect of Ghrelin on the Cardiovascular System. Biology (Basel). 2022;11(8). Epub 2022/08/27. doi: 10.3390/biology11081190 ; PubMed Central PMCID: PMC9405061. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Jo C, Kim B, Lee K, Choi HY. Vascular Relaxation and Blood Pressure Lowering Effects of Prunus mume in Rats. Bioengineering (Basel). 2023;10(1). Epub 2023/01/22. doi: 10.3390/bioengineering10010074 ; PubMed Central PMCID: PMC9854816. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Chan YL, Orie NN, Dyson A, Taylor V, Stidwill RP, Clapp LH, et al. Inhibition of vascular adenosine triphosphate-sensitive potassium channels by sympathetic tone during sepsis. Crit Care Med. 2012;40(4):1261–8. Epub 2012/03/20. doi: 10.1097/CCM.0b013e31823da98d . [DOI] [PubMed] [Google Scholar]
- 30.Huang BP, Zhao L, Zhao XM, Zhai M, Huang Y, Zhou Q, et al. Serum Potassium Levels and Mortality in Hospitalized Heart Failure Patients. Rev Cardiovasc Med. 2023;24(8):228. Epub 2024/07/30. doi: 10.31083/j.rcm2408228 ; PubMed Central PMCID: PMC11266832. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Sordi R, Fernandes D, Heckert BT, Assreuy J. Early potassium channel blockade improves sepsis-induced organ damage and cardiovascular dysfunction. Br J Pharmacol. 2011;163(6):1289–301. Epub 2011/03/18. doi: 10.1111/j.1476-5381.2011.01324.x ; PubMed Central PMCID: PMC3144541. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Wetmore JB, Yan H, Peng Y, Gilbertson DT, Herzog CA. Development and outcomes of hyperkalemia in hospitalized patients: potential implications for care. Am Heart J. 2021;241:59–67. Epub 2021/07/23. doi: 10.1016/j.ahj.2021.07.006 . [DOI] [PubMed] [Google Scholar]
- 33.McMahon GM, Mendu ML, Gibbons FK, Christopher KB. Association between hyperkalemia at critical care initiation and mortality. Intensive Care Med. 2012;38(11):1834–42. Epub 2012/07/19. doi: 10.1007/s00134-012-2636-7 . [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data analyzed in this study were sourced from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database, accessible via https://mimic.mit.edu/.