Abstract
Background
This study aimed to investigate the association between postoperative nadir serum potassium and 28-day mortality in patients undergoing surgical aortic valve replacement (SAVR) using the Medical Information Mart for Intensive Care (MIMIC-IV) database.
Methods
A retrospective cohort study included 2,877 patients undergoing SAVR from MIMIC-IV. The independent variable was postoperative nadir serum potassium, and the primary outcome was 28-day all-cause mortality. Logistic regression models adjusted for sex, age, BMI, use of beta-blockers, antidiabetic drugs, diuretics, comorbidities (COPD, heart failure, peripheral vascular disease), and indexes of kidney (BUN, Creatinine). Restricted cubic spline (RCS) regression assessed the potential nonlinear relationship, and Kaplan–Meier survival curves compared mortality risk across different serum potassium levels.
Results
A U-shaped relationship was observed between postoperative nadir potassium and 28-day all-cause mortality. Higher nadir potassium levels were significantly associated with reduced mortality (odds ratio [OR] = 0.40, 95% confidence interval [CI]: 0.24–0.69, p < 0.001). The optimal threshold was 3.55 mmol/L; patients below this level had a markedly increased risk of death.
Conclusion
Lower postoperative nadir serum potassium is independently associated with increased 28-day mortality in SAVR patients. Maintaining postoperative nadir serum potassium no lower than 3.5 mmol/L may be associated with reduced short-term mortality, although this threshold should not be interpreted as a target level to be constantly maintained. The nadir value represents only the lowest single measurement during ICU stay and does not reflect overall potassium homeostasis.
Supplementary Information
The online version contains supplementary material available at 10.1186/s13019-026-04239-x.
Keywords: Nadir serum potassium, Aortic valve replacement, Hypokalemia, MIMIC-Ⅳ
Background
Aortic stenosis (AS) is one of the most common heart valve diseases, characterized by progressive calcification of the aortic valve [1]. Aortic regurgitation (AR) is also a common valve disease, characterized by the backflow of the blood from the aorta to left ventricle during the diastolic period. Due to the influence of aortic valve disease on the left ventricle, patients often have impaired cardiac function and are more sensitive to homeostatic disturbance, especially in patients with AR. Most patients with aortic valve disease eventually require aortic valve replacement [2].
Potassium, the predominant intracellular cation, plays a vital role in maintaining cellular membrane potential and normal physiological function [3–8]. Even minor disruption in potassium homeostasis can lead to significant clinical consequences, including arrhythmias, muscle weakness, and increased mortality [9, 10].
However, the association of postoperative nadir serum potassium levels with short-term mortality in patients undergoing SAVR remains uncertain. This study aims to clarify the relationship of postoperative nadir serum potassium with short-term mortality.
Method
Data source
This retrospective cohort study utilized the data from the MIMIC-IV 3.1 database, which contains de-identified clinical information from over 70,000 ICU patients admitted to Beth Israel Deaconess Medical Center between 2008 and 2019. One of the researchers (Hao-ran Zhang) obtained a license for the database. Data access was granted through a certified data use agreement (ID: 13910140). Due to the de-identified nature of the data, institutional review board approval was not required. The written informed consent form was also waived.
Study population
Patients undergoing SAVR were identified using structured query language (PostgreSQL v15.0). Only the first ICU admission was included for the patients with multiple admission of ICU. Exclusion criteria were: age < 18 years and missing postoperative potassium data.
Data extraction
The database has an independent identifier, allowing investigators to identify individual patient information. All date-related content in the database has been de-identified to protect the confidentiality of patient.
Demographic information and laboratory data of the patients who admitted to the ICU for the first time due to SAVR were extracted, including: (1) Demographic information: age, sex, race, and BMI; (2) Comorbidities: hypertension, diabetes, coronary heart disease, chronic kidney disease, cerebral infarction, anemia, atrial fibrillation, myocardial infarction, heart failure, chronic obstructive pulmonary disease (COPD), and cancer; (3) Preoperative medication: Angiotensin Converting Enzyme Inhibitor(ACE-I), beta receptor blocker, Angiotensin Ⅱ Receptor Blocker, calcium antagonist, statin, antidiabetic drug, inotropic agent, and diuretic; (4) Scoring systems: Glasgow Coma Scale (GCS), Sequential Organ Failure Assessment (SOFA), Acute Physiology and Chronic Health Evaluation III (APACHE III), Logistic Organ Dysfunction System (LODS), and Oxford Acute Severity of Illness Score (OASIS); (5) Laboratory results: hemoglobin, hematocrit, platelet count, red blood cells (RBC), white blood cells (WBC), sodium, chloride, creatinine, and blood urea nitrogen; (6) Clinical outcomes: acute kidney injury (AKI), hospital length of stay, ICU length of stay, 28-day mortality, and 90-day mortality.
Missing data handling
Variables with > 20% missing values were excluded. The remaining variables with missing data were addressed using multiple imputation method.
Definitions
The postoperative nadir serum potassium was defined as the lowest serum potassium value recorded from all measurements taken during the patient’s ICU stay following aortic valve replacement. Based on the different postoperative nadir serum potassium values, patients were classified into the three groups, with serum potassium ranges of < 3.5, 3.5–4.0, and > 4.0 mmol/L [11]. Comorbidities were extracted using ICD-9/10 codes. Severity of illness was assessed using GCS, SOFA, APACHE III, LODS, and OASIS scores. AKI was defined according to Kidney Disease: Improving Global Outcome (KDIGO) criteria. In this study, the 28-day all-cause mortality was defined as the primary outcome. The secondary outcomes included 90-day all-cause mortality, the incidence of AKI, hospital length of stay, and ICU length of stay.
Statistical analysis
Continuous variables were first tested for normality using the Shapiro-Wilk test. Normally distributed data are presented as mean ± standard deviation (SD), and comparisons among groups were performed using one-way analysis of variance (ANOVA). Nonnormally distributed data are presented as median with interquartile range (IQR), and group differences were assessed using the Kruskal-Wallis test. Categorical variables are expressed as numbers and percentages, and group comparisons were performed using the chi-square test. Kaplan–Meier analysis was performed to compare survival curves among different groups of postoperative nadir serum potassium. Restricted Cubic Spline (RCS) analysis was used to explore the relationship of postoperative nadir serum potassium with the outcomes. The optimal postoperative nadir potassium threshold was determined using ROC Area under curve (AUC) and the Youden index. Sensitivity analyses were conducted to assess consistency across patient strata. Multiple logistic regression models were constructed for the adjustment of different confounders. Variables were selected for multivariable models based on clinical expertise and prior literature, combined with univariate analysis where variables with P < 0.05 were considered for inclusion. For logistic regression analyses, the reference level for serum potassium was set at 3.5 mmol/L based on the conventional clinical definition of hypokalemia. Logistic regression models with the same sequential adjustment strategy were used to examine the association between postoperative nadir serum potassium and arrhythmia.
Results
Baseline characteristics
Of 3,015 patients undergoing AVR, 2877 met the inclusion criteria (Fig. 1). The median of BMI was 29.3 (IQR: 25.9–33.5), and 78.5% were White. The 28-day and 90-day mortality rate was 1.6% and 3.2%, respectively (Table 1). Patients with postoperative nadir serum potassium > 4.0 mmol/L were more likely to be male, younger, and have fewer comorbidities. Patients with potassium > 4.0 mmol/L also had lower rate of AKI and blood transfusion, as well as shorter hospital and ICU length of stay. Patients with a nadir serum potassium > 4.5 mmol/L were extremely rare, accounting for less than 5% of the cohort, which precluded independent statistical analysis.
Fig. 1.
Flow chart illustrating the inclusion and exclusion of participants
Table 1.
Comparison of baseline characteristics among groups with different serum potassium levels
| Variable | Overall (n = 2 877) | < 3.5 mmol/L (n = 910) | 3.5–4.0 mmol/L (n = 1 539) | > 4.0 mmol/L (n = 428) | P value |
|---|---|---|---|---|---|
| Demographics | |||||
| Age, yrs | 70.0 (62.0, 78.0) | 72.0 (63.0, 79.0) | 70.0 (62.0, 78.0) | 68.5 (60.5, 77.0) | 0.001 |
| Male sex, n (%) | 1 878 (65.3%) | 497 (54.6%) | 1 028 (66.8%) | 353 (82.5%) | < 0.001 |
| White race, n (%) | 2 259 (78.5%) | 703 (77.3%) | 1 222 (79.4%) | 334 (78.0%) | 0.436 |
| BMI | 29.3 (25.9, 33.5) | 29.1 (25.7, 34.1) | 29.3 (26.0, 33.0) | 30.0 (26.4, 33.7) | 0.106 |
| Comorbidities, n (%) | |||||
| Cancer | 573 (19.9%) | 167 (18.4%) | 312 (20.3%) | 94 (22.0%) | 0.267 |
| Hypertension | 1 720 (59.8%) | 524 (57.6%) | 941 (61.1%) | 255 (59.6%) | 0.220 |
| Coronary heart disease | 653 (22.7%) | 202 (22.2%) | 333 (21.6%) | 118 (27.6%) | 0.032 |
| Chronic renal failure | 111 (3.9%) | 48 (5.3%) | 45 (2.9%) | 18 (4.2%) | 0.013 |
| Cerebral infarction | 58 (2.0%) | 25 (2.7%) | 26 (1.7%) | 7 (1.6%) | 0.165 |
| Anemia | 1 371 (47.7%) | 502 (55.2%) | 669 (43.5%) | 200 (46.7%) | < 0.001 |
| Diabetes | 429 (14.9%) | 156 (17.1%) | 198 (12.9%) | 75 (17.5%) | 0.004 |
| Atrial fibrillation | 205 (7.1%) | 69 (7.6%) | 111 (7.2%) | 25 (5.8%) | 0.504 |
| Myocardial infarction | 313 (10.9%) | 111 (12.2%) | 150 (9.7%) | 52 (12.1%) | 0.112 |
| Heart failure | 913 (31.7%) | 388 (42.6%) | 407 (26.4%) | 118 (27.6%) | < 0.001 |
| COPD | 113 (3.9%) | 36 (4.0%) | 55 (3.6%) | 22 (5.1%) | 0.336 |
| Preoperative medications, | |||||
| ACE inhibitor | 163 (5.7%) | 61 (6.7%) | 74 (4.8%) | 28 (6.5%) | 0.102 |
| β-blocker | 340 (11.8%) | 137 (15.1%) | 159 (10.3%) | 44 (10.3%) | 0.001 |
| ARB | 37 (1.3%) | 19 (2.1%) | 12 (0.8%) | 6 (1.4%) | 0.021 |
| Calcium antagonist | 78 (2.7%) | 31 (3.4%) | 34 (2.2%) | 13 (3.0%) | 0.191 |
| Statin | 152 (5.3%) | 61 (6.7%) | 70 (4.5%) | 21 (4.9%) | 0.066 |
| Antidiabetic drug | 167 (5.8%) | 72 (7.9%) | 78 (5.1%) | 17 (4.0%) | 0.003 |
| Cardiotonic agent | 79 (2.7%) | 45 (4.9%) | 26 (1.7%) | 8 (1.9%) | < 0.001 |
| Diuretic | 357 (12.4%) | 167 (18.4%) | 145 (9.4%) | 45 (10.5%) | < 0.001 |
| Scoring systems | |||||
| GCS | 15.0 (15.0, 15.0) | 15.0 (15.0, 15.0) | 15.0 (15.0, 15.0) | 15.0 (15.0, 15.0) | 0.039 |
| SOFA (24 h) | 3.0 (2.0, 4.0) | 3.0 (2.0, 4.0) | 3.0 (2.0, 4.0) | 4.0 (3.0, 5.0) | < 0.001 |
| APACHE III | 34.0 (27.0, 46.0) | 38.5 (30.0, 52.0) | 33.0 (26.0, 44.0) | 34.0 (27.0, 43.0) | < 0.001 |
| LODS | 4.0 (3.0, 6.0) | 5.5 (4.0, 8.0) | 4.0 (3.0, 6.0) | 4.0 (3.0, 6.0) | < 0.001 |
| OASIS | 26.0 (22.0, 31.0) | 27.0 (22.0, 32.0) | 26.0 (22.0, 30.0) | 26.0 (21.0, 30.0) | < 0.001 |
| Laboratory events | |||||
| Hemoglobin, g/dL | 9.3 (8.2, 10.6) | 9.0 (7.9, 10.5) | 9.4 (8.2, 10.7) | 9.4 (8.2, 10.6) | < 0.001 |
| Hematocrit, % | 27.9 (24.5, 31.8) | 27.5 (23.8, 31.5) | 27.9 (24.7, 31.8) | 28.4 (25.3, 31.8) | 0.010 |
| Platelet count, 10³/µL | 141 (111, 183) | 142 (107, 191) | 140 (114, 177) | 142 (114, 187) | 0.545 |
| RBC, 10⁶/µL | 3.1 (2.7, 3.5) | 3.0 (2.6, 3.5) | 3.1 (2.7, 3.5) | 3.1 (2.8, 3.5) | 0.049 |
| WBC, 10³/µL | 11.2 (8.3, 15.0) | 11.2 (8.1, 15.2) | 11.2 (8.3, 14.8) | 11.7 (8.6, 15.2) | 0.401 |
| Sodium, mmol/L | 135 (133, 136) | 135 (132, 136) | 135 (133, 136) | 135 (133, 136) | < 0.001 |
| Chloride, mmol/L | 107 (104, 109) | 105 (101, 109) | 107 (104, 109) | 107 (105, 109) | < 0.001 |
| Creatinine, mg/dL | 0.8 (0.7, 1.0) | 0.8 (0.6, 1.0) | 0.8 (0.7, 1.0) | 0.9 (0.8, 1.1) | < 0.001 |
| BUN, mg/dL | 15.0 (12.0, 18.0) | 14.0 (11.0, 18.0) | 14.0 (11.0, 18.0) | 17.0 (13.0, 21.0) | < 0.001 |
| Clinical outcomes, n (%) | |||||
| Hospital stay los | 7.0(5.0, 11.0) | 10(6.0, 17.0) | 6(5.0, 9.0) | 6.0(5.0, 9.0) | < 0.001 |
| ICU stay los | 2.1(1.3, 3.4) | 3.4(2.0, 6.5) | 1.7(1.2, 3.1) | 1.3(1.2, 2.2) | < 0.001 |
| AKI | 798 (27.7%) | 326 (35.8%) | 355 (23.1%) | 117 (27.3%) | < 0.001 |
| 28-day mortality | 47 (1.6%) | 25 (2.7%) | 18 (1.2%) | 4 (0.9%) | 0.006 |
| 90-day mortality | 91 (3.2%) | 55 (6.0%) | 26 (1.7%) | 9 (2.1%) | < 0.001 |
Continuous variables were compared using ANOVA (normally distributed) or Kruskal-Wallis test (non-normally distributed). Categorical variables were compared using the chi-square test. P values indicate overall differences among the three groups
Multiple model analysis
To reduce bias, we established five models to investigate the association of postoperative nadir serum potassium with the primary and secondary outcomes. In model 1, univariate logistic regression was applied; in model 2, covariates were adjusted according to age, sex, and BMI; in model 3, adjustments were made for age, sex, BMI, beta-blockers, diuretics, and antidiabetic drugs; in model 4, adjustments were made for age, sex, BMI, beta-blockers, diuretics, antidiabetic drugs, chronic obstructive pulmonary disease, heart failure, and peripheral vascular disease; in model 5, adjustments were made for age, sex, BMI, beta-blockers, diuretics, antidiabetic drugs, chronic obstructive pulmonary disease, heart failure, peripheral vascular disease, BUN, and Creatinine (Table 2). We found that four multivariate logistic regression models, the odds ratios (OR) for 28-day mortality were 0.40 (0.24, 0.69), 0.39 (0.23, 0.66), 0.43 (0.25, 0.75), 0.47 (0.27, 0.82), and 0.42 (0.24, 0.75) respectively.
Table 2.
Association between nadir potassium level and clinical outcomes (28-day mortality, ICU length of stay, hospital length of stay, and acute kidney injury) across multivariable adjusted models
| Variables | Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| Beta/OR(95%CI) | P-value | Beta/OR(95%CI) | P-value | Beta/OR(95%CI) | P-value | Beta/OR(95%CI) | P-value | Beta/OR(95%CI) | P-value | |
| 28 day Mortality | 0.40(0.24, 0.69) | < 0.001 | 0.39(0.23, 0.66) | < 0.001 | 0.43(0.25, 0.75) | 0.003 | 0.47(0.27, 0.82) | 0.007 | 0.42(0.24, 0.75) | 0.003 |
| ICU LOS | -4.76(-5.16, -4.37) | < 0.001 | -4.85(-5.25, -4.45) | < 0.001 | -4.73(-5.13, -4.33) | < 0.001 | -4.62(-5.02, -4.22) | < 0.001 | -4.70(-5.10, -4.30) | < 0.001 |
| Hospital LOS | -7.04(-7.73, -6.35) | < 0.001 | -7.27(-7.96, -6.57) | < 0.001 | -6.28(-6.91, -5.65) | < 0.001 | -6.06(-6.69, -5.44) | < 0.001 | -6.20(-6.83, -5.57) | < 0.001 |
| AKI | 0.59(0.49, 0.71) | < 0.001 | 0.61(0.50, 0.73) | < 0.001 | 0.64(0.53, 0.77) | < 0.001 | 0.66(0.55, 0.80) | < 0.001 | 0.54(0.44, 0.66) | < 0.001 |
Kaplan–meier analysis
Kaplan-Meier analysis demonstrated that higher postoperative nadir serum potassium levels were significantly associated with lower 28-day and 90-day mortality (p = 0.006). Among the three groups of patients stratified by the postoperative nadir serum potassium level, the patients with the serum potassium level > 4 mmol/L had the lowest mortality rate, and the advantage in survival rate became more pronounced with the prolongation of follow-up time (Fig. 2A). During the initial two weeks, the difference of survival rates among the three group was not significant, with the group of postoperative nadir serum potassium level between 3.5 and 4 mmol/L demonstrating the highest survival rate. However, subsequent analysis revealed that the group with serum potassium levels exceeding 4 mmol/L experienced a significantly reduced rate of survival decline and ultimately maintained the highest survival rate. The mortality rate of the group with postoperative serum potassium level below 3.5 mmol/L was markedly higher compared to the other two groups till 90 days (Fig. 2B).
Fig. 2.
K-M curve of 28-day and 90-day survival in SAVR patients. (A) Kaplan-Meier survival curve for 28-day all-cause mortality (P = 0.006). (B) Kaplan-Meier survival curve for 90-day all-cause mortality (P < 0.001)
RCS and ROC analysis
RCS showed a non-linear relationship between postoperative nadir serum potassium and mortality (Fig. 3). The optimal threshold was identified as 3.55 mmol/L. However, given the difficulty of achieving precise decimal-level control in clinical practice, a practical interpretation is that postoperative nadir serum potassium should not fall below 3.5 mmol/L. Each 1 mmol/L increase of postoperative nadir serum potassium was associated with a 60% reduction in 28-day mortality (OR = 0.40; 95% CI: 0.24–0.69). Before covariate adjustment, the area under the curve (AUC) of the predicted values of the 28-day and 90-day mortality rates of SAVR patients by the postoperative lowest serum potassium level was 0.636 and 0.697 respectively, and the optimal thresholds were both 3.55 mmol/L. After adjustment for age, gender, BMI, heart failure, COPD, beta-blockers, antidiabetic drugs, diuretics, peripheral vascular diseases, BUN, and Creatinine, the AUC of the predicted values of the 28-day and 90-day mortality rates of SAVR patients by the postoperative lowest serum potassium level was 0.770 and 0.789 respectively, and the optimal thresholds were 2.70 and 3.03 mmol/L, respectively (Fig. 4).
Fig. 3.
Nonlinear association between nadir potassium and 28-day mortality
Fig. 4.
ROC curves of nadir postoperative potassium for predicting 28-day and 90-day mortality in SAVR patients: Comparison of unadjusted and adjusted models
Secondary outcomes
In the unadjusted model and the models adjusted by sequentially adding demographic characteristics (sex, age, BMI), preoperative medications (beta-blockers, antidiabetic drugs, diuretics), comorbidities (heart failure, COPD, peripheral vascular diseases), and indexes of kidney (BUN, Creatinine), the odds ratios (OR) of AKI were 0.59 (0.49, 0.71), 0.61 (0.50, 0.73), 0.64 (0.53, 0.77), 0.66 (0.55, 0.80), and 0.54 (0.44, 0.66) respectively. We also observed a significant negative association between postoperative serum potassium level and the risk of AKI. However, with the gradual inclusion of demographic characteristics, preoperative medications and comorbidities, the OR value slowly increased from 0.59 to 0.66. The standardized regression coefficients (β) of the length of hospital stay were − 7.04 (− 7.73, − 6.35), − 7.27 (− 7.96, − 6.57), − 6.28 (− 6.91, − 5.65), − 6.06 (− 6.69, − 5.44), and − 6.20 (-6.83, -5.57) respectively; the β values of the length of ICU stay were − 4.76 (− 5.16, − 4.37), − 4.85 (− 5.25, − 4.45), − 4.73 (− 5.13, − 4.33), − 4.62 (− 5.02, − 4.22), and − 4.70 (-5.10, -4.30) respectively. In Model 2 adjusted for demographics, the absolute values of β for the length of hospital stay and ICU stay increased, while in Models 3 and 4, the absolute values of β gradually decreased. We also examined the association between postoperative nadir serum potassium and arrhythmia using the same sequential adjustment strategy. In the unadjusted model (OR = 0.995, 95% CI: 0.357–2.770, P = 0.992) and the fully adjusted model (OR = 1.229, 95% CI: 0.293–5.154, P = 0.778), no statistically significant association was observed. Detailed results are presented in Supplementary Table S1.
Subgroup analysis
Figure 5 presents the forest plot of subgroup analyses. The association between postoperative nadir serum potassium and 28-day mortality was consistent across most subgroups, including sex (female: 0.23[0.19–0.28]; male: 0.48[0.39–0.58]), age (< 65 years: 0.73[0.60–0.89]; ≥65 years: 0.29[0.24–0.35]), BMI (< 30: 0.23[0.19–0.28]; ≥30: 0.97[0.80–1.18]), heart failure (no: 0.27[0.22–0.33]; yes: 0.69[0.57–0.84]), COPD (no: 0.40[0.33–0.49]; yes: 0.35[0.29–0.43]), beta- blocker (no: 0.44[0.36–0.54]; yes: 0.37[0.30–0.45]), antidiabetic drug (no: 0.35[0.29–0.43]; yes: 1.34[1.10–1.63]), diuretic (no: 0.39[0.32–0.47]; yes: 0.59[0.48–0.72]), and peripheral vascular disease (no: 0.42[0.35–0.51]; yes: 0.36[0.30–0.44]). While sub-group analysis also suggested potentially stronger effects in female patients (OR = 0.23), older adults aged ≥ 65 years (OR = 0.29), non-obesity individuals (BMI < 30, OR = 0.23), and those without heart failure (OR = 0.27). However, interaction tests did not reach statistical significance in these subgroups (all p > 0.05). A borderline interaction was observed for heart failure status (p = 0.080). Notably, among patients receiving antidiabetic drugs, higher nadir potassium was unexpectedly associated with increased 28-day mortality (OR = 1.34), which contrasts with the protective effect observed in other subgroups. This paradoxical finding may be explained by residual confounding, such as the severity of diabetes, insulin use versus oral agents, or concurrent renal impairment. It is also possible that antidiabetic drugs affecting potassium homeostasis (e.g., SGLT2 inhibitors) were differentially distributed. Given the small sample size in this subgroup (n = 167, 5.8%), chance findings cannot be excluded, and this result should be interpreted with caution.
Fig. 5.
Forest plot of odds ratios for the association between nadir potassium and 28-day mortality across subgroups, with interaction P-values
Discussion
This study found a U-shaped association between postoperative nadir serum potassium and 28-day mortality in patients undergoing SAVR. Patients with postoperative nadir serum potassium level > 4.0mmol/L had the lowest 28-day and 90-day mortality rates, shorter hospital and ICU length of stay, and a lower risk of postoperative AKI. Additionally, as the postoperative nadir serum potassium level increased, the OR value of 28-day mortality showed a significant downward trend. The association of higher postoperative nadir serum potassium with short term mortality was consistent across different patient subgroups.
In the present study, the reference level for serum potassium in logistic regression models was set at 3.5 mmol/L. This cutoff was chosen based on the conventional clinical definition of hypokalemia (serum potassium < 3.5 mmol/L), which is widely accepted in clinical practice and has been used as a reference in numerous prior studies investigating potassium-related outcomes [11, 12, 21]. Using this clinically meaningful reference level facilitates the interpretation of odds ratios and enhances the comparability of our findings with existing literature.
These findings are consistent with previous studies that demonstrated the risks of both hypokalemia and hyperkalemia. Moreover, this is the first study to identify a specific threshold of postoperative nadir serum potassium in patients undergoing SAVR. To date, the “optimal” range of potassium remains unclear. Although 3.5-5.5mmol/L is currently considered safe, it is still difficult to define a critical value associated with the lowest risk of death and other clinical outcomes in patients undergoing SAVR [12–14]. Hypokalemia leads to an increased concentration gradient of potassium ions across the myocardial cell membrane, reduced potassium efflux, and a decrease in the resting potential negativity of myocardial cells, thereby affecting cell excitability [15, 16]. Since cardiac repolarization depends on potassium influx, hypokalemia prolongs the action potential and increases QT dispersion (reflecting electrical heterogeneity). Hypokalemia is also a strong independent predictor of mortality in heart failure (HF), and patients taking non-potassium-sparing diuretics have higher all-cause and cardiac mortality rates in patients with HF [17, 18]. In a study of 6,073 patients with HF, the risk of death was relatively increased by 2.3 times in patients with a serum potassium level < 3.5mmol/L [19]. In a cohort of 142,087 patients from the US Veterans Affairs database, the risk of death was 1.3 times and 1.6 times higher in acute decompensated heart failure (ADHF) patients with intermittent and persistent hypokalemia, respectively [20]. A study on coronary artery bypass grafting (CABG) showed that the 90-day mortality risk was increased in patients with postoperative hyperkalemia, hypokalemia, and normal-range hypokalemia (3.5-3.9mmol/L), while the relative risk was smaller in patients with normal range of potassium (4.7-5.0mmol/L), and the all-cause mortality rate increased in patients with potassium levels outside the range of 4.2-5.2mmol/L [21]. Another study showed that patients with a serum potassium level below 3.5 mmol/L (HR 1.36, 95% CI 1.12–1.66) and those with a serum potassium level above 4.6 mmol/L (HR 2.46, 95% CI 1.66–2.74) had an increased mortality rate within 60 days in patients with hypertension, compared with those with a serum potassium level of 3.5–4.6 mmol/L. Moreover, serum potassium level below 3.0 mmol/L was significantly associated with mortality (HR 2.46, 95% CI 1.05–5.74), and patients with serum potassium level between 3.0 and 3.4 mmol/L also had an increased risk of death [22]. A study of 2,660,609 patients from the US National Inpatient Sample database demonstrated a 1.9-fold higher in-hospital mortality risk in hypokalemic patients [23], and multiple prior studies have linked hypokalemia to increased mortality in cardiovascular disease patients [24–28]. Consistent with the U-shaped association between postoperative nadir serum potassium and 28-day mortality observed in our study, unadjusted mortality in unselected cardiac intensive care unit (CICU) patients was also U-shaped related to admission serum potassium, including subgroups with ACS, HF, AKI and CKD [29]. This U-shaped relationship has been further confirmed in other studies: a UK cohort of 21,334 newly diagnosed HF patients showed a 2-fold higher mortality risk in any hypokalemia (< 3.5 mmol/L) after mortality correction [30]; a Danish National Registry study of 19,549 chronic HF patients found that serum potassium < 3.8 mmol/L was associated with a 3-fold higher 90-day mortality risk compared with 4.2–4.4 mmol/L and a 2.4-fold higher risk compared with 3.5-5.0 mmol/L [31]; similar correlations were also observed between post-randomization potassium levels and cardiovascular mortality in HFpEF patients [32], and between serum potassium and all-cause mortality as summarized in a JACC review [33]. Although the U-shaped potassium-mortality association exists in both SAVR and HF patients, potassium exerts a distinct perioperative-specific effect on death risk in SAVR patients. Postoperatively, SAVR patients have stunned myocardium due to surgical trauma and ischemia-reperfusion injury, with extremely fragile electrophysiological and systolic functions, resulting in far greater sensitivity to hypokalemia than HF patients. In the specific context of SAVR, the U-shaped association between postoperative nadir potassium and mortality carries distinct clinical implications. First, surgical aortic valve replacement involves cardiopulmonary bypass, aortic cross-clamping, and cardioplegic arrest, all of which predispose the myocardium to reperfusion injury and electrical instability. In this setting, even transient hypokalemia may act as a proximate trigger for malignant arrhythmias, whereas hyperkalemia is often iatrogenic and less common. Second, the nadir potassium typically occurs early postoperatively due to hemodilution, diuresis, and stress-related catecholamine surge, making it a modifiable and time-sensitive target. Third, the U-shaped relationship underscores that both ends of the potassium spectrum are harmful, but the mechanisms differ: hypokalemia primarily increases arrhythmic risk, while hyperkalemia may reflect renal dysfunction or excessive supplementation. Therefore, potassium management after SAVR should aim to avoid extreme deviations rather than achieving a single “optimal” value. Mild hypokalemia can directly induce malignant arrhythmias, exacerbate myocardial injury, and synergize with abrupt postoperative volume load changes to increase acute cardiogenic adverse event risk, acting as a key trigger for perioperative acute death—whereas hypokalemia-related risks in HF patients are mostly cumulative effects of chronic pathological processes, and mild hypokalemia generally does not directly cause acute death. In addition, potassium disorders in SAVR patients are mostly short-term stress-induced postoperative abnormalities due to cardiopulmonary bypass and volume fluctuations, rather than the chronic compensatory abnormalities from long-term RAAS activation and chronic medication use in HF patients. Rigorous perioperative monitoring and rapid intervention also significantly attenuate the lethal effect of hyperkalemia in SAVR patients, unlike in HF patients where hyperkalemia is also a major mortality risk factor. Furthermore, prompt perioperative correction of hypokalemia in SAVR patients can directly block the acute pathological cascade with marked immediate benefits and a narrow intervention window, thus requiring a more stringent and aggressive potassium management strategy than that for HF patients to effectively avoid hypokalemia-related short-term mortality. However, an ICU study showed that patients with an average serum potassium of 3.5-4.0 mmol/L and low potassium variability had the lowest mortality rate (3.7%), and even those with mild hypokalemia (3.0-3.5 mmol/L) had significantly lower in-hospital mortality than patients with 4.0-4.5 mmol/L [34]. For SAVR patients, cardiac function and electrolyte balance are critical for postoperative recovery, and hypokalemia exerts a more pronounced impact on cardiac function, potentially triggering complications such as arrhythmias and increasing mortality. While overactivated RAAS and diuretic use can cause hypokalemia, and ACEI/ARB and mineralocorticoid receptor antagonist (MRA) use can lead to hyperkalemia in both populations, the underlying physiological mechanisms differ. HF patients often develop hyperkalemia due to reduced potassium excretion from insufficient renal perfusion, whereas aortic regurgitation patients undergoing SAVR have left ventricular dilation and dysfunction caused by diastolic blood regurgitation, which alters myocardial cell physiological characteristics and renders this population more sensitive to potassium fluctuations. Additionally, previous studies analyzing the correlation between postoperative serum potassium and 90-day mortality in CABG patients focused on the first serum potassium level within 7 days postoperatively, enabling simultaneous evaluation of hyperkalemia and hypokalemia on mortality [21]. In contrast, the present study analyzes postoperative nadir serum potassium in SAVR patients, which provides greater guiding significance for maintaining the minimum postoperative serum potassium level and preventing postoperative hypokalemia.
There are several limitations in this study. First, this is a retrospective, single-center study, which precludes causal inference. Second, the predominantly U.S.-based cohort may limit external generalizability. Third, unmeasured confounders, such as intraoperative factors or dietary intake, could also influence the outcomes. It is also important to note that the nadir serum potassium represents a single extreme value rather than the average or cumulative potassium status during the postoperative period. Therefore, our findings should not be interpreted as advocating for maintaining potassium at or above 3.55 mmol/L at all times, but rather as identifying a threshold below which the risk of mortality increases significantly. Fourth, patients with lower nadir potassium levels had higher APACHE III scores and were older, indicating greater baseline illness severity. Therefore, nadir potassium may serve as a prognostic marker rather than a direct causal factor, and residual confounding cannot be fully excluded. Fifth, among the samples excluded due to the lack of blood potassium measurement levels in the ICU, the majority also had missing data on the length of ICU stay. This part of the samples might have had data missing because of a relatively short ICU stay, which could lead to the exclusion of samples of early postoperative deaths, causing a certain degree of bias. Additionally, the unexpected interaction between antidiabetic drug use and nadir potassium warrants further investigation in larger, prospective cohorts with detailed medication classification and diabetes severity adjustment.
Conclusion
In patients undergoing SAVR, the postoperative nadir serum potassium level is independently associated with 28-day mortality. A threshold of 3.55 mmol/L was identified, which can be practically interpreted as maintaining postoperative nadir serum potassium no lower than 3.5 mmol/L to reduce short-term mortality. However, this threshold should not be interpreted as a target level to be maintained constantly, nor does it imply that higher potassium levels are beneficial for prognosis. The nadir value represents only the lowest single measurement during ICU stay and does not reflect overall potassium homeostasis throughout the hospitalization period.
Electronic Supplementary Material
Acknowledgements
Not applicable.
Abbreviations
- SBP
Systolic Blood Pressure
- DBP
Diastolic Blood Pressure
- SpO₂
Peripheral Oxygen Saturation
- ICU
Intensive Care Unit
- AKI
Acute Kidney Injury
- Hb
Hemoglobin
- Hct
Hematocrit
- PLT
Platelet count
- RBC
Red Blood Cell count
- WBC
White Blood Cell count
- BUN
Blood Urea Nitrogen
- APACHE III
Acute Physiology and Chronic Health Evaluation III
- LODS
Logistic Organ Dysfunction Score
- OASIS
Oxford Acute Severity of Illness Score
- COPD
Chronic Obstructive Pulmonary Disease
- ACE inhibitor
Angiotensin-Converting Enzyme inhibitor
- ARB
Angiotensin II Receptor Blocker
- GCS
Glasgow Coma Scale
- SOFA
Sequential Organ Failure Assessment
- BMI
Body Mass Index
- HF
Heart Failure
- CABG
Coronary Artery Bypass Grafting
Author contributions
HZ and SZ analyzed and interpreted the patient data regarding the hematological disease and the transplant, performed the histological examination of the kidney, and were major contributors in writing the manuscript. SX and XL analyzed and interpreted the patient data regarding the hematological disease and the transplant, and were major contributors in writing the manuscript. XG and ML performed the histological examination of the kidney, and were major contributors in writing the manuscript. All authors read and approved the final manuscript.
Funding
There was no funding for this study.
Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
Due to the de-identified nature of the data, institutional review board approval was not required. The written informed consent form was also waived.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Ambrosy AP, Go AS, Leong TK, Garcia EA, Chang AJ, Slade JJ, et al. Temporal trends in the prevalence and severity of aortic stenosis within a contemporary and diverse community-based cohort. Int J Cardiol. 2023;384:107–11. [DOI] [PubMed] [Google Scholar]
- 2.Carabello BA, Paulus WJ. Aortic stenosis. Lancet. 2009;373(9667):956–66. [DOI] [PubMed] [Google Scholar]
- 3.Chen Y, Sang Y, Ballew SH, Tin A, Chang AR, Matsushita K, et al. Race, serum potassium, and associations with ESRD and mortality. Am J Kidney Dis. 2017;70:244–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Dhondup T, Qian Q. Electrolyte and acid-base disorders in chronic kidney disease and end-stage kidney failure. Blood Purif. 2017;43(1–3):179–88. [DOI] [PubMed] [Google Scholar]
- 5.Luo J, Brunelli SM, Jensen DE, Yang A. Association between serum potassium and outcomes in patients with reduced kidney function. Clin J Am Soc Nephrol. 2016;11(1):90–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Xia Z, Huang X, Chen K, Wang H, Xiao J, He K, et al. Proapoptotic role of potassium ions in liver cells. Biomed Res Int. 2016;2016:1729135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chen Y, Chang AR, McAdams DeMarco MA, Inker LA, Matsushita K, Ballew SH et al. Serum potassium, mortality, and kidney outcomes in the Atherosclerosis Risk in Communities Study. Mayo Clin Proc. 2016;91(10):1403-12. [DOI] [PMC free article] [PubMed]
- 8.Eliacik E, Yildirim T, Sahin U, Kizilarslanoglu C, Tapan U, Aybal-Kutlugun A, et al. Potassium abnormalities in current clinical practice: frequency, causes, severity and management. Med Princ Pract. 2015;24(3):271–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Tromp J, Ter Maaten JM, Damman K, O’Connor CM, Metra M, Dittrich HC, et al. Serum potassium levels and outcome in acute heart failure (data from the PROTECT and COACH trials). Am J Cardiol. 2017;119:290–6. [DOI] [PubMed] [Google Scholar]
- 10.Engelhardt LJ, Balzer F, Müller MC, Grunow JJ, Spies CD, Christopher KB, et al. Association between potassium concentrations, variability and supplementation, and in-hospital mortality in ICU patients: a retrospective analysis. Ann Intensive Care. 2019;9:100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Alper AB, Campbell RC, Anker SD, Bakris G, Wahle C, Love TE, et al. A propensity-matched study of low serum potassium and mortality in older adults with chronic heart failure. Int J Cardiol. 2008;137:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Macdonald JE, Struthers AD. What is the optimal serum potassium level in cardiovascular patients? J Am Coll Cardiol. 2004;43:155–61. [DOI] [PubMed] [Google Scholar]
- 13.Goyal A, Spertus JA, Gosch K, Venkitachalam L, Jones PG, Van den Berghe G, et al. Serum potassium levels and mortality in acute myocardial infarction. JAMA. 2012;307:157–64. [DOI] [PubMed] [Google Scholar]
- 14.Korgaonkar S, Tilea A, Gillespie BW, Kiser M, Eisele G, Finkelstein F, et al. Serum potassium and outcomes in CKD: insights from the RRI-CKD cohort study. Clin J Am Soc Nephrol. 2010;5:762–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Fisch C, Knoebel SB, Feigenbaum H, Greenspan K. Potassium and the monophasic action potential, electrocardiogram, conduction and arrhythmias. Prog Cardiovasc Dis. 1966;8:387–418. [DOI] [PubMed] [Google Scholar]
- 16.Gettes L, Surawicz B. Effects of low and high concentrations of potassium on the simultaneously recorded Purkinje and ventricular action potentials of the perfused pig moderator band. Circ Res. 1968;23:717–29. [DOI] [PubMed] [Google Scholar]
- 17.Cleland JG, Dargie HJ, Ford I. Mortality in heart failure: clinical variables of prognostic value. Br Heart J. 1987;58:572–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Cooper HA, Dries DL, Davis CE, Shen YL, Domanski MJ. Diuretics and risk of arrhythmic death in patients with left ventricular dysfunction. Circulation. 1999;100:1311–5. [DOI] [PubMed] [Google Scholar]
- 19.Hoss S, Elizur Y, Luria D, Keren A, Lotan C, Gotsman I. Serum potassium levels and outcome in patients with chronic heart failure. Am J Cardiol. 2016;118:1868–74. [DOI] [PubMed] [Google Scholar]
- 20.Matsushita K, Sang Y, Yang C, Ballew SH, Grams ME, Coresh J, et al. Dyskalemia, its patterns, and prognosis among patients with incident heart failure: a nationwide study of US veterans. PLoS ONE. 2019;14:e0219899. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kjeldgaard M, Mæng MO, Torp-Pedersen C, Søgaard P, Kragholm KH, Andreasen JJ, et al. 90-day mortality risk related to postoperative potassium levels in patients undergoing coronary bypass surgery. J Mol Cell Cardiol Plus. 2023;4:100035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Krogager ML, Søgaard P, Torp-Pedersen C, Bøggild H, Lee CJ, Bonde A, et al. Impact of plasma potassium normalization on short-term mortality in patients with hypertension and hypokalemia or low normal potassium. BMC Cardiovasc Disord. 2020;20(1):386. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Basnet S, Dhital R, Tharu B, Ghimire S, Poudel DR, Donato A. Influence of abnormal potassium levels on mortality among hospitalized heart failure patients in the US: data from National Inpatient Sample. J Community Hosp Intern Med Perspect. 2019;9:103–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Alvis BD, Hughes CG. Physiology considerations in geriatric patients. Anesthesiol Clin. 2015;33(3):447–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Islas-Granillo H, Medina-Solís CE, de Lourdes Márquez-Corona M, de la Rosa-Santillana R, Fernández-Barrera MÁ, Villalobos-Rodelo JJ, et al. Prevalence of multimorbidity in subjects aged ≥ 60 years in a developing country. Clin Interv Aging. 2018;13:1129–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kojima T, Mizokami F, Akishita M. Geriatric management of older patients with multimorbidity. Geriatr Gerontol Int. 2020;20(12):1105–11. [DOI] [PubMed] [Google Scholar]
- 27.Abensur Vuillaume L, Ferreira JP, Asseray N, Trombert-Paviot B, Montassier E, Legrand M, et al. Hypokalemia is frequent and has prognostic implications in stable patients attending the emergency department. PLoS ONE. 2020;15(8):e0236934. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Steen B. Hypokalemia–clinical spectrum and etiology. Acta Med Scand Suppl. 1981;647:61–6. [DOI] [PubMed] [Google Scholar]
- 29.Brueske B, Sidhu MS, Schulman-Marcus J, Kashani KB, Barsness GW, Jentzer JC. Hyperkalemia is associated with increased mortality among unselected cardiac intensive care unit patients. J Am Heart Assoc. 2019;8(7):e011814. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Linde C, Qin L, Bakhai A, Furuland H, Evans M, Ayoubkhani D, et al. Serum potassium and clinical outcomes in heart failure patients: results of risk calculations in 21 334 patients in the UK. ESC Heart Fail. 2019;6(2):280–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Aldahl M, Jensen AC, Davidsen L, Eriksen MA, Møller Hansen S, Nielsen BJ, et al. Associations of serum potassium levels with mortality in chronic heart failure patients. Eur Heart J. 2017;38(38):2890–6. [DOI] [PubMed] [Google Scholar]
- 32.Ferreira JP, Butler J, Rossignol P, Pitt B, Anker SD, Kosiborod M, et al. Abnormalities of potassium in heart failure: JACC state-of-the-art review. J Am Coll Cardiol. 2020;75(22):2836–50. [DOI] [PubMed] [Google Scholar]
- 33.Desai AS, Liu J, Pfeffer MA, Claggett B, Fleg J, Lewis EF, et al. Incident hyperkalemia, hypokalemia, and clinical outcomes during spironolactone treatment of heart failure with preserved ejection fraction: analysis of the TOPCAT trial. J Card Fail. 2018;24:313–20. [DOI] [PubMed] [Google Scholar]
- 34.Engelhardt LJ, Balzer F, Müller MC, Grunow JJ, Spies CD, Christopher KB, et al. Association between potassium concentrations, variability and supplementation, and in-hospital mortality in ICU patients: a retrospective analysis. Ann Intensive Care. 2019;9(1):100. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.