Abstract
Our study aimed at assessing the impact of β-blocker treatment on the mortality of patients with sepsis-associated acute kidney injury (SA-AKI). Clinical data for patients with SA-AKI were collected from the Medical Information Mart for Intensive Care (MIMIC)-IV database. Baseline characteristics between β-blocker users and non-users were adjusted through propensity score matching (PSM). Kaplan-Meier curves and the Cox proportional hazards model were employed to analyze the relationship between β-blocker use and mortality, with the primary outcome being 30-day mortality. This study included data from 8,620 patients, with 3,159 receiving treatment and 5,461 not. After PSM, 2,896 matched pairs were created. The analysis revealed that 30-day (hazard ratio [HR]: 0.58, 95% CI: 0.52–0.65; p < 0.001), 90-day (HR: 0.55, 95% CI: 0.50–0.61; p < 0.001), and 365-day mortality (HR: 0.59, 95% CI: 0.54–0.64; p < 0.001) were significantly reduced in β-blocker users. Treatment with long-acting β-blockers was linked to notable improvement in 30-day (17.0 vs. 29.2%; p < 0.001), 90-day (24.1 vs. 36.0%; p < 0.001), and 365-day survival (31.9 vs. 45.0%; p < 0.001). In contrast, short-acting β-blockers showed no significant effect on 30-day (33.8 vs. 28.3%; p = 0.08), 90-day (37.5 vs. 34.2%; p = 0.07), or 365-day mortality (46.3 vs. 38.8%; p = 0.14). β-blocker therapy was associated with reduced mortality rates at 30, 90, and 365 days in SA-AKI patients. Long-acting β-blockers have demonstrated significant protective effects, while short-acting β-blockers, like esmolol, failed to improve patient survival in SA-AKI.
Keywords: β-blockers, sepsis-associated acute kidney injury, mortality, MIMIC-IV
Introduction
Sepsis-associated acute kidney injury (SA-AKI) is a prevalent and severe complication among ICU patients. Recent studies suggest a rising incidence of SA-AKI, with up to 50% of sepsis patients developing acute kidney injury (AKI) [1]. SA-AKI is associated prolonged ICU stay and significantly increased mortality [2], making it a critical research topic in intensive care medicine. The pathophysiological mechanisms of SA-AKI involve multiple factors, including inflammatory responses, microcirculation dysfunction, and nephrotoxicity, all of which contribute to acute renal failure [3,4].
β-blockers are widely used in the management of cardiovascular diseases such as chronic heart failure [5] and myocardial infarction [6]. Attention has been gained for their potential in improving the prognosis of AKI and sepsis patients. The rationale for β-blocker use in SA-AKI lies in their ability to regulate the overactive sympathetic nervous system activity in sepsis [7,8]. By alleviating the harmful effects of excessive catecholamine release [9], β-blockers may enhance hemodynamic stability and reduce kidney ischemia; that are key factors in preventing kidney damage development. However, the current evidence on β-blocker use in SA-AKI is limited, and the findings remain contradictory. Some research has shown that β-blockers can lower heart rate and reduce both morbidity and mortality in SA-AKI patients [10–13], potentially improving kidney function and overall prognosis. However, other research suggests possible adverse effects, such as impaired kidney perfusion [14]. While some evidence suggests that β-blockers may exert protective effects on kidney function, their specific impact on overall patient outcomes (especially mortality, functional recovery, and long-term kidney function) remains unclear.
Given the complexity of SA-AKI and the uncertainty about the role of β-blocker in the treatment of SA-AKI patients, we used the MIMIC-IV database to conduct the current study in assessing whether β-blockers can improve outcomes in SA-AKI patients, in order to provide solid scientific evidence for clinical decision-making.
Materials and methods
Database introduction
The MIMIC-IV database is a comprehensive public resource providing anonymized clinical information for ICU patients hospitalized at Beth Israel Deaconess Medical Center between 2008 and 2019. It provides comprehensive details of demographics, laboratory results, prescriptions, vital signs, and diagnostic codes, serving as an essential resource for research in critical care, epidemiology, and health outcomes. In this study, the first author, Chaonan Fan, who had completed the required training in human subjects research (Certification No.: 64447378), was responsible for data extraction and processing in compliance with ethical guidelines and the integrity of the research, and fully complied with the ethical and data access requirements of the MIMIC-IV database.
Study population
According to the Sepsis-3 definition, sepsis was identified as life-threatening organ dysfunction caused by infection, reflected by an acute increase in SOFA score of ≥2 points from baseline. For patients without documented pre-hospital SOFA scores, a baseline SOFA of 0 was assumed [15,16]. AKI was defined based on the KDIGO criteria as any of the following: an increase in serum creatinine by ≥0.3 mg/dL within 48 h; or an increase to ≥1.5 times the baseline level within 7 days; or a urine output of <0.5 mL/kg/h for at least 6 h [17]. The exclusion criteria included the following: (1) Patients with a hospital stay <48 h. (2) Individuals under the age of 18. This study compared outcomes in SA-AKI patients who were administered β-blockers following AKI diagnosis with those who were not.
β-blocker exposure
Exposure to β-blockers was defined as having a prescription for a β-blocker following AKI diagnosis, including Metoprolol, Bisoprolol, Atenolol, Propranolol, Nadolol, and Esmolol (Supplementary Tables S1 and S2).
Data collection
Data for the research were extracted from a PostgreSQL database containing detailed hospitalization records. Only data from the first day of hospitalization were used to ensure consistency and reduce variability due to changes in clinical status. The collected variables include demographic data (sex, age, ethnicity), vital signs (diastolic blood pressure [DBP], systolic blood pressure [SBP], oxygen saturation [SpO2]), heart rate, and clinical conditions (hypertension, diabetes, heart failure, arrhythmias). Treatment-related variables include mechanical ventilation, continuous kidney replacement therapy (CRRT), and the use of vasopressors. Laboratory data comprised of serum creatinine, platelet count, urine output, and white blood cell count (WBC). Additionally, severity scores, such as the Glasgow Coma Scale (GCS) and the Sequential Organ Failure Assessment (SOFA), were recorded. Data were collected on the use of long-acting β-blockers (e.g., propranolol, atenolol, bisoprolol, nadolol, and metoprolol) and short-acting β-blockers (esmolol), including drug name, start time, dosage, administration route, and end time.
Outcomes
The 30-day mortality was set as the primary outcome, with secondary outcomes comprising of 90-day mortality, 365-day mortality, and duration of patients hospital stay.
Statistical analysis
All study variables had missing data below 5%. The missing data were then imputed using the ‘mice’ package in R. The experimental group included SA-AKI patients who received β-blockers during hospitalization, while the control group consisted of those who did not. For non-normally distributed continuous variables, the Mann-Whitney U test was used. Results were presented as medians and interquartile ranges. Categorical variables were represented by counts and percentages, and analyzed by the chi-square test. To ensure balance in baseline characteristics between the two groups, PSM with a 0.05 caliper value was applied. The following variables were included in the PSM model: age, sex, race, heart failure, arrhythmias, hypertension, diabetes, heart rate, SBP, DBP, SpO2, CRRT, mechanical ventilation, vasopressin use, urine output, serum creatinine, WBC, platelet count, GCS, SOFA, and AKI stage. Kaplan–Meier curves were used to depict mortality at 30, 90, and 365 days. Subsequently, a multivariable Cox proportional hazards regression was performed in the matched cohort as a doubly robust method, adjusting for β-blocker use and the same covariates as in the PSM model. Further subgroup analyses were carried out based on race, sex, age, AKI stage, comorbidities, GCS score, SOFA score, CRRT use, and vasoactive drug administration, with hazard ratios (HRs) and 95% confidence intervals (CIs) calculated for each subgroup. Statistical analyses were performed using R software (version 4.2.2), with significance set at p < 0.05.
Results
Patient characteristics
Patient selection process was described in Figure 1. We identified 26,774 individuals who were first-time ICU admissions and met the Sepsis 3.0 diagnostic criteria. After applying exclusion criteria, 12,484 patients met the KDIGO definition for AKI. Of these, 3,864 who had received β-blockers before hospital admission or AKI diagnosis were excluded. The remaining 8,620 patients were classified as SA-AKI, including 3,159 who received β-blockers and 5,461 who did not. PSM analysis resulted in 2,896 matched pairs of SA-AKI patients for further study. Table 1 depicts significant differences in variables between the β-blocker and non-β-blocker groups before PSM (p < 0.05). The β-blocker group was older in age (70 vs. 64, p < 0.001) and higher in incidences of heart failure (36 vs. 25%, p < 0.001) and hypertension (46 vs. 30%, p < 0.001). Differences in race distribution were also noted, with a higher percentage of Caucasian patients in the β-blocker group (71 vs. 65%, p < 0.001). Notable differences in vital signs were also observed, including DBP (60 vs. 61 mmHg, p < 0.001) and SpO2 (98 vs. 97%, p < 0.001). Moreover, the β-blocker group had greater use of mechanical ventilation (84 vs. 69%, p < 0.001), vasopressin (18 vs. 16%, p = 0.020), and CRRT (10 vs. 7.3%, p < 0.001). Lab values such as platelet count and creatinine also showed statistically significant differences. After PSM, these differences in variables were balanced between the two groups, with p-values exceeding 0.05 for most variables and the standardized mean differences (SMDs) below 0.1 (Figure 2).
Figure 1.
Flow chart of the research.
Table 1.
Baseline characteristics of the patient.
| Variables | Before PSM |
After PSM |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Total N = 8,620a |
Non-blockers N = 5,461a |
Blockers N = 3,159a | SMDb | p-Valuec | Total N = 5,792a |
Non blockers N = 2,896a |
Blockers N = 2,896a |
SMDb | p-Valuec | |
| Age | 66.00 (54.00, 78.00) | 64.00 (52.00, 76.00) | 70.00 (59.00, 80.00) | −0.32 | <0.001 | 68.00 (57.50, 79.50) | 68.00 (57.00, 80.00) | 69.00 (58.00, 79.00) | 0.00 | >0.9 |
| Sex, n (%) | 0.05 | 0.019 | 0.02 | 0.5 | ||||||
| Female | 3,664 (43%) | 2,373 (43%) | 1,291 (41%) | 2,409 (42%) | 1,217 (42%) | 1,192 (41%) | ||||
| Male | 4,956 (57%) | 3,088 (57%) | 1,868 (59%) | 3,383 (58%) | 1,679 (58%) | 1,704 (59%) | ||||
| Race, n (%) | 0.12 | <0.001 | 0.01 | >0.9 | ||||||
| Caucasian | 5,809 (67%) | 3,575 (65%) | 2,234 (71%) | 4,018 (69%) | 2,002 (69%) | 2,016 (70%) | ||||
| African | 732 (8.5%) | 515 (9.4%) | 217 (6.9%) | 421 (7.3%) | 212 (7.3%) | 209 (7.2%) | ||||
| Other | 2,079 (24%) | 1,371 (25%) | 708 (22%) | 1,353 (23%) | 682 (24%) | 671 (23%) | ||||
| Heart failure, n (%) | 0.25 | <0.001 | 0.02 | 0.5 | ||||||
| No | 6,103 (71%) | 4,097 (75%) | 2,006 (64%) | 3,804 (66%) | 1,914 (66%) | 1,890 (65%) | ||||
| Yes | 2,517 (29%) | 1,364 (25%) | 1,153 (36%) | 1,988 (34%) | 982 (34%) | 1,006 (35%) | ||||
| Arrhythmias, n (%) | 0.01 | 0.6 | 0.01 | 0.6 | ||||||
| No | 8,147 (95%) | 5,167 (95%) | 2,980 (94%) | 5,468 (94%) | 2,730 (94%) | 2,738 (95%) | ||||
| Yes | 473 (5.5%) | 294 (5.4%) | 179 (5.7%) | 324 (5.6%) | 166 (5.7%) | 158 (5.5%) | ||||
| Hypertension, n (%) | 0.32 | <0.001 | 0.02 | 0.4 | ||||||
| No | 5,515 (64%) | 3,798 (70%) | 1,717 (54%) | 3,325 (57%) | 1,678 (58%) | 1,647 (57%) | ||||
| Yes | 3,105 (36%) | 1,663 (30%) | 1,442 (46%) | 2,467 (43%) | 1,218 (42%) | 1,249 (43%) | ||||
| Diabetes, n (%) | 0.04 | 0.11 | 0.00 | 0.9 | ||||||
| No | 8,285 (96%) | 5,235 (96%) | 3,050 (97%) | 5,586 (96%) | 2,794 (96%) | 2,792 (96%) | ||||
| Yes | 335 (3.9%) | 226 (4.1%) | 109 (3.5%) | 206 (3.6%) | 102 (3.5%) | 104 (3.6%) | ||||
| Heart rate (beats/min) | 86.00 (75.00, 99.00) | 86.00 (74.00, 99.00) | 86.00 (77.00, 99.00) | −0.06 | 0.012 | 86.00 (76.00, 99.00) | 87.00 (75.00, 100.00) | 86.00 (77.00, 98.00) | 0.00 | 0.8 |
| SBP (mmHg) | 112.00 (104.00, 123.00) | 112.00 (104.00, 124.00) | 112.00 (105.00, 123.00) | −0.01 | 0.3 | 112.00 (105.00, 123.00) | 113.00 (104.00, 124.00) | 112.00 (105.00, 123.00) | 0.01 | >0.9 |
| DBP (mmHg) | 60.00 (54.00, 67.00) | 61.00 (55.00, 68.00) | 60.00 (54.00, 66.00) | 0.13 | <0.001 | 60.00 (54.00, 66.00) | 60.00 (54.00, 67.00) | 60.00 (54.00, 66.00) | 0.01 | 0.6 |
| SpO2 (%) | 97.00 (96.00, 99.00) | 97.00 (96.00, 99.00) | 98.00 (96.00, 99.00) | −0.12 | <0.001 | 98.00 (96.00, 99.00) | 98.00 (96.00, 99.00) | 98.00 (96.00, 99.00) | −0.02 | >0.9 |
| CRRT, n (%) | 0.10 | <0.001 | 0.01 | 0.7 | ||||||
| No | 7,898 (92%) | 5,060 (93%) | 2,838 (90%) | 5,236 (90%) | 2,613 (90%) | 2,623 (91%) | ||||
| Yes | 722 (8.4%) | 401 (7.3%) | 321 (10%) | 556 (9.6%) | 283 (9.8%) | 273 (9.4%) | ||||
| Ventilation, n (%) | 0.35 | <0.001 | 0.01 | 0.6 | ||||||
| No | 2,219 (26%) | 1,701 (31%) | 518 (16%) | 1,012 (17%) | 499 (17%) | 513 (18%) | ||||
| Yes | 6,401 (74%) | 3,760 (69%) | 2,641 (84%) | 4,780 (83%) | 2,397 (83%) | 2,383 (82%) | ||||
| Vasopressin, n (%) | 0.05 | 0.020 | 0.01 | 0.7 | ||||||
| No | 7,187 (83%) | 4,592 (84%) | 2,595 (82%) | 4,752 (82%) | 2,371 (82%) | 2,381 (82%) | ||||
| Yes | 1,433 (17%) | 869 (16%) | 564 (18%) | 1,040 (18%) | 525 (18%) | 515 (18%) | ||||
| Urine output (mL) | 80.00 (40.00, 160.00) | 85.00 (40.00, 175.00) | 75.00 (40.00, 150.00) | 0.14 | <0.001 | 75.00 (37.00, 150.00) | 75.00 (35.00, 150.00) | 75.00 (40.00, 150.00) | −0.04 | 0.15 |
| Creatinine (mg/dL) | 1.00 (0.70, 1.60) | 1.00 (0.70, 1.60) | 1.00 (0.70, 1.60) | 0.00 | 0.012 | 1.00 (0.70, 1.60) | 1.00 (0.70, 1.60) | 1.00 (0.70, 1.60) | −0.01 | 0.6 |
| WBC (×109/L) | 10.00 (7.00, 13.60) | 9.90 (6.80, 13.70) | 10.10 (7.30, 13.50) | 0.00 | 0.089 | 10.00 (7.10, 13.70) | 9.80 (6.90, 13.80) | 10.10 (7.30, 13.50) | −0.01 | 0.2 |
| Platelets (×109/L) | 160.50 (106.00, 226.00) | 164.00 (106.00, 230.00) | 155.00 (105.00, 217.00) | 0.08 | <0.001 | 158.00 (107.00, 219.00) | 160.00 (107.00, 219.50) | 155.00 (106.00, 218.00) | 0.00 | 0.4 |
| GCS | 15.00 (15.00, 15.00) | 15.00 (15.00, 15.00) | 15.00 (15.00, 15.00) | 0.02 | 0.001 | 15.00 (15.00, 15.00) | 15.00 (15.00, 15.00) | 15.00 (15.00, 15.00) | 0.01 | 0.10 |
| SOFA | 6.00 (4.00, 9.00) | 6.00 (4.00, 9.00) | 6.00 (4.00, 9.00) | −0.03 | 0.031 | 6.00 (4.00, 9.00) | 6.00 (4.00, 9.00) | 6.00 (4.00, 9.00) | 0.02 | 0.8 |
| AKI stage, n (%) | 0.18 | <0.001 | 0.01 | >0.9 | ||||||
| Stage 1 | 1,487 (17%) | 1,071 (20%) | 416 (13%) | 822 (14%) | 414 (14%) | 408 (14%) | ||||
| Stage 2 | 4,009 (47%) | 2,506 (46%) | 1,503 (48%) | 2,761 (48%) | 1,372 (47%) | 1,389 (48%) | ||||
| Stage 3 | 3,124 (36%) | 1,884 (34%) | 1,240 (39%) | 2,209 (38%) | 1,110 (38%) | 1,099 (38%) | ||||
SMD: standardized mean difference; WBC: white blood cell count; SBP: systolic blood pressure; SpO2: blood oxygen saturation; DBP: diastolic blood pressure; CRRT: continuous renal replacement therapy; SOFA: Sequential Organ Failure Assessment; GCS: Glasgow Coma Scale.
aMedian (Q1, Q3); n (%).
bStandardized mean difference.
cWilcoxon rank sum test; Pearson’s chi-squared test.
Figure 2.
Standardized mean differences for variables before and after PSM. SBP: systolic blood pressure; WBC: white blood cell count; DBP: diastolic blood pressure; SpO2: blood oxygen saturation; CRRT: continuous renal replacement therapy; SOFA: Sequential Organ Failure Assessment; GCS: Glasgow Coma Scale.
β-blocker usage and clinical outcomes
After PSM, 5,792 patients were included, with 2,896 in the β-blocker group and 2,896 in the non-β-blocker group. Kaplan-Meier analysis (Figure 3) showed that the 30-day, 90-day, and 365-day survival rates were considerably higher in the β-blocker group compared to the non-β-blocker group (p < 0.001). After adjusting for potential confounding variables in the multivariate Cox regression model, the use of β-blockers remained significantly associated with improved survival outcomes. As presented in Table 2, the 30-day mortality rates were 858/2,896 (29.6%) in the non-β-blocker group and 504/2,896 (17.4%) in the β-blocker group (p < 0.001), with an HR of 0.58 (95% CI: 0.52–0.65). The 90-day mortality rates were 1,059/2,896 (36.6%) and 702/2,896 (24.2%), respectively (p < 0.001), with an HR of 0.55 (95% CI: 0.50–0.61). The 365-day mortality rate was also significantly lower in the β-blocker group, at 934/2,896 (32.3%) compared to 1,298/2,896 (44.8%) in the non-β-blocker group (p < 0.001), with an HR of 0.59 (95% CI: 0.54–0.64). Additionally, patients receiving β-blockers had a longer median hospital stay than those who did not (12.6 [7.5–21.0] vs. 11.0 [6.7–18.0] days, p < 0.001).
Figure 3.
The Kaplan-Meier survival analysis comparing mortality at 30, 90, and 365 days in patients receiving and not receiving β-blockers. The red line indicates the β-blocker group, while the blue line indicates the non β-blocker group.
Table 2.
Clinical out comes associated with β-blocker treatment.
| Variables | Non β-blocker | β-blocker | p-Value | HR (95% CI) |
|---|---|---|---|---|
| After PSM | n = 2,896 | n = 2,896 | ||
| Primary outcome | ||||
| 30-day mortality, n (%) | 858 (29.6) | 504 (17.4) | <0.001 | 0.58 (0.52–0.65) |
| Secondary outcomes | ||||
| 90-day mortality, n (%) | 1,059 (36.6) | 702 (24.2) | <0.001 | 0.55 (0.50–0.61) |
| 365-day mortality, n (%) | 1,298 (44.8) | 934 (32.3) | <0.001 | 0.59 (0.54–0.64) |
| Length of hospital stay (days) | 11.0 [6.7, 18.2] | 12.6 [7.5, 21.0] | <0.001 | |
Long-acting β-blocker usage and clinical outcomes
After PSM, 2,844 patients not prescribed with long-acting β-blockers were matched with the same number of patients who received (Supplementary Table S3). These drugs include atenolol, metoprolol, nadolol, bisoprolol, and propranolol. Kaplan-Meier survival analysis (Figure 4) revealed better survival outcomes at 30, 90, and 365 days in the long-acting β-blocker group, in contrast to the non-user group (p < 0.001). The impact of long-acting β-blocker treatment on mortality at 30, 90, and 365 days was examined using multivariate Cox regression analysis, with further adjustments for residual confounding factors. This model estimates hazard ratios while accounting for time-to-event data and the influence of multiple covariates. Therefore, the additional multivariate adjustments after PSM provided better control of potential confounders and enhanced the robustness of our findings. As indicated in Table 3, the 30-day mortality rate was notably reduced in the long-acting β-blocker group in comparison to the non-user group (486/2,844 [17.0%] vs. 831/2,844 [29.2%], p < 0.001, HR 0.55, 95% CI: 0.49–0.61). Similarly, 90-day mortality was reduced in the long-acting β-blocker group (686/2,844 [24.1%] vs. 1,026/2,844 [36.0%], p < 0.001, HR 0.54, 95% CI: 0.49–0.59). The 365-day mortality was also reduced in the long-acting β-blocker group (910/2,844 [31.9%] vs. 1,280/2,844 [45.0%], p < 0.001), with an HR of 0.57 (95% CI: 0.52–0.62). Additionally, patients receiving long-acting β-blockers experienced longer median hospital stay than those who did not (12.6 [7.5–21.0] vs.11.1 [6.7–18.2], p < 0.001).
Figure 4.
The Kaplan-Meier survival analysis comparing mortality at 30, 90, and 365 days in patients receiving and not receiving long-acting β-blockers. The red line indicates the long-acting β-blocker group, while the blue line indicates the non long-acting β-blocker group.
Table 3.
Clinical outcomes associated with long-acting β-blocker treatment.
| Variables | Non long-acting β-blocker | Long-acting β-blocker | p-Value | HR (95% CI) |
|---|---|---|---|---|
| After PSM | n = 2,844 | n = 2,844 | ||
| Primary outcome | ||||
| 30-day mortality, n (%) | 831 (29.2) | 486 (17.0) | <0.001 | 0.55 (0.49–0.61) |
| Secondary outcomes | ||||
| 90-day mortality, n (%) | 1,026 (36.0) | 686 (24.1) | <0.001 | 0.54 (0.49–0.59) |
| 365-day mortality, n (%) | 1,280 (45.0) | 910 (31.9) | <0.001 | 0.57 (0.52–0.62) |
| Length of hospital stay (days) | 11.1 [6.7, 18.2] | 12.6 [7.5, 21.0] | <0.001 | |
Short-acting β-blocker usage and clinical outcomes
After PSM, 317 patients were enrolled in this analysis, with 80 patients assigned to the short-acting β-blocker (esmolol) group and 237 patients to the non-β-blocker group (Supplementary Table S4). The Kaplan-Meier survival curve indicates no notable differences in 30-day (p = 0.29), 90-day (p = 0.45), or 365-day (p = 0.20) survival rates between the two groups (Figure 5). The correlation between short-acting β-blockers and mortality at 30, 90, and 365 days were evaluated using multivariate Cox regression analysis (Table 4). The 30-day survival showed no difference between the two groups (27/80 [33.8%] vs. 67/237 [28.3%], p = 0.08). Similarly, there was no significant difference in 90-day mortality (30/80 [37.5%] vs. 81/237 [34.2%], p = 0.07) or 365-day mortality (37/80 [46.3%] vs. 92/237 [38.8%], p = 0.14). Patients receiving short-acting β-blockers experienced longer median hospital stays than those who did not (13.2 [6.2–19.0] vs. 13.0 [7.6–24.0] days, p < 0.001).
Figure 5.
The Kaplan-Meier survival analysis comparing mortality at 30, 90, and 365 days in patients receiving and not receiving short-acting β-blockers. The red line indicates the short-acting β-blocker group, while the blue line indicates the non short-acting β-blocker group.
Table 4.
Clinical outcomes associated with short-acting β-blocker treatment.
| Variables | Non short-acting β-blocker | Short-acting β-blocker | p-Value | HR (95% CI) |
|---|---|---|---|---|
| After PSM | n = 237 | n = 80 | ||
| Primary outcome | ||||
| 30-day mortality, n (%) | 67 (28.3) | 27 (33.8) | 0.08 | 0.70 (0.47–1.05) |
| Secondary outcomes | ||||
| 90-day mortality, n (%) | 81 (34.2) | 30 (37.5) | 0.07 | 0.71 (0.49–1.03) |
| 365-day mortality, n (%) | 92 (38.8) | 37 (46.3) | 0.14 | 0.77 (0.55–1.09) |
| Length of hospital stay (days) | 13.0 [7.6, 24.0] | 13.2 [6.2, 19.0] | <0.001 | |
Subgroup analyses
The subgroups were categorized by age, sex, race, comorbidities, mechanical ventilation, CRRT, vasopressor use, and AKI stage. In the subgroup analysis of SA-AKI patients, β-blocker treatment showed a consistent and significant protective effect on 30-day mortality in most subgroups, as illustrated in Figure 6. The overall interaction p-values demonstrated no significant interactions across most subgroups, except for the vasopressor subgroup, which showed a notable interaction (interaction p-value = 0.008).
Figure 6.
Subgroup analysis exploring the correlation of β-blocker administration with 30-day mortality rates. HR: hazard ratio; CRRT: continuous renal replacement therapy; CI: confidence interval.
Discussion
The therapeutic role of β-blockers in SA-AKI patients remains unclear, and prior observational studies have assessed their impact on patient prognosis. According to our study, long-acting β-blocker therapy reduces the mortality risk at 30, 90, and 365 days in ICU patients with SA-AKI. Conversely, short-acting β-blockers, including esmolol, did not lead to a notable decrease in mortality among these patients.
Sepsis is a life-threatening condition with high mortality caused by infection and a dysregulated host immune response. AKI is a common and severe complication of sepsis [18,19]. The pathogenesis of sepsis-associated AKI is complex. Excessive activation of the sympathetic nervous system (SNS) playing a critical role. Overactivation of the SNS not only induces renal vasoconstriction and hypoperfusion, exacerbating ischemia and hypoxia, but also promotes the release of pro-inflammatory cytokines, intensifying inflammatory responses and accelerating renal dysfunction [20–23]. In our study, we found that administration of β-blockers after AKI diagnosis was associated with significant improvement in clinical outcomes, suggesting that β-blockade may alleviate sepsis-associated AKI by inhibiting sympathetic overactivation and modulating inflammatory responses. This finding is consistent with Liu et al.’s report that perioperative β-blocker use significantly reduced the risk of contrast-induced AKI [24]. Moreover, as conventional cardiovascular agents, β-blockers have multiple beneficial effects, including heart rate control, reduction of myocardial oxygen consumption, and anti-inflammatory properties, supporting their potential therapeutic value in sepsis and AKI.
The therapeutic value of β-blockers in treating sepsis and SA-AKI may stem from their ability to inhibit excessive sympathetic activation and reduce the release of inflammatory mediators. Several investigations have indicated that β-blockers could decrease mortality in patients with sepsis and SA-AKI. Morelli et al. initiated a randomized controlled trial (RCT) evaluating the clinical outcomes of β-blocker treatment in sepsis patients. They found that esmolol therapy markedly lowered 28-day mortality in contrast to placebo (49.4 vs. 80.5%, p < 0.001), with no reported adverse events or safety concerns [10]. A secondary analysis by Fuchs et al. of a prospective observational analysis at a single center suggested that maintaining β-blocker therapy was linked to reduced 90-day mortality in sepsis patients (40.7 vs. 52.7%; p = 0.046) [25]. A single-center observational cohort analysis reported that long-term β-blocker treatment reduced mortality among sepsis in patients in the medical ward (OR 0.38, 95% CI: 0.148–0.976) [11]. Furthermore, subgroup analysis demonstrated consistent survival benefits associated with β-blocker use across various patient subgroups, with no significant interactions observed in most subgroups.
The effect of esmolol in treating sepsis has been evaluated in many studies. Evidence suggests that ultra-short-acting β-blockers, like esmolol and landiolol, can reduce 28-day mortality in patients with tachycardic sepsis [12]. Milrinone and esmolol together have been found to improve 28-day survival in sepsis patients [26]. However, Whitehouse et al. conducted the STRESS-L randomized clinical trial and found that landiolol did not significantly reduce organ dysfunction in patients with septic shock, nor did it improve 28- or 90-day mortality compared to standard care [27]. The multicenter randomized clinical trial demonstrated that the ultra-short-acting β-blocker landiolol effectively controlled heart rate in patients with septic shock and persistent tachycardia without increasing vasopressor requirements, with similar safety profiles and 28-day mortality compared to standard care [28]. In our study, 80 patients were administered with esmolol within 48 h of admission, while 237 matched controls did not receive β-blocker therapy. Our results indicated that, following PSM, esmolol did not significantly affect 30-day, 90-day, or 365-day mortality. Our findings differ from previous studies, is likely due to the following reasons: Firstly, although PSM was employed to balance baseline covariates, unobserved confounding factors may still exist. For instance, esmolol-treated patients may have more severe underlying conditions (e.g., higher cardiovascular risk or multiple organ dysfunction), which could offset its treatment effects, resulting in no significant reduction assumption in mortality. Secondly, the dose and administration time of esmolol are key factors affecting its efficacy. Different dosing regimens may be used in different studies. For example, higher doses or earlier interventions may provide greater protection, whereas insufficient doses or delayed interventions could reduce its efficacy. This study did not report specific doses or heart rate control targets, potentially limiting the interpretability of the results. Thirdly, deaths in ICU patients are usually the result of multiple factors, such as the type of infection pathogens (bacteria, viruses, or fungi), severity of sepsis, organ failure, and the use of other drugs or treatments (such as mechanical ventilation, CRRT), etc. The impact of long-acting β-blockers on sepsis-related mortality was unclear. As early as in 1969, Berk et al. found that dogs receiving endotoxin before propranolol treatment showed lower mortality and less organ congestion [29]. Atenolol, a selective β-blocker, has been reported to modulate immune responses and potentially increase survival among patients with sepsis [30]. In this large-scale study involving 2,844 sepsis patients, PSM was applied to balance baseline characteristics. Our findings align with previous studies, which have reported a notable mortality reduction in SA-AKI patients administered long-acting β-blockers.
Interestingly, the β-blocker group experienced longer hospital stays than the control group, possibly due to several factors: Firstly, the extended hospital stays may reflect the need for more intensive monitoring and for recovery. Specifically, reduced mortality in the β-blocker group suggests more patients with severe conditions survived but required longer monitoring and observation. Secondly, although β-blockers may help stabilize patient conditions during the acute phase, their usage is most likely associated with a longer recovery time as patients undergo further restoration of organ function. Additionally, it is plausible that patients in the β-blocker group had more cardiovascular comorbidities, necessitating prolonged hospitalization. We believe that these factors are the underlying contributors to the extended length of hospitalization. Subgroup analysis revealed a notable interaction between β-blocker treatment and 30-day outcomes within the vasopressor subgroup (interaction p = 0.008). The HR was lower in patients using vasopressors, suggesting greater efficacy of β-blockers in this subgroup. This result may be explained by more severe conditions in vasopressor users, where β-blockers offer greater protection by improving cardiac function, mitigating sympathetic overactivation, and lowering arrhythmia risk. Additionally, intensive monitoring and treatment in these patients may enhance efficacy, while selection bias and patient characteristics may also influence the results.
Limitations of current study should be noted. Firstly, the lack of detailed documentation on the dosage of β-blockers used. Due to the variety of β-blockers used in this study, including metoprolol, bisoprolol, atenolol, propranolol, nadolol, and esmolol, each with different pharmacokinetic properties, half-lives, and receptor selectivities, it was not possible to provide a standardized dosage (such as equivalent doses or standard daily doses). These differences may impact the comparability of the results, and thus, future prospective studies should more systematically record and analyze β-blocker dosages to more comprehensively assess their potential impact on patient outcomes. Secondly, due to its retrospective design, this study does not confirm a causal link between β-blocker therapy and mortality. Thirdly, although PSM was applied to minimize baseline differences, it may not fully account for time-varying confounders such as the clinical condition at the time of vasopressin or β-blocker administration. Moreover, restricting the β-blocker group to patients who received the drug only after SA-AKI diagnosis may have introduced both selection bias and immortal time bias. Patients who died early could not have received β-blockers and were therefore classified into the control group, potentially leading to an overestimation of mortality in this group. Conversely, patients who initiated β-blockers later were required to have survived the initial untreated period, potentially leading to an underestimation of mortality in the β-blocker group. This limitation could introduce residual bias and should be considered when interpreting the findings. Fourthly, since this study was carried out in a single-center cohort primarily involving critically ill individuals, the generalizability of its findings to broader populations or varied clinical settings may be limited. Finally, there is a critical need for a rigorous, randomized, double-blind prospective trial to investigate the possible benefits of β-blockers, including different formulations, dosages, and administration routes, in SA-AKI patients to further investigate their impact on prognosis.
Conclusion
β-blocker therapy is linked to better survival outcomes in patients with SA-AKI at 30, 90, and 365 days. Long-acting β-blocker treatment may protect SA-AKI patients by reducing mortality at 30, 90, and 365 days. In contrast, short-acting β-blockers, such as esmolol, show no impact on mortality in SA-AKI patients.
Supplementary Material
Funding Statement
This research was supported by a grant from the National Natural Science Foundation of China (No. 82170746).
Ethical approval
The study is an analysis of a public database that has been approved by the Institutional Review Board. The Institutional Review Board of Beth Israel Deaconess Medical Center has approved data collection and MIMIC-IV for research purposes and agreed to waive informed consent.
Authors contributions
C.F. designed the study, searched and analyzed the data, and wrote the manuscript. W.G. analyzed the data. F.Y., H.W., Y.Y., and J.D. assisted with data analysis. G.Z. supervised the study and contributed to writing-review and editing. L.W. acquired funding and designed and supervised the study.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
The author can provide raw data upon request.
References
- 1.Zarbock A, Nadim MK, Pickkers P, et al. Sepsis-associated acute kidney injury: consensus report of the 28th Acute Disease Quality Initiative workgroup. Nat Rev Nephrol. 2023;19(6):401–417. doi: 10.1038/s41581-023-00683-3. [DOI] [PubMed] [Google Scholar]
- 2.Zarbock A, Koyner JL, Gomez H, et al. Sepsis-associated acute kidney injury-treatment standard. Nephrol Dial Transplant. 2023;39(1):26–35. Erratum in: Nephrol Dial Transplant. 2023 Nov 30;38(12):2858. doi: 10.1093/ndt/gfad198. doi: 10.1093/ndt/gfad142. [DOI] [PubMed] [Google Scholar]
- 3.Gómez H, Kellum JA.. Sepsis-induced acute kidney injury. Curr Opin Crit Care. 2016;22(6):546–553. doi: 10.1097/MCC.0000000000000356. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.He FF, Wang YM, Chen YY, et al. Sepsis-induced AKI: from pathogenesis to therapeutic approaches. Front Pharmacol. 2022;13:981578. doi: 10.3389/fphar.2022.981578. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kubon C, Mistry NB, Grundvold I, et al. The role of beta-blockers in the treatment of chronic heart failure. Trends Pharmacol Sci. 2011;32(4):206–212. doi: 10.1016/j.tips.2011.01.006. [DOI] [PubMed] [Google Scholar]
- 6.Cucherat M. Quantitative relationship between resting heart rate reduction and magnitude of clinical benefits in post-myocardial infarction: a meta-regression of randomized clinical trials. Eur Heart J. 2007;28(24):3012–3019. doi: 10.1093/eurheartj/ehm489. [DOI] [PubMed] [Google Scholar]
- 7.Annane D, Trabold F, Sharshar T, et al. Inappropriate sympathetic activation at onset of septic shock: a spectral analysis approach. Am J Respir Crit Care Med. 1999;160(2):458–465. doi: 10.1164/ajrccm.160.2.9810073. [DOI] [PubMed] [Google Scholar]
- 8.Dünser MW, Hasibeder WR.. Sympathetic overstimulation during critical illness: adverse effects of adrenergic stress. J Intensive Care Med. 2009;24(5):293–316. Erratum in: J Intensive Care Med. 2016 Sep;31(8):NP1. doi: 10.1177/0885066616661952. doi: 10.1177/0885066609340519. [DOI] [PubMed] [Google Scholar]
- 9.Sanfilippo F, Santonocito C, Morelli A, et al. Beta-blocker use in severe sepsis and septic shock: a systematic review. Curr Med Res Opin. 2015;31(10):1817–1825. doi: 10.1185/03007995.2015.1062357. [DOI] [PubMed] [Google Scholar]
- 10.Morelli A, Ertmer C, Westphal M, et al. Effect of heart rate control with esmolol on hemodynamic and clinical outcomes in patients with septic shock: a randomized clinical trial. JAMA. 2013;310(16):1683–1691. doi: 10.1001/jama.2013.278477. [DOI] [PubMed] [Google Scholar]
- 11.Guz D, Buchritz S, Guz A, et al. β-blockers, tachycardia, and survival following sepsis: an observational cohort study. Clin Infect Dis. 2021;73(4):e921–e926. doi: 10.1093/cid/ciab034. [DOI] [PubMed] [Google Scholar]
- 12.Hasegawa D, Sato R, Prasitlumkum N, et al. Effect of ultrashort-acting β-blockers on mortality in patients with sepsis with persistent tachycardia despite initial resuscitation: a systematic review and meta-analysis of randomized controlled trials. Chest. 2021;159(6):2289–2300. doi: 10.1016/j.chest.2021.01.009. [DOI] [PubMed] [Google Scholar]
- 13.Zhang J, Chen C, Liu Y, et al. Benefits of esmolol in adults with sepsis and septic shock: an updated meta-analysis of randomized controlled trials. Medicine. 2022;101(27):e29820. doi: 10.1097/MD.0000000000029820. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gore DC, Wolfe RR.. Hemodynamic and metabolic effects of selective beta1 adrenergic blockade during sepsis. Surgery. 2006;139(5):686–694. doi: 10.1016/j.surg.2005.10.010. [DOI] [PubMed] [Google Scholar]
- 15.Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315(8):801–810. doi: 10.1001/jama.2016.0287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.White KC, Serpa-Neto A, Hurford R, et al. Sepsis-associated acute kidney injury in the intensive care unit: incidence, patient characteristics, timing, trajectory, treatment, and associated outcomes. A multicenter, observational study. Intensive Care Med. 2023;49(9):1079–1089. doi: 10.1007/s00134-023-07138-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120(4):c179–c184. doi: 10.1159/000339789. [DOI] [PubMed] [Google Scholar]
- 18.Rudd KE, Johnson SC, Agesa KM, et al. Global, regional, and national sepsis incidence and mortality, 1990–2017: analysis for the Global Burden of Disease Study. Lancet. 2020;395(10219):200–211. doi: 10.1016/S0140-6736(19)32989-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Liu AB, Tan B, Yang P, et al. The role of inflammatory response and metabolic reprogramming in sepsis-associated acute kidney injury: mechanistic insights and therapeutic potential. Front Immunol. 2024;15:1487576. doi: 10.3389/fimmu.2024.1487576. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Zhang X, Zhang Y, Yuan S, et al. The potential immunological mechanisms of sepsis. Front Immunol. 2024;15:1434688. doi: 10.3389/fimmu.2024.1434688. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ferreira JA, Bissell BD.. Misdirected sympathy: the role of sympatholysis in sepsis and septic shock. J Intensive Care Med. 2018;33(2):74–86. doi: 10.1177/0885066616689548. [DOI] [PubMed] [Google Scholar]
- 22.Zhang W, Li Z, Li Z, et al. The role of the superior cervical sympathetic ganglion in ischemia reperfusion-induced acute kidney injury in rats. Front Med. 2022;9:792000. doi: 10.3389/fmed.2022.792000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Tan S, Zhou F, Zhang Z, et al. Beta-1 blocker reduces inflammation and preserves intestinal barrier function after open abdominal surgery. Surgery. 2021;169(4):885–893. doi: 10.1016/j.surg.2020.11.004. [DOI] [PubMed] [Google Scholar]
- 24.Liu J, Sun G, He Y, et al. Early β-blockers administration might be associated with a reduced risk of contrast-induced acute kidney injury in patients with acute myocardial infarction. J Thorac Dis. 2019;11(4):1589–1596. doi: 10.21037/jtd.2019.04.65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Fuchs C, Wauschkuhn S, Scheer C, et al. Continuing chronic beta-blockade in the acute phase of severe sepsis and septic shock is associated with decreased mortality rates up to 90 days. Br J Anaesth. 2017;119(4):616–625. doi: 10.1093/bja/aex231. [DOI] [PubMed] [Google Scholar]
- 26.Wang Z, Wu Q, Nie X, et al. Combination therapy with milrinone and esmolol for heart protection in patients with severe sepsis: a prospective, randomized trial. Clin Drug Investig. 2015;35(11):707–716. doi: 10.1007/s40261-015-0325-3. [DOI] [PubMed] [Google Scholar]
- 27.Whitehouse T, Hossain A, Perkins GD, et al. Landiolol and organ failure in patients with septic shock: the STRESS-L randomized clinical trial. JAMA. 2023;330(17):1641–1652. doi: 10.1001/jama.2023.20134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Rehberg S, Frank S, Černý V, et al. Landiolol for heart rate control in patients with septic shock and persistent tachycardia. A multicenter randomized clinical trial (Landi-SEP). Intensive Care Med. 2024;50(10):1622–1634. doi: 10.1007/s00134-024-07587-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Berk JL, Hagen JF, Beyer WH, et al. The treatment of endotoxin shock by beta adrenergic blockade. Ann Surg. 1969;169(1):74–81. doi: 10.1097/00000658-196901000-00007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Hong SY, Lai CC, Teng NC, et al. Premorbid use of selective beta-blockers improves sepsis incidence and course: human cohort and animal model studies. Front Med. 2023;10:1105894. doi: 10.3389/fmed.2023.1105894. [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 author can provide raw data upon request.