Introduction: Patients on extracorporeal membrane oxygenation (ECMO) often experience worse renal outcomes and higher mortality rates as the severity of kidney injury increases. Nevertheless, the in-hospital mortality risks of patients with end-stage renal disease (ESRD) are poorly understood. This study evaluated several prognostic factors associated with in-hospital mortality in patients with ESRD receiving ECMO therapy. Methods: This study reviewed the medical records of 90 adult patients with ESRD on venoarterial ECMO in intensive care units in Linkou Chang Gung Memorial Hospital between March 2009 and February 2022. Fourteen patients who died within 24 h of receiving ECMO support were excluded; the remaining 76 patients were enrolled. Demographic, clinical, and laboratory variables were retrospectively collected as survival predictors. The primary outcome was in-hospital mortality. Results: The overall in-hospital mortality rate was 69.7%. The most common diagnosis requiring ECMO support was postcardiotomy cardiogenic shock, and the most frequent ECMO-associated complication was infection. Multiple logistic regression analysis revealed that the Acute Physiology and Chronic Health Evaluation II (APACHE II) score on day 1 of ECMO support was an independent risk factor for in-hospital mortality. The APACHE II score demonstrated satisfactory discriminative power (0.788 ± 0.057) in the area under the receiver operating characteristic curve. The cumulative survival rates at the 6-month follow-up differed significantly (p < 0.001) between patients with APACHE II score ≤ 29 versus those with APACHE II score >29. Conclusion: For patients with ESRD on ECMO, the APACHE II score is an excellent predictor of in-hospital mortality.

Extracorporeal membrane oxygenation (ECMO) is used in patients in critical condition with severe reversible myocardial dysfunction or life-threatening respiratory failure. ECMO can provide temporary circulatory support for patients who do not respond well to conventional treatment. However, patients on ECMO often experience poor prognosis and high mortality rates; this is particularly true of those with kidney injuries, and the incidence of adverse outcomes and the overall mortality rate rise considerably as the severity of acute kidney injury (AKI) increases [1‒4]. In long-term follow-up, patients with AKI requiring dialysis exhibit higher rates of chronic kidney disease (CKD) and end-stage renal disease (ESRD) compared with those of patients with AKI who do not require dialysis [5]. Furthermore, CKD is associated with increased risks of in-hospital and mid-term mortality in patients receiving ECMO treatment [6]. These findings indicate that renal function is intimately associated with the prognosis of patients on ECMO.

Patients with ESRD experience substantially higher hospitalization rates, longer hospital stays, and higher in-hospital mortality rates than do those with normal renal function [7, 8]. These patients often present with serious comorbidities, require surgery, and have higher perioperative mortality and morbidity. Recent research revealed that such patients have increased postoperative mortality in elective surgeries, with an odds ratio ranging from 4.0 to 10.8 [9]. Additionally, patients with ESRD exhibit increased postoperative cardiovascular and infectious complications, such as myocardial infarction, stroke, sepsis, surgical site infection, and pneumonia [10].

ECMO implantation is an emergency operation that is typically performed under unstable hemodynamic conditions; the operative risk for this procedure is considerably higher than that of general elective surgeries. Although the outcomes of patients with AKI and CKD on ECMO have been analyzed, the risk and prognosis for patients with ESRD on ECMO have not been thoroughly investigated. Therefore, this study evaluated the prognosis of patients with ESRD receiving ECMO.

Patient Information and Data Collection

This study was approved by the Institutional Review Board of Chang Gung Memorial Hospital, and because the study retrospectively analyzed patient data, the requirement for informed consent was waived. We examined the medical records of 90 adult patients with ESRD on venoarterial ECMO in intensive care units (ICUs) of Linkou Chang Gung Memorial Hospital between March 2009 and February 2022. Patients with ESRD underwent either hemodialysis 3 times a week for 4 h each session or continuous ambulatory peritoneal dialysis for 4 exchanges each day. We excluded 14 patients who died within 24 h after receiving ECMO support and included the remaining 76 patients.

The following clinical information was collected retrospectively: demographic data and ICU scoring system (Acute Physiology and Chronic Health Evaluation II [APACHE II], Sequential Organ Failure Assessment [SOFA], Organ System Failure [OSF], and Simplified Acute Physiology Score II [SAPS II]) data on day 1 of ECMO support, the primary diagnosis indicating a need for ECMO support, underlying comorbidities, and ECMO-associated complications. Calculations for the ICU scoring system were completed using the worst physiological and laboratory values on day 1 of ECMO support. The primary outcome was in-hospital mortality. Follow-up information at 6 months after hospital discharge was obtained through a review of medical charts.

Clinical Management

A CAPIOX EBS system (Terumo, Tokyo, Japan) and a Medos Deltastream System (Medos, Aachen, Germany) were employed for ECMO support. A hollow fiber oxygenator (Hilite LT 700; Medos, Aachen, Germany) or a silicone oxygenator (Medtronic, Minneapolis, MN, USA) was incorporated into the ECMO circuit. Percutaneous cannulation with a 19–21 French cannula (inflow) and a 17–19 French cannula (outflow) (DLP Medtronic, Minneapolis, MN, USA) was performed using the cut-down method. A 6-French distal perfusion catheter was implanted to enhance the distal perfusion of the limb on cannulation. Initially, the ECMO flow was set to maintain the mean arterial pressure between 60 and 80 mm Hg, and the flow of the oxygenator was titrated on the basis of arterial blood gas results. During the first 24 h of ECMO support, heparin was routinely withheld to minimize postoperative bleeding. Then heparin was restarted and dose was adjusted to maintain the activated clotting time between 180 and 200 s. Continuous infusion of midazolam (0.04–0.2 mg/kg/h) was used for sedation and periodically withheld for neurological evaluation. Laboratory data were checked daily to maintain the platelet count at ≥100,000/mm3, hemoglobin at ≥10 g/dL, and fibrinogen at ≥100 mg/dL to minimize the risk of bleeding during ECMO support [11, 12]. Blood transfusion and inotropic agents were administered and titrated by clinicians to meet acceptable clinical targets. Ultrafiltration amount during each dialysis was evaluated and adjusted by nephrologists according to the clinical condition and fluid status.

Statistical Analysis

The primary analysis involved a comparison of in-hospital survivors with nonsurvivors. The Kolmogorov-Smirnov test was used to test all variables for normal distributions. Student’s t-test was employed to compare means of continuous variables and normally distributed data; the Mann-Whitney U test was applied in all other cases. The χ2 test or Fisher’s exact test was used to test categorical data. Descriptive statistics are expressed as means ± standard deviations. Risk factors were identified through univariate analysis. Statistically significant variables (p < 0.05) in the univariate analysis were further examined in multivariate analysis through multiple logistic regression with forward data elimination.

The Hosmer-Lemeshow goodness-of-fit test was used for calibration assessment to compare the number of observed and predicted deaths in risk groups for the entire range of death probabilities. The area under the receiver operating characteristic curve (AUROC) was applied for discrimination evaluation and compared with a nonparametric approach. AUROC analysis was also employed to calculate cutoff values, sensitivity, specificity, and overall correctness. Subsequently, the best Youden index (sensitivity + specificity – 1) was determined to calculate cutoff points. Cumulative survival curves were plotted as a function of time by using the Kaplan-Meier method and were compared using the log-rank test. All statistical tests were two-tailed, and statistical significance was set at p < 0.05. All data were analyzed using SPSS Statistics 26.0 for Windows (International Business Machines, Armonk, NY, USA).

Patient Characteristics

Between March 2009 and February 2022, 76 adult patients with ESRD received ECMO support in the ICU. Table 1 presents the demographic data and clinical characteristics of the survivors and nonsurvivors. The overall in-hospital mortality was 69.7% (53/76). The patients had an average age of 63 years and were mostly men (76%, 58/76). Most patients underwent hemodialysis (89%, 68/76), but the modality and duration of dialysis were not significantly associated with mortality. The nonsurviving patients had lower mean arterial pressures and higher scores on ICU scoring systems than surviving patients did. Table 2 lists the primary diagnoses indicating a need for ECMO support, underlying comorbidities, and ECMO-associated complications. Postcardiotomy cardiogenic shock (52.6%) was the most common indication for ECMO support, and hypertension and coronary artery disease (82.9%) were the most common comorbidities. Regarding ECMO-associated complications, new-onset infection (22.4%), surgical wound bleeding (18.4%), and gastrointestinal bleeding (17.1%) occurred most frequently. For the primary diagnoses indicating a need for ECMO support, only the incidence of acute myocardial infarction differed significantly between the survivors and nonsurvivors.

Table 1.

Patients’ demographic data and clinical characteristics on the first day of ECMO support

All patients (n = 76)Survivors (n = 23)Nonsurvivors (n = 53)p value
Age, years 63±10 60±12 64±9 NS (0.132) 
Gender (male/female) 58/18 19/4 39/14 NS (0.395) 
Dialysis modality (HD/PD) 68/8 21/2 47/6 NS (0.732) 
Dialysis duration (HD), years 8.0±7.8 8.8±7.9 7.6±7.9 NS (0.578) 
Dialysis duration (PD), years 5.0±2.6 4.0±3.6 5.5±2.5 NS (0.573) 
Duration of ECMO support, days 7±6 7±5 7±6 NS (0.866) 
ICU duration, days 23±30 34±41 18±22 NS (0.089) 
IABP (yes/no) 46/30 16/7 30/23 NS (0.288) 
Mean arterial pressure, mm Hg 47±13 53±12 45±13 0.010 
WBC count, cu/mm × 1,000 14±9 13±8 14±9 NS (0.552) 
Hemoglobin, g/dL 8.0±1.3 8.4±1.3 7.8±1.3 NS (0.061) 
Blood urea nitrogen, mg/dL 55.9±21.2 56.4±24.7 55.7±19.7 NS (0.894) 
Serum creatinine, mg/dL 7.14±3.15 7.61±3.64 6.94±2.93 NS (0.395) 
Sodium, meq/L 141±7 139±6 142±8 NS (0.077) 
Potassium, meq/L 4.2±1.2 4.4±1.4 4.1±1.2 NS (0.368) 
Calcium, mg/dL 8.1±1.1 8.2±1.0 8.0±1.2 NS (0.503) 
Phosphorus, mg/dL 6.3±2.7 6.4±3.2 6.2±2.5 NS (0.796) 
Magnesium, meq/L 1.8±0.3 1.8±0.3 1.8±0.4 NS (0.984) 
Albumin, g/dL 2.55±0.43 2.74±0.53 2.47±0.36 NS (0.071) 
Total bilirubin, mg/dL 1.2±0.9 1.0±0.5 1.3±1.0 NS (0.217) 
Troponin I, ng/mL 31±33 28±27 32±36 NS (0.679) 
Lactate, mg/dL 112±56 105±60 115±55 NS (0.507) 
AaDO2 276±188 288±188 271±190 NS (0.724) 
APACHE II score 33±8 28±6 35±8 <0.001 
SOFA score 13±3 11±2 13±3 0.004 
OSF score 4.1±0.8 3.6±0.7 4.3±0.8 0.001 
SAPS II 72±15 63±9 76±16 <0.001 
All patients (n = 76)Survivors (n = 23)Nonsurvivors (n = 53)p value
Age, years 63±10 60±12 64±9 NS (0.132) 
Gender (male/female) 58/18 19/4 39/14 NS (0.395) 
Dialysis modality (HD/PD) 68/8 21/2 47/6 NS (0.732) 
Dialysis duration (HD), years 8.0±7.8 8.8±7.9 7.6±7.9 NS (0.578) 
Dialysis duration (PD), years 5.0±2.6 4.0±3.6 5.5±2.5 NS (0.573) 
Duration of ECMO support, days 7±6 7±5 7±6 NS (0.866) 
ICU duration, days 23±30 34±41 18±22 NS (0.089) 
IABP (yes/no) 46/30 16/7 30/23 NS (0.288) 
Mean arterial pressure, mm Hg 47±13 53±12 45±13 0.010 
WBC count, cu/mm × 1,000 14±9 13±8 14±9 NS (0.552) 
Hemoglobin, g/dL 8.0±1.3 8.4±1.3 7.8±1.3 NS (0.061) 
Blood urea nitrogen, mg/dL 55.9±21.2 56.4±24.7 55.7±19.7 NS (0.894) 
Serum creatinine, mg/dL 7.14±3.15 7.61±3.64 6.94±2.93 NS (0.395) 
Sodium, meq/L 141±7 139±6 142±8 NS (0.077) 
Potassium, meq/L 4.2±1.2 4.4±1.4 4.1±1.2 NS (0.368) 
Calcium, mg/dL 8.1±1.1 8.2±1.0 8.0±1.2 NS (0.503) 
Phosphorus, mg/dL 6.3±2.7 6.4±3.2 6.2±2.5 NS (0.796) 
Magnesium, meq/L 1.8±0.3 1.8±0.3 1.8±0.4 NS (0.984) 
Albumin, g/dL 2.55±0.43 2.74±0.53 2.47±0.36 NS (0.071) 
Total bilirubin, mg/dL 1.2±0.9 1.0±0.5 1.3±1.0 NS (0.217) 
Troponin I, ng/mL 31±33 28±27 32±36 NS (0.679) 
Lactate, mg/dL 112±56 105±60 115±55 NS (0.507) 
AaDO2 276±188 288±188 271±190 NS (0.724) 
APACHE II score 33±8 28±6 35±8 <0.001 
SOFA score 13±3 11±2 13±3 0.004 
OSF score 4.1±0.8 3.6±0.7 4.3±0.8 0.001 
SAPS II 72±15 63±9 76±16 <0.001 

ECMO, extracorporeal membrane oxygenation; NS, not significant; HD, hemodialysis; PD, peritoneal dialysis; ICU, intensive care unit; IABP, intra-aortic balloon pump; WBC, white blood cell; AaDO2, alveolar-arterial oxygen tension difference; APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, Sequential Organ Failure Assessment; OSF, Organ System Failure; SAPS, Simplified Acute Physiology Score.

Table 2.

Primary diagnosis requiring ECMO support, underlying comorbidities, and complications

All patients, n = 76 (%)Survivors, n = 23 (%)Nonsurvivors, n = 53 (%)p value
Primary diagnosis for ECMO support 
Postcardiotomy cardiogenic shock 40 (52.6) 10 (43.5) 30 (56.6) NS (0.292) 
Acute myocardial infarction 19 (25.0) 10 (43.5) 9 (17.0) 0.014 
Sudden collapse status post CPCR 10 (13.2) 2 (8.7) 8 (15.1) NS (0.448) 
Cardiogenic shock with desaturation 5 (6.6) 1 (4.3) 4 (7.5) NS (1.000) 
Myocarditis 1 (1.3) 0 (0) 1 (1.9) NS (1.000) 
Post heart transplantation 1 (1.3) 0 (0) 1 (1.9) NS (1.000) 
Underlying comorbidities 
Diabetes mellitus 40 (52.6) 15 (65.2) 25 (47.2) NS (0.148) 
Hypertension 63 (82.9) 21 (91.3) 42 (79.2) NS (0.200) 
Coronary artery disease 63 (82.9) 21 (91.3) 42 (79.2) NS (0.200) 
Peripheral artery disease 16 (21.1) 4 (17.4) 12 (22.6) NS (0.606) 
Malignancy 7 (9.2) 2 (8.7) 5 (9.4) NS (1.000) 
Cirrhosis 6 (7.9) 2 (8.7) 4 (7.5) NS (1.000) 
Old cerebrovascular accident 12 (15.8) 1 (4.3) 11 (20.8) NS (0.072) 
Gout 15 (19.7) 4 (17.4) 11 (20.8) NS (0.735) 
Complications 
Lower limb ischemia 10 (13.2) 2 (8.7) 8 (15.1) NS (0.448) 
Seizure 5 (6.6) 1 (4.3) 4 (7.5) NS (1.000) 
Cerebral infarction 6 (7.9) 1 (4.3) 5 (9.4) NS (0.661) 
Brain hypoxia 2 (2.6) 0 (0) 2 (3.8) NS (1.000) 
Intracerebral hemorrhage 4 (5.3) 1 (4.3) 3 (5.7) NS (1.000) 
Cannula wound bleeding 5 (6.6) 1 (4.3) 4 (7.5) NS (1.000) 
Surgical wound bleeding 14 (18.4) 5 (21.7) 9 (17.0) NS (0.623) 
Gastrointestinal bleeding 13 (17.1) 2 (8.7) 11 (20.8) NS (0.200) 
Hemothorax 3 (3.9) 1 (4.3) 2 (3.8) NS (1.000) 
Other bleeding 7 (9.2) 3 (13.0) 4 (7.5) NS (0.426) 
New-onset infection 17 (22.4) 6 (26.1) 11 (20.8) NS (0.608) 
All patients, n = 76 (%)Survivors, n = 23 (%)Nonsurvivors, n = 53 (%)p value
Primary diagnosis for ECMO support 
Postcardiotomy cardiogenic shock 40 (52.6) 10 (43.5) 30 (56.6) NS (0.292) 
Acute myocardial infarction 19 (25.0) 10 (43.5) 9 (17.0) 0.014 
Sudden collapse status post CPCR 10 (13.2) 2 (8.7) 8 (15.1) NS (0.448) 
Cardiogenic shock with desaturation 5 (6.6) 1 (4.3) 4 (7.5) NS (1.000) 
Myocarditis 1 (1.3) 0 (0) 1 (1.9) NS (1.000) 
Post heart transplantation 1 (1.3) 0 (0) 1 (1.9) NS (1.000) 
Underlying comorbidities 
Diabetes mellitus 40 (52.6) 15 (65.2) 25 (47.2) NS (0.148) 
Hypertension 63 (82.9) 21 (91.3) 42 (79.2) NS (0.200) 
Coronary artery disease 63 (82.9) 21 (91.3) 42 (79.2) NS (0.200) 
Peripheral artery disease 16 (21.1) 4 (17.4) 12 (22.6) NS (0.606) 
Malignancy 7 (9.2) 2 (8.7) 5 (9.4) NS (1.000) 
Cirrhosis 6 (7.9) 2 (8.7) 4 (7.5) NS (1.000) 
Old cerebrovascular accident 12 (15.8) 1 (4.3) 11 (20.8) NS (0.072) 
Gout 15 (19.7) 4 (17.4) 11 (20.8) NS (0.735) 
Complications 
Lower limb ischemia 10 (13.2) 2 (8.7) 8 (15.1) NS (0.448) 
Seizure 5 (6.6) 1 (4.3) 4 (7.5) NS (1.000) 
Cerebral infarction 6 (7.9) 1 (4.3) 5 (9.4) NS (0.661) 
Brain hypoxia 2 (2.6) 0 (0) 2 (3.8) NS (1.000) 
Intracerebral hemorrhage 4 (5.3) 1 (4.3) 3 (5.7) NS (1.000) 
Cannula wound bleeding 5 (6.6) 1 (4.3) 4 (7.5) NS (1.000) 
Surgical wound bleeding 14 (18.4) 5 (21.7) 9 (17.0) NS (0.623) 
Gastrointestinal bleeding 13 (17.1) 2 (8.7) 11 (20.8) NS (0.200) 
Hemothorax 3 (3.9) 1 (4.3) 2 (3.8) NS (1.000) 
Other bleeding 7 (9.2) 3 (13.0) 4 (7.5) NS (0.426) 
New-onset infection 17 (22.4) 6 (26.1) 11 (20.8) NS (0.608) 

ECMO, extracorporeal membrane oxygenation; NS, not significant; CPCR, cardiopulmonary cerebral resuscitation.

Number (%) of patients with the condition who survived or died.

Hospital Mortality and Short-Term Prognosis

Table 3 presents the results of univariate analysis on all variables presented in Table 1. Five variables were determined to have prognostic value. Multivariate analysis was employed to analyze these 5 variables, and APACHE II score was identified as an independent prognostic factor. The logit of death for each patient was calculated using regression coefficients of the variables as follows: The logarithm of death odds = −4.14 + 0.160 × APACHE II score.

Table 3.

Variables showing prognostic significance

ParameterBeta coefficientStandard errorOdds ratios (95% CI)p value
Univariate logistic regression 
Mean arterial pressure −0.054 0.022 0.948 (0.908–0.989) 0.014 
APACHE II score 0.160 0.046 1.173 (1.071–1.284) 0.001 
SOFA score 0.286 0.107 1.331 (1.079–1.642) 0.008 
OSF score 1.177 0.389 3.245 (1.514–6.952) 0.002 
SAPS II 0.077 0.025 1.080 (1.029–1.133) 0.002 
Multivariate logistic regression 
APACHE II score 0.160 0.046 1.173 (1.071–1.284) 0.001 
Constant −4.140 1.400 0.016 0.003 
ParameterBeta coefficientStandard errorOdds ratios (95% CI)p value
Univariate logistic regression 
Mean arterial pressure −0.054 0.022 0.948 (0.908–0.989) 0.014 
APACHE II score 0.160 0.046 1.173 (1.071–1.284) 0.001 
SOFA score 0.286 0.107 1.331 (1.079–1.642) 0.008 
OSF score 1.177 0.389 3.245 (1.514–6.952) 0.002 
SAPS II 0.077 0.025 1.080 (1.029–1.133) 0.002 
Multivariate logistic regression 
APACHE II score 0.160 0.046 1.173 (1.071–1.284) 0.001 
Constant −4.140 1.400 0.016 0.003 

CI, confidence interval; APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, Sequential Organ Failure Assessment; OSF, Organ System Failure; SAPS, Simplified Acute Physiology Score.

Table 4 presents the results regarding the goodness-of-fit, calculated using the Hosmer-Lemeshow chi-squared statistic of predicted mortality risk and the values indicating the predictive accuracy of APACHE II score, SOFA score, OSF score, and SAPS II. The discriminatory power of the 4 scoring systems was compared, and the APACHE II score achieved the highest discriminatory power in the AUROC analysis. Table 5 presents the sensitivity, specificity, and overall correctness that enabled comparison of the predictive ability of each scoring system for in-hospital mortality. The APACHE II score had the best Youden index value and the highest overall prediction correctness. Figure 1 depicts the significant differences in cumulative survival rates (p < 0.001) between the patients on ECMO support with an APACHE II score ≤29 and those with an APACHE II score >29.

Table 4.

Comparison of calibration and discrimination of scoring methods for predicting in-hospital mortality

CalibrationDiscrimination
Hosmer-Lemeshow χ2dfp valueAUROC±SE95% CIp value
APACHE II score 5.526 0.700 0.788±0.057 0.676–0.899 <0.001 
SOFA score 13.906 0.031 0.697±0.064 0.572–0.821 0.007 
OSF score 0.182 0.913 0.720±0.062 0.599–0.841 0.002 
SAPS II 7.702 0.360 0.749±0.057 0.637–0.861 0.001 
CalibrationDiscrimination
Hosmer-Lemeshow χ2dfp valueAUROC±SE95% CIp value
APACHE II score 5.526 0.700 0.788±0.057 0.676–0.899 <0.001 
SOFA score 13.906 0.031 0.697±0.064 0.572–0.821 0.007 
OSF score 0.182 0.913 0.720±0.062 0.599–0.841 0.002 
SAPS II 7.702 0.360 0.749±0.057 0.637–0.861 0.001 

df, degree of freedom; AUROC, areas under the receiver operating characteristic curve; SE, standard error; CI, confidence intervals; APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, Sequential Organ Failure Assessment; OSF, Organ System Failure; SAPS, Simplified Acute Physiology Score.

Table 5.

Subsequent in-hospital mortality predicted on the first day of ECMO support

Predictive factorsCutoff pointYouden IndexSensitivity, %Specificity, %Overall correctness, %
APACHE II score 29a 0.49 75.5 73.9 74.7 
SOFA score 10a 0.30 86.8 43.5 65.2 
OSF score 3a 0.31 83.0 47.8 65.4 
SAPS II 72a 0.48 56.6 91.3 74.0 
Predictive factorsCutoff pointYouden IndexSensitivity, %Specificity, %Overall correctness, %
APACHE II score 29a 0.49 75.5 73.9 74.7 
SOFA score 10a 0.30 86.8 43.5 65.2 
OSF score 3a 0.31 83.0 47.8 65.4 
SAPS II 72a 0.48 56.6 91.3 74.0 

ECMO, extracorporeal membrane oxygenation; APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, Sequential Organ Failure Assessment; OSF, Organ System Failure; SAPS, Simplified Acute Physiology Score.

aValue giving the best Youden index.

Fig. 1.

Cumulative survival rate at 6-month follow-up for 76 patients with end-stage renal disease (ESRD) based on their Acute Physiology and Chronic Health Evaluation II (APACHE II) score on the first day of receiving extracorporeal membrane oxygenation (ECMO) support.

Fig. 1.

Cumulative survival rate at 6-month follow-up for 76 patients with end-stage renal disease (ESRD) based on their Acute Physiology and Chronic Health Evaluation II (APACHE II) score on the first day of receiving extracorporeal membrane oxygenation (ECMO) support.

Close modal

The overall in-hospital mortality rate was 69.7%, which is higher than the average short-term mortality rate in patients with cardiogenic shock receiving ECMO support (61.0%) [13]. In terms of renal function, this overall mortality rate was similar to that of patients receiving ECMO with AKI (57.0%–67.5%) and those with CKD (69.5%) but much higher than that of patients receiving ECMO without AKI (19.5%–45.3%) [2, 6, 14‒17]. This finding indicates that with ECMO support, patients with ESRD have a comparable risk of mortality to that of patients with AKI and CKD and a higher risk than that observed in patients without these conditions. In recent studies, patients receiving ECMO who have AKI and require dialysis experience a higher long-term mortality rate than that of patients receiving ECMO who have AKI but do not require dialysis. Patients receiving ECMO who do not recover from dialysis-requiring AKI exhibit worse long-term outcome than that of patients receiving ECMO who recover from dialysis-requiring AKI [5]. Besides, acute kidney disease (AKD) and CKD are associated with increased risks of all-cause mortality in patients receiving ECMO [4, 6, 18]. Therefore, renal function proves to have a significant influence on the outcome of patients receiving ECMO, which is consistent with the result of our study.

In subgroup comparisons, patients with ESRD with postcardiotomy cardiogenic shock and sudden collapse after cardiopulmonary cerebral resuscitation were discovered to have worse in-hospital mortality rates (75.0% vs. 60.2% and 80.0% vs. 69.0%), whereas patients with acute myocardial infarction experienced better outcome (47.4% vs. 64.2%) [13]. Of the 19 patients with acute myocardial infarction in our study, 4 experienced cardiogenic shock during cardiac catheterization or at the beginning of coronary artery bypass graft surgery. Therefore, immediate ECMO insertion may stabilize hemodynamic circulation soon and decrease overall mortality in this subgroup. Regarding ECMO-associated complications, our patients experienced a higher incidence of gastrointestinal hemorrhage and new-onset infection compared with those in patients receiving ECMO without ESRD [19]. Increased gastrointestinal bleeding may result from uremic platelet dysfunction, hypergastrinemia, and vascular anomaly of the gastrointestinal tract mucosa [20, 21]. By contrast, increases in new-onset infection are associated with immune dysfunction, multiple comorbidities, and immunosuppression in patients with ESRD [22‒24].

The incidence of cardiogenic shock in patients with ESRD has increased in recent years. The overall mortality rate for this condition has improved slightly but remains high without significant change in the trend [25]. Patients with ESRD have a higher risk of experiencing cardiogenic shock in acute myocardial infarctions and heart surgeries than patients without ESRD do [26‒29]. Our study demonstrated that patients with ESRD with ECMO support also experience a higher risk of cardiogenic shock and myocardial infarction, a finding that may be explained by several factors. First, patients with ESRD often have multiple comorbidities, such as diabetes mellitus, hypertension, heart failure, and coronary artery diseases, all of which are risk factors for myocardial infarction with cardiogenic shock [30]. In our study, at least 50% of the patients had diabetes mellitus, and more than 80% had hypertension and/or coronary artery disease, contributing to an increased risk of cardiogenic shock requiring ECMO support. Second, hyperphosphatemia is common and associated with vascular calcification in patients with ESRD [31]. Hyperphosphatemia has multiple adverse effects on vascular smooth muscle cells and causes dysregulation of fibroblast growth factor-23 (FGF-23) and Klotho crucial to the development of vascular calcification and cardiovascular disease [32]. In this study, hyperphosphatemia was observed in both the survival and nonsurvival groups, and it may have played a role in the similarly high mortality rates between the groups. Third, uremic toxins and oxidative stress in patients with ESRD also contribute to vascular calcification. Uremic toxins such as p-cresyl sulfate and indoxyl sulfate are associated with the transition and apoptosis of vascular smooth muscle cells and the dysfunction of endothelial cells. Moreover, oxidative stress and excess production of reactive oxygen species mediate osteogenic transition and calcification of vascular smooth muscle cells [32]. Fourth, cardiogenic shock triggers the release of many inflammatory cytokines (interleukin-1β, interleukin-6, interleukin-8, tumor necrosis factor-α, etc.), and baseline levels of proinflammatory cytokines are predictive of cardiogenic shock development and subsequent mortality [33]. Additionally, ESRD induces a chronic proinflammatory state marked by increased levels of inflammatory cytokines. Therefore, patients with a combination of ESRD and cardiogenic shock may experience worse outcomes than those patients without these conditions.

The APACHE II score is one of the most common ICU scoring system used to evaluate disease severity and predict outcomes. It comprises several physiological parameters, age, and chronic health status for comprehensive evaluation, and various subtle changes of organ dysfunction can be reflected by the variations of physiological parameters [34]. Patients with ESRD usually have comorbidities that increase the risk and possibility of multiorgan failure. Studies have revealed that the APACHE II score is significantly associated with overall mortality in patients with ECMO who require continuous renal replacement therapy or have AKI, ST-segment elevation myocardial infarction, or postcardiotomy shock [4, 35‒37]. In this study, the APACHE II score exhibited the best discriminatory power and Youden index among the 4 scoring systems, indicating it is a practical tool that can be used to predict the outcomes of patients with ESRD on ECMO.

Although our predictions were satisfactorily accurate, this study has some limitations. First, this study was retrospective and conducted at a single tertiary-care medical center, and the sample size was relatively small; therefore, the statistical power and the generalizability of the findings may be limited. Second, some collected laboratory data, such as serum creatinine and electrolytes, were influenced by hemodialysis, with levels differing depending on whether patients underwent hemodialysis treatment on day 1 of ECMO support. The mode and timing of hemodialysis were determined on the basis of the patients’ clinical conditions; to minimize the effects of hemodialysis on our results, we chose the worst laboratory values of the first day of ECMO support for the calculation of ICU scoring system. Third, some laboratory data could not be obtained for all patients because of the retrospective study design. Finally, although the underlying comorbidities and dialysis modes did not differ substantially between survivors and nonsurvivors, the heterogeneity and diversity of the study population may limit the generalizability of our findings to other patient populations. Therefore, additional large-scale studies are required to conduct further validation and comparison.

In conclusion, this study revealed that patients with ESRD have high overall in-hospital mortality while on ECMO support, with mortality rates similar to those of patients with CKD or AKI, but substantially higher than those of patients without AKI. Additionally, the APACHE II score on day 1 of ECMO support is a strong predictor of in-hospital mortality for patients with ESRD on ECMO and is superior to various other ICU scoring systems in predicting mortality. The APACHE II score is a practical tool that can be used to predict outcomes in patients with ESRD on ECMO.

Tsung-Yu Tsai was supported by the Chang Gung Medical Research Program CMRPG5P0011. Pei-Chun Fan was supported by the Chang Gung Medical Research Program CMRPG5M0111.

This study protocol was reviewed and approved by the Institutional Review Board of Chang Gung Memorial Hospital (Approval No. 202401152B0). The requirement for informed consent was waived by the Institutional Review Board of Chang Gung Memorial Hospital because the study retrospectively analyzed patient data.

The authors have no conflicts of interest to declare.

This study was not supported by any sponsor or funder.

Tsung-Yu Tsai: conceptualization, data collection, statistical analysis, and manuscript drafting; Pei-Chun Fan: data collection and manuscript editing; Cheng-Chia Lee: data interpretation and curation; Shao-Wei Chen: data curation and methodology; Jia-Jin Chen: statistical analysis and data interpretation; Ming-Jen Chan: statistical analysis, methodology, and visualization; Ji-Tseng Fang and Yung-Chang Chen: visualization and manuscript reviewing; and Chih-Hsiang Chang: supervision, conceptualization, and manuscript reviewing.

The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author C.H.C. (E-mail: [email protected]) upon reasonable request.

1.
Askenazi
DJ
,
Selewski
DT
,
Paden
ML
,
Cooper
DS
,
Bridges
BC
,
Zappitelli
M
, et al
.
Renal replacement therapy in critically ill patients receiving extracorporeal membrane oxygenation
.
Clin J Am Soc Nephrol
.
2012
;
7
(
8
):
1328
36
.
2.
Tsai
TY
,
Chien
H
,
Tsai
FC
,
Pan
HC
,
Yang
HY
,
Lee
SY
, et al
.
Comparison of RIFLE, AKIN, and KDIGO classifications for assessing prognosis of patients on extracorporeal membrane oxygenation
.
J Formos Med Assoc
.
2017
;
116
(
11
):
844
51
.
3.
Thongprayoon
C
,
Cheungpasitporn
W
,
Lertjitbanjong
P
,
Aeddula
NR
,
Bathini
T
,
Watthanasuntorn
K
, et al
.
Incidence and impact of acute kidney injury in patients receiving extracorporeal membrane oxygenation: a meta-analysis
.
J Clin Med
.
2019
;
8
(
7
):
981
.
4.
Hsu
CK
,
Wu
IW
,
Chen
YT
,
Tsai
TY
,
Tsai
FC
,
Fang
JT
, et al
.
Acute kidney disease stage predicts outcome of patients on extracorporeal membrane oxygenation support
.
PLoS One
.
2020
;
15
(
4
):
e0231505
.
5.
Chen
SW
,
Lu
YA
,
Lee
CC
,
Chou
AH
,
Wu
VCC
,
Chang
SW
, et al
.
Long-term outcomes after extracorporeal membrane oxygenation in patients with dialysis-requiring acute kidney injury: a cohort study
.
PLoS One
.
2019
;
14
(
3
):
e0212352
.
6.
Lu
YA
,
Chen
SW
,
Lee
CC
,
Wu
VCC
,
Fan
PC
,
Kuo
G
, et al
.
Mid-term survival of patients with chronic kidney disease after extracorporeal membrane oxygenation
.
Interact Cardiovasc Thorac Surg
.
2020
;
31
(
5
):
595
602
.
7.
Kumar
VA
,
Ledezma
ML
,
Idroos
ML
,
Burchette
RJ
,
Rasgon
SA
.
Hospitalization rates in daily home hemodialysis versus peritoneal dialysis patients in the United States
.
Am J Kidney Dis
.
2008
;
52
(
4
):
737
44
.
8.
Lee
CC
,
Hsu
CC
,
Lin
MH
,
Chen
KH
,
Wu
IW
.
Hospitalization in patients with dialysis in Taiwan: a nationwide population-based observational study
.
J Formos Med Assoc
.
2022
;
121
(
Suppl 1
):
S39
46
.
9.
Palamuthusingam
D
,
Nadarajah
A
,
Pascoe
EM
,
Craig
J
,
Johnson
DW
,
Hawley
CM
, et al
.
Postoperative mortality in patients on chronic dialysis following elective surgery: a systematic review and meta-analysis
.
PLoS One
.
2020
;
15
(
6
):
e0234402
.
10.
Palamuthusingam
D
,
Nadarajah
A
,
Johnson
DW
,
Pascoe
EM
,
Hawley
CM
,
Fahim
M
.
Morbidity after elective surgery in patients on chronic dialysis: a systematic review and meta-analysis
.
BMC Nephrol
.
2021
;
22
(
1
):
97
.
11.
Wu
MY
,
Lin
PJ
,
Lee
MY
,
Tsai
FC
,
Chu
JJ
,
Chang
YS
, et al
.
Using extracorporeal life support to resuscitate adult postcardiotomy cardiogenic shock: treatment strategies and predictors of short-term and midterm survival
.
Resuscitation
.
2010
;
81
(
9
):
1111
6
.
12.
Tseng
YH
,
Wu
TI
,
Liu
YC
,
Lin
PJ
,
Wu
MY
.
Venoarterial extracorporeal life support in post-traumatic shock and cardiac arrest: lessons learned
.
Scand J Trauma Resusc Emerg Med
.
2014
;
22
:
12
.
13.
Alba
AC
,
Foroutan
F
,
Buchan
TA
,
Alvarez
J
,
Kinsella
A
,
Clark
K
, et al
.
Mortality in patients with cardiogenic shock supported with VA ECMO: a systematic review and meta-analysis evaluating the impact of etiology on 29,289 patients
.
J Heart Lung Transpl
.
2021
;
40
(
4
):
260
8
.
14.
Lee
SW
,
Yu
MY
,
Lee
H
,
Ahn
SY
,
Kim
S
,
Chin
HJ
, et al
.
Risk factors for acute kidney injury and in-hospital mortality in patients receiving extracorporeal membrane oxygenation
.
PLoS One
.
2015
;
10
(
10
):
e0140674
.
15.
Antonucci
E
,
Lamanna
I
,
Fagnoul
D
,
Vincent
JL
,
De Backer
D
,
Silvio Taccone
F
.
The impact of renal failure and renal replacement therapy on outcome during extracorporeal membrane oxygenation therapy
.
Artif Organs
.
2016
;
40
(
8
):
746
54
.
16.
Sakhuja
A
,
McCarthy
P
,
Hayanga
JA
,
Turley
S
,
Smith
G
,
Kellum
JA
.
Acute kidney injury in extracorporeal membrane oxygenation patients: national analysis of impact of age
.
Blood Purif
.
2022
;
51
(
7
):
567
76
.
17.
Ding
X
,
Xie
H
,
Yang
F
,
Wang
L
,
Hou
X
.
Risk factors of acute renal injury and in-hospital mortality in adult patients with postcardiotomy cardiogenic shock requiring veno-arterial extracorporeal membrane oxygenation: utility of MELD-XI score
.
Perfusion
.
2022
;
37
(
5
):
505
14
.
18.
Chan
MJ
,
Chen
SW
,
Fan
PC
,
Lee
CC
,
Chen
JJ
,
Kuo
G
, et al
.
Long-term clinical outcomes of acute kidney disease in patients receiving extracorporeal membrane oxygenation
.
Cardiorenal Med
.
2024
;
14
(
1
):
294
306
.
19.
Lo Coco
V
,
Lorusso
R
,
Raffa
GM
,
Malvindi
PG
,
Pilato
M
,
Martucci
G
, et al
.
Clinical complications during veno-arterial extracorporeal membrane oxigenation in post-cardiotomy and non post-cardiotomy shock: still the achille's heel
.
J Thorac Dis
.
2018
;
10
(
12
):
6993
7004
.
20.
Wasse
H
,
Gillen
DL
,
Ball
AM
,
Kestenbaum
BR
,
Seliger
SL
,
Sherrard
D
, et al
.
Risk factors for upper gastrointestinal bleeding among end-stage renal disease patients
.
Kidney Int
.
2003
;
64
(
4
):
1455
61
.
21.
Kalman
RS
,
Pedrosa
MC
.
Evidence-based review of gastrointestinal bleeding in the chronic kidney disease patient
.
Semin Dial
.
2015
;
28
(
1
):
68
74
.
22.
Cohen
G
,
Haag-Weber
M
,
Hörl
WH
.
Immune dysfunction in uremia
.
Kidney Int Suppl
.
1997
;
62
:
S79
82
.
23.
Girndt
M
,
Sester
U
,
Sester
M
,
Kaul
H
,
Köhler
H
.
Impaired cellular immune function in patients with end-stage renal failure
.
Nephrol Dial Transpl
.
1999
;
14
(
12
):
2807
10
.
24.
Chavers
BM
,
Solid
CA
,
Gilbertson
DT
,
Collins
AJ
.
Infection-related hospitalization rates in pediatric versus adult patients with end-stage renal disease in the United States
.
J Am Soc Nephrol
.
2007
;
18
(
3
):
952
9
.
25.
Dalia
T
,
Pothuru
S
,
Chan
WC
,
Mehta
H
,
Goyal
A
,
Farhoud
H
, et al
.
Trends and outcomes of cardiogenic shock in patients with end-stage renal disease: insights from USRDS database
.
Circ Heart Fail
.
2023
;
16
(
8
):
e010462
.
26.
Beattie
JN
,
Soman
SS
,
Sandberg
KR
,
Yee
J
,
Borzak
S
,
Garg
M
, et al
.
Determinants of mortality after myocardial infarction in patients with advanced renal dysfunction
.
Am J Kidney Dis
.
2001
;
37
(
6
):
1191
200
.
27.
Vallabhajosyula
S
,
Ya'Qoub
L
,
Kumar
V
,
Verghese
D
,
Subramaniam
AV
,
Patlolla
SH
, et al
.
Contemporary national outcomes of acute myocardial infarction-cardiogenic shock in patients with prior chronic kidney disease and end-stage renal disease
.
J Clin Med
.
2020
;
9
(
11
):
3702
.
28.
Guo
M
,
St Pierre
E
,
Clemence
J
Jr
,
Wu
X
,
Tang
P
,
Romano
M
, et al
.
Impact of chronic renal failure on surgical outcomes in patients with infective endocarditis
.
Ann Thorac Surg
.
2021
;
111
(
3
):
828
35
.
29.
Sandhyavenu
H
,
Ullah
W
,
Badu
I
,
Patel
BA
,
Polam
AR
,
Taha
A
, et al
.
Trend, predictors, and outcomes of transcatheter aortic valve implantation in patients with end-stage renal disease: insights from the nationwide readmissions database
.
Cardiovasc Revasc Med
.
2023
;
50
:
1
7
.
30.
Reynolds
HR
,
Hochman
JS
.
Cardiogenic shock: current concepts and improving outcomes
.
Circulation
.
2008
;
117
(
5
):
686
97
.
31.
Adeney
KL
,
Siscovick
DS
,
Ix
JH
,
Seliger
SL
,
Shlipak
MG
,
Jenny
NS
, et al
.
Association of serum phosphate with vascular and valvular calcification in moderate CKD
.
J Am Soc Nephrol
.
2009
;
20
(
2
):
381
7
.
32.
Ding
N
,
Lv
Y
,
Su
H
,
Wang
Z
,
Kong
X
,
Zhen
J
, et al
.
Vascular calcification in CKD: new insights into its mechanisms
.
J Cel Physiol
.
2023
;
238
(
6
):
1160
82
.
33.
Shpektor
A
.
Cardiogenic shock: the role of inflammation
.
Acute Card Care
.
2010
;
12
(
4
):
115
8
.
34.
Prondzinsky
R
,
Lemm
H
,
Swyter
M
,
Wegener
N
,
Unverzagt
S
,
Carter
JM
, et al
.
Intra-aortic balloon counterpulsation in patients with acute myocardial infarction complicated by cardiogenic shock: the prospective, randomized IABP SHOCK Trial for attenuation of multiorgan dysfunction syndrome
.
Crit Care Med
.
2010
;
38
(
1
):
152
60
.
35.
Chung
SY
,
Tong
MS
,
Sheu
JJ
,
Lee
FY
,
Sung
PH
,
Chen
CJ
, et al
.
Short-term and long-term prognostic outcomes of patients with ST-segment elevation myocardial infarction complicated by profound cardiogenic shock undergoing early extracorporeal membrane oxygenator-assisted primary percutaneous coronary intervention
.
Int J Cardiol
.
2016
;
223
(
223
):
412
7
.
36.
Erdoğan
SB
,
Bastopcu
M
,
Usca
MK
,
Çakmak
AY
,
Sargın
M
,
Aka
S
.
The utility of risk scores in postcardiotomy extracorporeal membrane oxygenation
.
Perfusion
.
2024
;
39
(
3
):
578
84
.
37.
Tsai
TY
,
Tsai
FC
,
Chang
CH
,
Jenq
CC
,
Hsu
HH
,
Chang
MY
, et al
.
Prognosis of patients on extracorporeal membrane oxygenation plus continuous arteriovenous hemofiltration
.
Chang Gung Med J
.
2011
;
34
(
6
):
636
43
.