Pediatric extracorporeal membrane oxygenation support at a single center: Outcomes by indication, prognostic factors, and duration Authors: L. Lily Rosenthal, Andreas W. Flemmer, Kai Martin Förster, Mathias Bruegel, Jürgen Hörer, Nikolaus A. Haas, Christian Hagl, Sebastian Sadoni, Oliver J Muensterer, Sebastian Georg Michel Published: 2026 ABSTRACT Background:Pediatric extracorporeal membrane oxygenation (pECMO) provides life-saving cardiopulmonary support for children with severe cardiac or respiratory failure. Identifying prognostic factors may improve risk stratification and clinical decision-making.Methods:We retrospectively reviewed 149 patients (<18 years) who received pECMO at a single center between 2003 and 2025. Patients with multiple extracorporeal membrane oxygenation (ECMO) runs during the same hospitalization were excluded. Demographic, clinical, laboratory, and outcome data were collected. The primary endpoint was 30-day survival.Results:Among 149 patients, the median age was 245 days, and the median weight was 6.3 kg. Indications for ECMO were acute cardiac failure (56%), respiratory failure (37%), and extracorporeal cardiopulmonary resuscitation (ECPR; 11%). Kaplan–Meier analysis demonstrated significant survival differences among groups (log-rank P = 0.010). Mean survival times were 24.07 days (95% confidence interval [CI] 20.72–27.41) for cardiac failure, 28.99 days (95% CI 24.64–33.35) for respiratory failure, and 14.13 days (95% CI 7.72–20.55) for ECPR. In multivariable Cox regression analysis, pre-ECMO lactate (hazard ratio [HR] 1.042, 95% CI 1.008–1.078; P = 0.017) and post-ECMO lactate (HR 1.067, 95% CI 1.024–1.110; P = 0.002) were independent predictors of mortality. Pre- and post-ECMO aspartate aminotransferase and pH were not significantly associated with mortality.Conclusions:Elevated lactate levels before and after ECMO initiation independently predicted mortality in pediatric patients receiving ECMO. Lactate may serve as a useful biomarker for prognostication and risk stratification in this population. INTRODUCTION Extracorporeal membrane oxygenation (ECMO) has become an essential life-support modality for children with severe cardiac or respiratory failure refractory to conventional therapy. Since its introduction, ECMO technology and clinical protocols have evolved significantly, resulting in improved survival rates, yet substantial morbidity persists. Children undergoing ECMO are particularly vulnerable due to immature organ systems, small body size, and complex underlying pathologies, which can lead to complications including acute kidney injury (AKI), neurological injury, and hepatic dysfunction.[1234] Improvements in circuit maintenance, nutritional support, and clinical decision-making have reduced many complications once considered insurmountable.[5] Cardiogenic shock, particularly post-cardiotomy, presents a high-risk scenario where survival is closely tied to pre-ECMO organ dysfunction, timing of initiation, and adequacy of perfusion. In contrast, neonatal respiratory failure, including conditions such as meconium aspiration syndrome or pulmonary hypertension due to diaphragmatic hernia, may have more favorable outcomes if ECMO is initiated before irreversible hypoxemia or multi-organ damage occurs,[67891011] and neurological events such as seizures, stroke, and intracranial hemorrhage. Despite advances, ECMO remains associated with significant complications. Renal failure, defined according to Kidney Disease: Improving Global Outcomes (KDIGO) criteria, occurs in 25%–45% of patients and frequently necessitates renal replacement therapy (RRT). Neurological injury, including seizures, intracranial hemorrhage, and ischemic stroke, affects up to one-third of pediatric patients. Hepatic dysfunction, characterized by elevated transaminases or coagulopathy, reflects systemic hypoperfusion, inflammation, or ischemia-reperfusion injury. Understanding these complications in the context of specific indications is critical for improving outcomes.[1213141516] Förster et al. previously reported our clinical results in a subgroup of patients with meconium aspiration syndrome. In the present study, we provide an updated analysis of these data.[17] Single-center analyses provide valuable insight by allowing standardized assessment of clinical protocols, cannulation strategies, anticoagulation practices, and supportive care. Large heterogeneous cohorts also enable identification of prognostic factors such as pre-ECMO lactate, acid-base status, and organ function markers that can guide patient selection and timing of ECMO initiation. The findings aim to inform best practices, refine risk stratification, and improve care strategies for critically ill children requiring ECMO support. METHODS Ethical statement The study was approved by the Institutional Review Board of Ludwig Maximilian University (LMU) University Hospital Munich (approval number 25-0892) and conducted in accordance with the Declaration of Helsinki. All patient data were anonymized to ensure confidentiality. Study design This retrospective observational study included pediatric patients (<18 years of age) who received ECMO support at LMU University Hospital Munich between January 2003 and July 2025. Patients were identified through the institutional ECMO registry and electronic medical records. A total of 178 patients were screened for eligibility. Twenty-nine patients were excluded because they (1) underwent complex tracheal or esophageal repair with cardiopulmonary bypass without ECMO support, (2) required multiple ECMO runs during the same hospitalization, or (3) represented atypical cases previously reported separately, including one patient bridged to lung transplantation and one patient treated with a surgical Potts shunt in combination with ECMO. The final study cohort consisted of 149 patients. Patients were categorized according to the primary indication for ECMO support into three groups: (1) cardiac failure, including cardiogenic shock and dilated cardiomyopathy; (2) respiratory failure, including neonatal respiratory failure and perioperative support for congenital tracheal or esophageal malformations; and (3) cardiorespiratory failure, including acute respiratory distress syndrome (ARDS), severe pneumonia or airway infection, drowning, and patients requiring extracorporeal cardiopulmonary resuscitation (ECPR). These groups were analyzed to reflect the major clinical indications for pediatric ECMO at our institution. The primary outcome was 30-day survival following ECMO initiation. Secondary outcomes included renal failure according to KDIGO criteria, hepatic dysfunction defined by elevated serum transaminases (aspartate aminotransferase [AST] and alanine aminotransferase), and neurological complications including seizures, intracranial hemorrhage, and ischemic stroke occurring within 30 days of ECMO initiation. Demographic characteristics, underlying diagnoses, ECMO modality and duration, laboratory parameters, complications, and clinical outcomes were collected for analysis. Cannulation and extracorporeal membrane oxygenation management The choice of ECMO modality (veno-arterial [V-A] or veno-venous [V-V]) was based on the underlying pathology, degree of cardiorespiratory compromise, and anticipated duration of support. V-A ECMO was primarily used for patients with cardiac failure or combined cardiorespiratory failure, whereas V-V ECMO was preferred for isolated respiratory failure. Cannula size and configuration were selected according to patient age, weight, and institutional protocols. In neonates, peripheral V-A ECMO was typically established through cannulation of the right carotid artery and internal jugular vein. In older children, femoral cannulation with distal limb perfusion was the preferred approach when anatomically feasible. V-V ECMO was most commonly established using a dual-lumen Avalon® cannula inserted through the right internal jugular vein. Central ECMO support was established through median sternotomy with cannulation of the ascending aorta and right atrium. Cannula position was routinely confirmed by echocardiography and chest radiography. ECMO initiation was considered in patients with severe cardiac and/or respiratory failure refractory to maximal conventional medical management, including vasoactive support, mechanical ventilation, and adjunctive therapies, following multidisciplinary team assessment. Anticoagulation Systemic anticoagulation was achieved using unfractionated heparin according to institutional protocols. Anticoagulation monitoring included activated partial thromboplastin time, activated clotting time, anti-factor Xa activity, platelet count, and routine coagulation parameters. Target ranges were adjusted according to patient age, ECMO modality, and individual bleeding or thrombotic risk. Extracorporeal membrane oxygenation explantation Weaning from ECMO was performed following evidence of adequate cardiac and/or pulmonary recovery, as determined by clinical, hemodynamic, laboratory, and echocardiographic assessment. Flow reduction trials were conducted before decannulation whenever clinically appropriate. For V-V ECMO, percutaneous cannula removal was followed by manual compression to achieve hemostasis. For V-A ECMO, arterial decannulation involved removal of the arterial cannula with subsequent vascular reconstruction using direct suturing or patch repair when required. Venous cannulas were removed with direct vessel closure. Central ECMO explantation was performed surgically through median sternotomy. Statistical analysis Continuous variables are presented as mean ± standard deviation or median (interquartile range [IQR]), as appropriate. Categorical variables are reported as frequencies and percentages. Comparisons between groups were performed using the Chi-square test or Fisher’s exact test for categorical variables and the Student’s t-test or Mann–Whitney U test for continuous variables, as appropriate. Survival was evaluated using Kaplan–Meier analysis from the time of ECMO initiation, and differences between groups were assessed using the log-rank test. Univariable and multivariable Cox proportional hazards regression analyses were performed to identify predictors of mortality. Hazard ratios (HRs) with 95% confidence intervals (CIs) were reported. Variables considered clinically relevant or associated with mortality in univariable analyses were entered into multivariable models. The multivariable model included clinically relevant pre-ECMO and early post-ECMO laboratory variables, specifically serum lactate, pH, and AST. These variables were selected based on biological plausibility and prior literature linking tissue hypoperfusion and organ dysfunction to mortality in ECMO patients. A two-sided P < 0.05 was considered statistically significant. Statistical analyses were performed using IBM SPSS Statistics version 30 (IBM Corp., Armonk, NY, USA). RESULTS Baseline characteristics of the 149 included pediatric patients are summarized in Table 1a. Figure 1 presents our clinical strategy for ECMO support in pediatric patients, from admission to long-term follow-up. Figure 2 shows the highly variable distribution of ECMO support and mortality after ECMO support, which was not clearly reflected by laboratory and blood gas analyses. Figure 3 shows 30-day survival according to indication groups. The highest survival was observed in the respiratory failure group. Patients with respiratory failure were the youngest (median age approximately 0.07 months) and smallest (median weight 3.5 kg), reflecting a predominantly neonatal population (e.g. meconium aspiration syndrome and congenital diaphragmatic hernia). Patients with cardiac failure were older and heavier (median age 26.1 months; median weight 9.3 kg), while those with cardiorespiratory failure, including ECPR cases, showed intermediate values (median age 23.2 months; median weight 10.0 kg). Body mass index and body surface area followed a similar distribution across groups. Sex distribution was comparable between groups. Timeline of extracorporeal membrane oxygenation (ECMO) Support: From pre-ECMO decision to long-term follow-up. ECMO: Extracorporeal membrane oxygenation Distribution of extracorporeal membrane oxygenation support (n) and mortality rates. ECMO: Extracorporeal membrane oxygenation Kaplan–Meier survival curves 30-day after extracorporeal membrane oxygenation initiation stratified by indication group. ECMO: Extracorporeal membrane oxygenation Table 1aDemographic data of patients at the time of extracorporeal membrane oxygenation implantation (n=149)Indication and diagnosis Median (IQR)Age (months)Weight (kg)BMI (kg/m2)BSAGender (male)Cardiac failure (n=83)26.067 (2.017–120.933)9.3 (3.5–25.50)14.542 (12.480–16.959)0.219 (0.053–0.92)43/83Respiratory failure (n=50)0.067 (0.033–2.633)3.5 (3.0–5.05)13.3 (12.31–17.57)0.049 (0.035–0.091)30/50Cardiorespiratory failure with and without ECPR (n=16)*23.20 (13.0–73.60)10.0 (7.0–20.0)15.60 (10.94–18.77)0.222 (0.126–0.556)7/16Values are presented as median (IQR) or n (%). *The number of patients in this subgroup was too small to allow meaningful statistical comparison between patients with and without ECPR. ECPR: Extracorporeal cardiopulmonary resuscitation, BMI: Body mass index, ECPR: Extracorporeal cardiopulmonary resuscitation, ECMO: Extracorporeal membrane oxygenation, IQR: Interquartile range, BSA: Body surface area Demographic data of patients at the time of extracorporeal membrane oxygenation implantation (n=149) Values are presented as median (IQR) or n (%). *The number of patients in this subgroup was too small to allow meaningful statistical comparison between patients with and without ECPR. ECPR: Extracorporeal cardiopulmonary resuscitation, BMI: Body mass index, ECPR: Extracorporeal cardiopulmonary resuscitation, ECMO: Extracorporeal membrane oxygenation, IQR: Interquartile range, BSA: Body surface area ECMO modality and duration according to indication are shown in Table 1b. In the cardiac group, both central and peripheral veno-arterial ECMO were used, with a median duration of 4.0 days (IQR 2.0–12.0). In the respiratory group, veno-venous ECMO was more frequently used, and median ECMO duration was 4.0 days (IQR 2.75–6.5). In the cardiorespiratory failure group, including ECPR patients, the median ECMO duration was 4.5 days (IQR 1.25–7.75). Overall, respiratory failure patients required the shortest duration of ECMO support. Figure 2 Kaplan–Meier survival curves after ECMO initiation stratified by indication group. Patients were grouped according to cardiac failure (green), respiratory failure (blue), and ECPR (red). Survival differed significantly between groups (log-rank test, P < 0.001). ECMO duration showed the shortest ECMO support in the respiratory failure group. Table 1bTypes and duration of extracorporeal membrane oxygenation support by diagnosisIndication and diagnosis (n=14)ECMO typeMedian duration (IQR) (days)Cardiac failure (n=83)Central ECMO (n=49) V-A-ECMO (n=34)4.0 (2.0–12.0)Respiratory failure (n=50)Central ECMO (n=2) V-A-ECMO (n=40) V-V-ECMO (n=8)4.0 (2.75–6.50)Cardiorespiratory failure with and without ECPR (n=16)Central ECMO (n=1) V-A-ECMO (n=12) V-V-ECMO (n=3)4.50 (1.25–7.75)V-V ECMO: Veno-venous extracorporeal membrane oxygenation, V-A ECMO: Veno-arterial extracorporeal membrane oxygenation, ECMO: Extracorporeal membrane oxygenation, ECPR: Extracorporeal cardiopulmonary resuscitation Types and duration of extracorporeal membrane oxygenation support by diagnosis V-V ECMO: Veno-venous extracorporeal membrane oxygenation, V-A ECMO: Veno-arterial extracorporeal membrane oxygenation, ECMO: Extracorporeal membrane oxygenation, ECPR: Extracorporeal cardiopulmonary resuscitation As shown in Table 2, adjunct pharmacological therapies during ECMO weaning and explantation: Adjunct pharmacological therapies varied according to underlying indication and were used to support organ recovery during ECMO weaning and preparation for explantation. In patients with cardiogenic shock, inotropic and vasoactive agents such as dobutamine, milrinone, levosimendan, noradrenaline, and vasodilators were used to optimize cardiac output, maintain systemic perfusion, and reduce afterload. Diuretics and antiarrhythmic agents were administered as clinically indicated. In neonatal respiratory failure (including meconium aspiration syndrome and diaphragmatic hernia), pulmonary vasodilators such as inhaled nitric oxide and sildenafil, in addition to milrinone and corticosteroids, were used to improve oxygenation and reduce pulmonary hypertension. In patients with tracheal or esophageal malformations, therapy consisted primarily of supportive care, antibiotics when infection was present, and pulmonary vasodilators when required. In dilated cardiomyopathy, pharmacological support included inotropes, afterload-reducing agents (angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers), beta-blockers after stabilization, diuretics, and anticoagulation as a bridge to ventricular assist device implantation. In ARDS, and other respiratory failure states, sedatives, neuromuscular blockade, corticosteroids, inhaled nitric oxide, diuretics, and antibiotics were used to optimize oxygenation and treat underlying infection or inflammation. In drowning-associated cardiorespiratory failure, supportive therapy included corticosteroids, antibiotics when indicated, and temperature management. Controlled rewarming was performed when hypothermia was present to avoid hemodynamic instability and support safe ECMO weaning. Table 2Adjunct pharmacological medications for each patient group for extracorporeal membrane oxygenation weaning and explantationPatient groupAdditional medicationsEffectCardiogenic shockDobutamine, milrinone, levosimendan, noradrenaline, diuretics, vasodilators (e.g., nitroglycerin), antiarrhythmicsSupport cardiac output, maintain blood pressure, reduce preload and afterload, control arrhythmiasNewborns with meconium aspirationSurfactant, iNO, sildenafil, milrinone, steroidsImprove lung compliance, reduce pulmonary hypertension, decrease inflammationNewborns with diaphragmatic herniaiNO, sildenafil, milrinone, steroidsPulmonary vasodilation, enhance cardiac and lung functionTracheal or esophageal malformationSupportive care, antibiotics, inhaled nitric oxide (if respiratory failure present), steroidsManage infection, reduce pulmonary hypertension, and inflammationDCM bridge to decisionDobutamine, milrinone, levosimendan, ACE inhibitors/ARBs, beta-blockers (post stabilization), diuretics, anticoagulants - to VAD implantationSupport heart function, reduce afterload, manage fluid status, prevent thrombosisARDSSedatives, muscle relaxants, steroids, inhaled nitric oxide, diuretics, antibioticsReduce inflammation, improve oxygenation, manage fluid balance, treat infectionsRespiratory failure to decisionSimilar to ARDS: Sedatives, Steroids, Inhaled Nitric Oxide, DiureticsSupport respiratory function and reduce lung injuryDrowning injury*Steroids, Antibiotics, Supportive care (including inhaled nitric oxide if indicated)Reduce lung inflammation, prevent secondary infections*The number of patients in this subgroup was too small to allow meaningful statistical comparison between patients with and without ECPR. In patients with drowning-related injury, hypothermia is frequently present due to cold water immersion. Controlled and gradual rewarming was performed during ECMO support to restore normothermia and avoid hemodynamic instability associated with rapid temperature correction. DCM: Dilated cardiomyopathy, ARDS: acute respiratory distress syndrome, iNO: Inhaled nitric oxide, ARB: Angiotensin receptor blockers, ACE: Angiotensin-converting enzyme, VAD: Ventricular assist device Adjunct pharmacological medications for each patient group for extracorporeal membrane oxygenation weaning and explantation *The number of patients in this subgroup was too small to allow meaningful statistical comparison between patients with and without ECPR. In patients with drowning-related injury, hypothermia is frequently present due to cold water immersion. Controlled and gradual rewarming was performed during ECMO support to restore normothermia and avoid hemodynamic instability associated with rapid temperature correction. DCM: Dilated cardiomyopathy, ARDS: acute respiratory distress syndrome, iNO: Inhaled nitric oxide, ARB: Angiotensin receptor blockers, ACE: Angiotensin-converting enzyme, VAD: Ventricular assist device Organ function at extracorporeal membrane oxygenation implantation, 24 h post-initiation, and pre-explantation Organ function parameters stratified by ECMO indication are summarized in Table 3. Across all groups, severe metabolic acidosis and elevated lactate levels were observed at ECMO initiation, with partial improvement during support. Liver and renal function markers showed variable changes over time, with higher transaminase and creatinine levels in the cardiorespiratory group compared with cardiac and respiratory groups. Inflammatory markers and cardiac biomarkers remained elevated throughout ECMO support, particularly in patients with prolonged or complicated courses [Table 4]. Table 3Organ function at extracorporeal membrane oxygenation implantation and explantation (analysis of variance table)Cardiac failure(83; 56%), mean±SDRespiratory failure(50; 34%) mean±SDCardiorespiratory failure (16; 11%) mean±SDBGA pH value7.268 ± 0.1207.255±0.1336.98±0.80 Base excess (mmol/L)−3.33±9.58−2.594±11.720−7.164±12.030 Lactate (mmol/L)6.81±7.486.056±5.499.80±6.71Liver function tests ALT (U/L)121±30398±409214±381 AST364±690201±626425±627Renal function test Creatinine (mg/dL)0.92±0.970.8±0.41.1±0.65 Urea (mg/dL)54.3±59.536±21.548±29Troponin T(ng/ml)18.7±4422±2513±36NT-proBNP15,857±21,72221,072±24,50017,939±29,682CRP4.3±63.3±3.57.3±10 24h post ECMO BGA* pH value0.3±0.45.45±0.77.4±4.1 Base excess (mmol/L)5.5±2.06.8±4.41.2±4-6 Lactate (mmol/L)5.0±2.55±2.43.9±4.9Liver function tests ALT (U/L)164±89189±238200±137 AST73±6585±49159±221Renal function test Creatinine (mg/dL)1.6±0.80.8±0.40.9±1.0 Urea (mg/dL)65±6748±5258±67Troponin T(ng/ml)0.8±1.50.07±0.50.9±0.8 Pre-explant BGA* PH value7.2 ±1.697.3±0.57.32±0.13 Base excess (mmol/L)−1.47±8.243.0±6.2−3.86±6 Lactate (mmol/L)4.6±2.32.3±2.26.7±8Liver function tests ALT (U/L)182±469252±11996±119 AST478±1035253±1048268±338Renal function test Creatinine (mg/dL)1±0.91.6±4.70.9±0.8 Urea (mg/dL)58±4157±4454±38Troponin T(ng/ml)10.3±272±51±1.4NT-proBNP19,649±27,3237000±5.5002800±4000CRP5.3±7.54±713±15*The number of patients in this subgroup was too small to allow meaningful statistical comparison between patients with and without ECPR. Measurements: Data are presented as mean ± SD. Laboratory parameters were collected at ECMO initiation, 24 h after initiation, and before explantation. BGA: arterial blood gas analysis, ALT: Alanine aminotransferase, AST: Aspartate aminotransferase, CRP: C-reactive protein, NT-proBNP: N-terminal pro-B-type natriuretic peptide, SD: Standard deviation, ECMO: Extracorporeal membrane oxygenation Organ function at extracorporeal membrane oxygenation implantation and explantation (analysis of variance table) *The number of patients in this subgroup was too small to allow meaningful statistical comparison between patients with and without ECPR. Measurements: Data are presented as mean ± SD. Laboratory parameters were collected at ECMO initiation, 24 h after initiation, and before explantation. BGA: arterial blood gas analysis, ALT: Alanine aminotransferase, AST: Aspartate aminotransferase, CRP: C-reactive protein, NT-proBNP: N-terminal pro-B-type natriuretic peptide, SD: Standard deviation, ECMO: Extracorporeal membrane oxygenation Table 4Risk factor analysis for early mortality at 30-day after extracorporeal membrane oxygenation support using univariable Cox regressionUnivariable Cox analysisHR (95% CI) P Patient’s data pre-ECMO implantation Age (years)0.952 (0.838-1–0.082)0.454 Gender0.984 (0.940–1.030)0.487 Body weight (kg)1.008 (0.995–1.021)0.234 BMI (kg/cm2)0.919 (0.838–1.007)0.071Diagnoses for ECMO implantation0.003 Cardiogenic shock (57; 39%)2.345 (0.830–6.622)0.108 Newborns with meconium aspiration1.081 (0.270–4.322)0.913 Newborns with PAH - diaphragmatic hernia1.182 (0.437–7.124)0.823 Tracheal or esophageal malformation0.558 (0.102–3.048)0.501 DCM bridge to decision0.587 (0.158–2.188)0.428 ARDS3.425 (0.918–12.778)0.067 Drowning injury3.364 (0.947–11.950)0.061 Respiratory failure to decision1.449 (0.324–6.477)0.627 AST1.000 (1.000–1.001)0.043 ALT1.000 (1.000–1.001)0.078 PH value0.715 (0.456–1.120)0.143 Base excess0.964 (0.939–0.989)0.006 Lactate1.079 (1.043–1.116)<0.001 NT-pro-BNP1.000 (1.000–1.000)0.947 Troponin T1.003 (0.996–1.009)0.384HR: Hazard ratio, CI: Confidence interval, ECMO: Extracorporeal membrane oxygenation, BMI: Body mass index, PAH: Pulmonary arterial hypertension, DCM: Dilated cardiomyopathy, ARDS: Acute respiratory distress syndrome, AST: Aspartate aminotransferase, ALT: Alanine aminotransferase, NT-Pro-BNP: N-terminal pro-B-type natriuretic peptide Risk factor analysis for early mortality at 30-day after extracorporeal membrane oxygenation support using univariable Cox regression HR: Hazard ratio, CI: Confidence interval, ECMO: Extracorporeal membrane oxygenation, BMI: Body mass index, PAH: Pulmonary arterial hypertension, DCM: Dilated cardiomyopathy, ARDS: Acute respiratory distress syndrome, AST: Aspartate aminotransferase, ALT: Alanine aminotransferase, NT-Pro-BNP: N-terminal pro-B-type natriuretic peptide DISCUSSION In this two-decade single-center study of 149 pediatric patients receiving ECMO, survival varied significantly according to underlying diagnosis, with the highest outcomes observed in respiratory failure and elective surgical indications, and the lowest in cardiogenic shock and drowning-related injury. Importantly, pre- and post-ECMO lactate levels were independently associated with mortality, highlighting the prognostic value of tissue hypoperfusion markers in this population. Our overall survival rate of 58% is consistent with previously reported pediatric ECMO outcomes, which range from 40% to 60% in heterogeneous cohorts.[12] Zens et al. reported in their review that vessel reconstruction did not appear to significantly impact short-term morbidity.[18] Lactate has been repeatedly identified as a marker of severity and mortality in ECMO patients, reflecting global tissue hypoxia and inadequate perfusion.[12] In addition, Wheeler and Bullock reported improved survival in patients with ARDS treated with ECMO combined with continuous RRT.[19] Better outcomes in respiratory failure likely reflect the reversibility of pulmonary pathology and the possibility of initiating ECMO before irreversible multiorgan dysfunction occurs.[202122] The heterogeneity of ECMO indications in our cohort highlights a major challenge in pediatric ECMO research, as differences in underlying diagnosis and clinical severity limit direct comparisons and the development of universal prognostic models. Our cohort included both emergent cardiogenic shock and elective perioperative cases, reflecting the wide spectrum of disease acuity encountered in clinical practice. Similar to Raj et al., who emphasized the importance of patient-specific factors such as underlying diagnosis, pre-ECMO physiological status, and comorbidities, our findings support the need for individualized risk assessment in pediatric ECMO patients.[23] In our cohort, markers of illness severity, including pre-ECMO acidosis and prolonged ECMO duration, as well as neurological complications, were associated with worse outcomes, underscoring the importance of early recognition and tailored management strategies. Future research and clinical protocols should focus on indication-specific risk stratification to better optimize the timing, modality, and duration of ECMO support. AKI and the need for RRT remain common complications during pediatric ECMO and contribute significantly to morbidity and mortality. In our cohort, renal complications occurred in 35% of patients, all of whom required dialysis, which is consistent with reported incidences of 27%–46% in the literature.[12] AKI during ECMO may result from multiple interrelated mechanisms, including pre-existing shock, hypoperfusion, hemolysis, and inflammation associated with extracorporeal circulation. The ECMO-induced systemic inflammatory response may further contribute to renal endothelial injury, capillary leak, and tissue oedema.[242526] These complications reflect multifactorial mechanisms, including hypoperfusion, inflammation, hemolysis, and anticoagulation-related effects. Although combined ECMO and RRT has been suggested to improve fluid and metabolic control in selected patients, this approach was not routinely used in our cohort.[272829] Neurological injury remains a major complication of ECMO, with multifactorial pathogenesis including altered cerebral perfusion, embolic events, and anticoagulation-related hemorrhage.[30] Hepatic dysfunction is similarly associated with hypoperfusion, congestion, and systemic inflammation.[10202729] In our cohort, prolonged ECMO duration was associated with increased complication rates and mortality, consistent with prior studies suggesting that longer support reflects higher disease severity and increased exposure to circuit-related risks.[10212830] Our data indicate an association between prolonged ECMO duration and increased complication rates and mortality. This finding is consistent with previous reports suggesting that extended ECMO support reflects greater illness severity or delayed recovery and increases exposure to complications such as bleeding, thrombosis, and infection.[242530] Despite advances in ECMO technology and management, renal, hepatic, and neurological complications remain common. In our cohort, renal injury occurred in 35% of patients, hepatic dysfunction in 28%, and neurological complications in 35%. These findings are consistent with published data, in which pediatric ECMO-associated AKI has been reported in 27%–46% of cases.[12252627] Early initiation of ECMO, before the development of severe metabolic derangements such as acidosis and hyperlactatemia, appears to be associated with improved outcomes. In our cohort, pre-ECMO pH was identified as an additional predictor of mortality, supporting the importance of timely cannulation strategies. However, the optimal timing of ECMO initiation remains challenging and requires balanced clinical judgment, potentially supported by standardized protocols incorporating biomarkers and physiological thresholds. Anticoagulation strategies vary considerably between centers due to differences in patient populations and institutional protocols. Careful individualized monitoring is therefore essential to balance bleeding and thrombotic risks in pediatric ECMO patients.[122324252630] Limitations This study is limited by its retrospective single-center design, long study period with evolving ECMO practices, and heterogeneity of indications. In addition, the predictive value of laboratory markers requires prospective validation. CONCLUSION Survival in pediatric ECMO varies significantly by indication and severity of illness. Serum lactate is an independent predictor of mortality and may aid in early risk stratification. Despite advances in management, organ dysfunction remains common, underscoring the need for timely initiation and individualized ECMO strategies. Conflicts of interest There are no conflicts of interest.