Use of extracorporeal membrane oxygenation in neonates with congenital anomalies of the kidney and urinary tract: a multicenter survey of current practices across CHNC and Pedi-ECMO centers Authors: Syed Ahmed, Burhan Mahmood, John Daniel, Sandy Johng, Amy Strong, John Ibrahim, June Hu, Kara Short, Kelsey Montgomery, Keri Drake, Kushal Bhakta, Leslie Lusk, Mary Revenis, Pratibha Thakkar, Ruth Seabrook, Samir Pandya, Sarah Keene, Sharada Gowda, Suma Hoffman, Tahagod Mohamed, Tara Beck, Tasnim Najaf, Robert DiGeronimo, Natalie Rintoul, Abhishek Makkar Published: 2026 ABSTRACT AbstractBackgroundThis study aimed to characterize practices and decision-making for extracorporeal membrane oxygenation (ECMO) for congenital anomalies of the kidney and urinary tract (CAKUT).MethodsGeneral practices (GP) section inquired about institutional practices and barriers, ECMO criteria, and dialysis. The hypothetical cases (HC) illustrated four clinical scenarios with varying degrees of renal severity for ECMO candidacy.ResultsThen, 99 (42 centers) and 91 (38 centers) physicians completed the GP and HC components, respectively. The majority considered ECMO on a case-by-case basis (66%). Bilateral renal agenesis was the most common diagnosis for exclusion (52%). Prenatal markers used for ECMO exclusion included anhydramnios (43%) and lung volumes (43%). The majority of centers had nephrology involved in ECMO decision-making. Challenges for implementing ECMO included disease heterogeneity (79%) and poor evidence on outcomes (66%). HC responses demonstrated variability in considering ECMO for CAKUT.ConclusionsVariability among providers and institutes underscores the need for consensus-based guidelines to optimize decision-making and outcomes.Graphical abstract A higher resolution version of the Graphical abstract is available as Supplementary informationSupplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s00467-026-07301-x.Keywords: ECMO, Congenital anomalies of the kidney and urinary tract, CAKUT, Neonates, CKRT, Peritoneal dialysis Introduction Congenital anomalies of the kidney and urinary tract (CAKUT) encompass a spectrum of structural and functional abnormalities, including autosomal recessive polycystic kidney disease (ARPKD), bilateral renal dysplasia, posterior urethral valves (PUV), and other obstructive or cystic conditions. Severe phenotypes are often associated with oligohydramnios, pulmonary hypoplasia, and persistent pulmonary hypertension of the newborn, contributing to high neonatal morbidity and mortality [1–5]. Extracorporeal membrane oxygenation (ECMO) remains a potentially life-saving intervention for neonates with refractory cardiorespiratory failure [1, 6]. However, its role in CAKUT remains controversial due to limited survival data, high complication rates, and uncertainty regarding long-term renal prognosis [3, 5, 6]. Neonates with intrinsic renal disease often require concurrent kidney replacement therapy (KRT), such as continuous kidney replacement therapy (CKRT) or peritoneal dialysis, and may face prolonged hospitalization, recurrent infections, and eventual renal transplantation [1, 6]. In a retrospective analysis of the Extracorporeal Life Support Organization (ELSO) registry, Bagdure et al. found an overall survival to discharge of 42% in neonates with CAKUT-associated respiratory failure, with higher survival in obstructive lesions compared with intrinsic renal disease (p = 0.004), and no survivors among patients with renal agenesis/dysgenesis or cystic kidney disease [6]. Cuestas et al. in 2021 expanded on these findings with a registry-based cohort of 1385 neonatal ECMO runs, noting that survival improved over time but remained lower for patients with pulmonary hypoplasia secondary to renal anomalies [2]. Long-term follow-up data remain sparse although Short et al. reported good neurodevelopmental outcomes and successful renal transplant in a small series of survivors [4]. Despite advances in ECMO technology, neonatal critical care, and KRT, there is no consensus on ECMO candidacy criteria for neonates with CAKUT. Prenatal prognostic markers, institutional exclusion criteria, and nephrology involvement vary widely across centers. This study aimed to evaluate current institutional practices, clinical thresholds, and multidisciplinary decision-making processes regarding ECMO candidacy in neonates with CAKUT across multiple neonatal centers. A clearer understanding of current practices across institutions may inform the development of consensus guidelines, standardize care, and optimize outcomes. Methods Study design and setting We conducted a cross-sectional, web-based survey targeting physicians involved in ECMO decision-making for neonates with CAKUT. The study was coordinated through the Children’s Hospital Neonatal Consortium (CHNC), a collaborative network of academic and community NICUs in North America. The reporting of this study was guided by Checklist for Reporting of Survey Studies (CROSS) guidelines to enhance transparency and completeness [7]. Participants and recruitment Eligible participants included neonatologists, pediatric intensivists, pediatric surgeons, and pediatric nephrologists involved in the management of critically ill neonates and ECMO initiation decisions. Email invitations containing a secure REDCap (Research Electronic Data Capture) [8, 9] survey link were distributed through the CHNC network to the ECMO and Renal Focus Groups, as well as the Pedi-ECMO group, a subgroup of the Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) network. Institutional ECMO leaders were encouraged to share the survey link with their colleagues in various specialties of interest, including nephrologists, pediatric surgeons, and ECMO core neonatologists. Since the total number of recipients could not be determined, the response rate was unable to be calculated. Participation was voluntary, and multiple respondents from a single institution were encouraged to capture intra-institutional variability. For questions addressing intra-institutional practices, variations in physician responses were reviewed, and the modal response within each institution was chosen as the institutional response. Informed consent was implied by survey completion. Survey instrument design The survey was developed by a multidisciplinary working subgroup, including neonatologists, intensivists, nephrologists, and pediatric surgeons, representing select members of the CHNC ECMO and Renal Focus Groups. The survey underwent content validation through review by multidisciplinary experts in CAKUT and ECMO management. It was pilot tested with a small group of clinicians to ensure clarity, relevance, and feasibility of questions. Feedback from the pilot was incorporated to refine wording and question clarity, response options, branching logics, and survey flow. The survey was also time-tested to require approximately 15–20 min for completion prior to broader distribution. The survey consisted of 2 main components: general practices and hypothetical cases. The general practices (GP) component consisted of 28 core questions organized into 7 domains, while the hypothetical cases (HC) component consisted of 4 case vignettes. The GP component consisted of 7 domains to include the following: Participant Characteristics and Institutional Practices for CAKUT, Prenatal Screening and Prognostic Indicators, Pre-ECMO Evaluation and Candidacy Criteria, Nephrology Involvement, Renal Replacement on ECMO, Decannulation and Post-ECMO Dialysis Transition, and Institutional and Clinical Feasibility. Supplemental Material A provides the complete set of questions for each domain for the GP component. The HC component included 4 clinical vignettes with varying CAKUT phenotypes, pulmonary hypoplasia severity, and comorbidities. Each vignette included structured questions on ECMO eligibility, venoarterial (VA) versus venovenous (VV) mode selection, and rationale for offering or withholding ECMO. Supplemental Material A outlines each of the clinical vignettes and questions for the HP component. Data collection Survey responses were collected over 6 weeks via REDCap, hosted on secure, encrypted servers at the University of Texas Southwestern Medical Center. No patient identifiers were collected, and all responses were anonymous. Email addresses were collected solely to allow follow-up with participants who had partial responses or incomplete sections, ensuring completeness of the survey data, and for follow-up if interest in future studies were desired. Outcome measures The primary outcome was the proportion of respondents offering ECMO to neonates with specific CAKUT diagnoses in defined scenarios. Secondary outcomes were reported as the proportion of respondents who had specific exclusion criteria, utilized prenatal and postnatal prognostic markers, reported nephrology involvement, and integrated kidney support therapy during ECMO. Data analysis Survey responses were analyzed using descriptive methods consistent with survey-based research. Quantitative items are reported as frequencies and percentage distributions, and open-ended responses were reviewed to identify recurring themes. Given the exploratory nature of the study, analyses were intended to characterize practice patterns rather than to test statistical hypotheses. All analyses were conducted on de-identified responses to maintain anonymity, and inclusion of the email field did not influence survey results. Certain results are reported at the individual participant level, whereas others are presented at the institutional (center) level, as appropriate to the survey item. Surveys with incomplete responses were excluded from the final analysis to ensure data completeness and accuracy. Ethics approval This study was approved by The University of Texas Southwestern Medical Center Institutional Review Board (IRB) and Ethics Committee (Approval No. STU20251149). All procedures followed institutional and federal research guidelines. Informed consent was implied by voluntary survey completion. Additional details regarding IRB/ethics committee approval are provided in Supplementary Material B. Results General practices component Section 1: participant information A total of 118 physicians responded to the survey, of whom only 99 (84%) completed the GP component and were included in the final analysis. These 99 participants represented 42 institutions: 38 (90%) centers within the CHNC network, and 4 (10%) at non-CHNC centers. Participants included 75 (76%) neonatologists, and 8 (8%) each for pediatric intensivists, pediatric nephrologists, and pediatric surgeons. Of the respondents, 84% reported NICU as the primary site for neonatal ECMO at their institution, while 12% reported PICU as the primary site for ECMO. Participant characteristics are summarized in Table 1. Table 1Participant and institutional characteristics (N = 99)CharacteristicsResponsesN (%)SpecialtyNeonatologyPediatric critical carePediatric nephrologyPediatric surgery75 (76)8 (8)8 (8)8 (8)Unit where neonatal ECMO is primarily managedNICUPICUCV-ICUVaries83 (84)12 (12)1 (1)3 (3)Established institutional ECMO candidacy protocol for CAKUTYesNoIn development8 (8)78 (79)13 (13)Experience in management of CAKUT patients on ECMO in last 3 yearsYesNo61 (62)38 (38)Institutional practice to offer ECMO in patients with CAKUTYesNoDepends on case-by-case basis30 (30)4 (4)65 (66)Institution has established protocol for ECMO use in patients with CAKUT(N = 42 centers)YesNoIn development5 (12)31 (74)6 (14)Abbreviations: ECMO extracorporeal membrane oxygenation, CAKUT congenital anomalies of the kidney and urinary tract, NICU neonatal intensive care unit, PICU pediatric intensive care unit, CV-ICU cardiovascular intensive care unit Participant and institutional characteristics (N = 99) Neonatology Pediatric critical care Pediatric nephrology Pediatric surgery 75 (76) 8 (8) 8 (8) 8 (8) NICU PICU CV-ICU Varies 83 (84) 12 (12) 1 (1) 3 (3) Yes No In development 8 (8) 78 (79) 13 (13) Yes No 61 (62) 38 (38) Yes No Depends on case-by-case basis 30 (30) 4 (4) 65 (66) Institution has established protocol for ECMO use in patients with CAKUT (N = 42 centers) Yes No In development 5 (12) 31 (74) 6 (14) Abbreviations: ECMO extracorporeal membrane oxygenation, CAKUT congenital anomalies of the kidney and urinary tract, NICU neonatal intensive care unit, PICU pediatric intensive care unit, CV-ICU cardiovascular intensive care unit Section 2: prenatal screening and prognostic indicators Among the 95 participants, 64 (67%) reported the influence of prenatal prognostic markers in ECMO decision-making. The most commonly reported included the presence of anhydramnios (43%), observed-to-expected lung volumes (43%), amniotic fluid index (23%), and/or oligohydramnios (23%). Even though anhydramnios was the most common prenatal marker, only 10 (11%) physicians would utilize the diagnosis of anhydramnios at a specific gestational age (GA) for ECMO exclusion, most commonly < 22 weeks GA (6/10, 60%) and < 26 weeks GA (2/10, 20%). The GA thresholds for the initial identification of oligohydramnios or anhydramnios were not considered by the majority of the participants as exclusion criteria for ECMO exclusion. Responses for the use of prenatal markers for ECMO candidacy and exclusion can be found in Table 2. Table 2Prenatal screening and pre-ECMO candidacy criteria for neonates with CAKUTCriteria categoryResponsesN (%)Absolute exclusion diagnoses for ECMO(N = 95)Bilateral renal agenesisMulticystic dysplastic kidneysPUVARPKDRVTAll are considered49 (52)9 (9)7 (7)3 (3)2 (2)39 (41)Prenatal markers for ECMO candidacy(N = 95)AnhydramniosO/E lung volumeOligohydramniosAmniotic fluid indexChest-to-abdomen ratioNone of the above41 (43)41 (43)22 (23)22 (23)14 (15)31 (33)Gestational age at the time of identification of oligohydramnios or anhydramnios for ECMO exclusion(N = 95)Oligohydramnios Yes NoAnhydramnios Yes No2 (2)93 (98)10 (11)85 (89)Birthweight threshold for ECMO exclusion(N = 99) < 1.6 kg < 1.8 kg < 2.0 kg < 2.5 kg < 2.8 kgNo specific cut-off5 (5)25 (25)33 (33)10 (10)1 (1)25 (25)Respiratory thresholds for ECMO exclusion(N = 99 participants)Best achieved PaCO₂ (N = 42)• > 100 mmHg• > 80 mmHg• > 60 mmHg• No specific threshold24 (57)8 (19)2 (5)8 (19)Best achieved preductal O2 saturation (N = 21)• < 70%• < 80%• < 90%• No specific threshold11 (52)4 (19)2 (10)4 (19)Best achieved blood pH (N = 25)• < 6.8 mg/dL• < 7.0 mg/dL• < 7.2 mg/dL• No specific threshold14 (56)8 (32)1 (4)2 (8)No respiratory threshold35 (35%)Renal thresholds for ECMO exclusion(N = 99)Urine outputSerum creatinineVolume overloadNo renal threshold5 (5)1 (1)2 (2)91 (92)Abbreviations: ECMO extracorporeal membrane oxygenation, CAKUT congenital anomalies of the kidney and urinary tract, PUV posterior urethral valves, ARPKD autosomal recessive polycystic kidney disease, RVT renal venous thrombosis, O/E observed-to-expected, PaCO2 partial pressure of carbon dioxide Prenatal screening and pre-ECMO candidacy criteria for neonates with CAKUT Absolute exclusion diagnoses for ECMO (N = 95) Bilateral renal agenesis Multicystic dysplastic kidneys PUV ARPKD RVT All are considered 49 (52) 9 (9) 7 (7) 3 (3) 2 (2) 39 (41) Prenatal markers for ECMO candidacy (N = 95) Anhydramnios O/E lung volume Oligohydramnios Amniotic fluid index Chest-to-abdomen ratio None of the above 41 (43) 41 (43) 22 (23) 22 (23) 14 (15) 31 (33) Gestational age at the time of identification of oligohydramnios or anhydramnios for ECMO exclusion (N = 95) Oligohydramnios Yes No Anhydramnios Yes No 2 (2) 93 (98) 10 (11) 85 (89) Birthweight threshold for ECMO exclusion (N = 99) < 1.6 kg < 1.8 kg < 2.0 kg < 2.5 kg < 2.8 kg No specific cut-off 5 (5) 25 (25) 33 (33) 10 (10) 1 (1) 25 (25) Respiratory thresholds for ECMO exclusion (N = 99 participants) Best achieved PaCO₂ (N = 42) • > 100 mmHg • > 80 mmHg • > 60 mmHg • No specific threshold 24 (57) 8 (19) 2 (5) 8 (19) Best achieved preductal O2 saturation (N = 21) • < 70% • < 80% • < 90% • No specific threshold 11 (52) 4 (19) 2 (10) 4 (19) Best achieved blood pH (N = 25) • < 6.8 mg/dL • < 7.0 mg/dL • < 7.2 mg/dL • No specific threshold 14 (56) 8 (32) 1 (4) 2 (8) Renal thresholds for ECMO exclusion (N = 99) Urine output Serum creatinine Volume overload No renal threshold 5 (5) 1 (1) 2 (2) 91 (92) Abbreviations: ECMO extracorporeal membrane oxygenation, CAKUT congenital anomalies of the kidney and urinary tract, PUV posterior urethral valves, ARPKD autosomal recessive polycystic kidney disease, RVT renal venous thrombosis, O/E observed-to-expected, PaCO2 partial pressure of carbon dioxide Section 3: pre-ECMO evaluation and candidacy criteria Overall, 30/99 (30%) participants reported that their institution offers ECMO to neonates with CAKUT, while 65/99 (66%) indicated that candidacy is determined on a case-by-case basis. Excluding the 4 participants who do not offer ECMO to patients with CAKUT, 49/95 (52%) physicians considered bilateral renal agenesis (BRA) an absolute contraindication to ECMO, while 39/95 (41%) would consider all renal anomalies for ECMO candidacy. Of the 99 participants, 42 (42%) selected PaCO2 levels as an important respiratory clinical parameter to be used for ECMO exclusion. Of these 42 participants, 24 (57%) stated that ECMO would not be offered if the best achieved PaCO2 level was > 100 mmHg. Birthweight cutoffs, respiratory and renal laboratory thresholds, and other exclusion criteria for ECMO are detailed in Table 2. Section 4: nephrology involvement Of the 99 participants, 61 (62%) reported that they had previous experience in managing patients with CAKUT on ECMO in the last 3 years, of which 48/61 (79%) were neonatologists, 5/61 (8%) were pediatric surgeons, 2/61 (3%) were pediatric intensivists, and 6/61 (10%) were pediatric nephrologists. Excluding the pediatric nephrologists, the 55 physicians (neonatologists, pediatric intensivists, and pediatric surgeons) were queried regarding the involvement of pediatric nephrology at their respective centers representing 21 centers. Nephrology was typically involved in ECMO decision-making and initiation in 16 (76%) centers. Nephrology participation during ECMO rounds occurred frequently (2–3 times weekly up to daily) in 15 (71%) of the centers, while 6 (29%) centers reported involvement as needed. In addition, 45/55 (82%) physicians reported that early consultation (within 24–48 h) improves patient care for patients with CAKUT on ECMO. Section 5: kidney replacement therapy on ECMO Thirty participants, representing 16 centers, who offer ECMO to patients with CAKUT were asked about CKRT usage. Of the 16 centers, CKRT was reported to be routinely used in 8 (50%) centers for neonates on ECMO. Four participants from 4 (25%) centers reported not having any recent CAKUT-ECMO cases. Of the 12 centers with recent CAKUT-ECMO cases, CKRT was initiated after ECMO cannulation based on deterioration in renal function in 7 (58%) centers and concurrently initiated with ECMO in 4 (34%) centers. The 26 participants who had CAKUT-ECMO cases at their center cited the following challenges in implementing KRT on ECMO: hemodynamic instability (50%), anticoagulation (42%), vascular access difficulties (23%), and institutional protocols and staff expertise (27%). KRT and anticoagulation approaches are summarized in Table 3. Table 3Kidney replacement therapy practices during ECMOInstitutional CKRT practicesResponses N (%)Frequency of CKRT use in CAKUT ECMO patients(N = 16 centers)RoutinelyOccasionallyRarelyNo CAKUT-ECMO cases yet8 (50)3 (19)1 (6)4 (25)Timing of CKRT initiation(N = 12 centers)Pre-ECMO initiationConcurrent with ECMO initiationPost-ECMO based on renal deteriorationDepends on case-by-case0 (0)4 (34)7 (58)1 (8)KRT-ECMO integration protocol present(N = 12 centers)YesNoIn development9 (75)2 (17)1 (8)Physician-specific experienceResponsesN (%)Challenges in implementation of CKRT(N = 26 participants)Vascular access difficultiesHemodynamic instabilityAnticoagulation managementEquipment compatibility w/ECMOInstitutional protocols/staff expertise6 (23)13 (50)11 (42)0 (0)7 (27)Anticoagulation approach(N = 26 participants)HeparinBivalirudinHeparin or bivalirudinCitrate11 (42)6 (23)9 (35)1 (4)Abbreviations: CKRT continuous kidney replacement therapy, ECMO extracorporeal membrane oxygenation, CAKUT congenital anomalies of the kidney and urinary tract, KRT kidney replacement therapy Kidney replacement therapy practices during ECMO Frequency of CKRT use in CAKUT ECMO patients (N = 16 centers) Routinely Occasionally Rarely No CAKUT-ECMO cases yet 8 (50) 3 (19) 1 (6) 4 (25) Timing of CKRT initiation (N = 12 centers) Pre-ECMO initiation Concurrent with ECMO initiation Post-ECMO based on renal deterioration Depends on case-by-case 0 (0) 4 (34) 7 (58) 1 (8) KRT-ECMO integration protocol present (N = 12 centers) Yes No In development 9 (75) 2 (17) 1 (8) Challenges in implementation of CKRT (N = 26 participants) Vascular access difficulties Hemodynamic instability Anticoagulation management Equipment compatibility w/ECMO Institutional protocols/staff expertise 6 (23) 13 (50) 11 (42) 0 (0) 7 (27) Anticoagulation approach (N = 26 participants) Heparin Bivalirudin Heparin or bivalirudin Citrate 11 (42) 6 (23) 9 (35) 1 (4) Abbreviations: CKRT continuous kidney replacement therapy, ECMO extracorporeal membrane oxygenation, CAKUT congenital anomalies of the kidney and urinary tract, KRT kidney replacement therapy Section 6: decannulation and post-ECMO dialysis transition Of the 26 participants with CAKUT-ECMO cases, 9 (35%) considered placing a peritoneal dialysis (PD) catheter during or after ECMO to facilitate immediate post-decannulation dialysis. Placement of a superior vena cava catheter at the time of decannulation was reported by 19 (73%) of respondents as the access of choice for dialysis immediately after ECMO. However, peritoneal dialysis was reported to be the most common modality for long-term dialysis by 13 (50%) of participants. The most common barriers reported for dialysis when transitioning off ECMO included hemodynamic instability (50%), dialysis access complications (50%), and multiorgan dysfunction (42%). Section 7: institutional and clinical feasibility Of the total 42 centers, a defined protocol for ECMO use in neonates with CAKUT existed in only 5 (12%) of institutions, while 6 (14%) reported that guidelines were currently in development. The 99 participants reported the following to be major challenges in the implementation of ECMO in patients with CAKUT: disease heterogeneity (79%), uncertainty in prognosis (66%), ethical concerns (62%), and high mortality rates (59%). Hypothetical cases component For the HC component there were a total of 77% (91/118) complete responses which included 71 (78%) neonatologists, 6 (7%) pediatric intensivists, 8 (9%) pediatric nephrologists, and 6 (7%) pediatric surgeons representing a total of 38 centers. Incomplete responses were excluded from analysis. Decision-making varied across the four clinical vignettes. Case 1 presented a neonate with moderate obstructive renal disease secondary to posterior urethral valves for which ECMO was considered appropriate by 83 (91%) of the participants. Case 2 represented a severe intrinsic renal disease as a result of bilateral renal agenesis for which ECMO was considered appropriate by only 10 (11%) of participants. Case 3 presented a patient with bilateral cystic kidneys resulting in moderate intrinsic renal disease for which 77 (85%) of the participants would consider ECMO. Lastly, case 4 presented a patient with trisomy 21 and uretero-pelvic junction obstruction resulting in moderate-severe obstructive renal disease for which ECMO was considered by 43 (47%) of participants. Rationales for ECMO inclusion or exclusion, preferred ECMO modes, and scenario-specific decision-making patterns are summarized in Table 4. Table 4Summary of responses for hypothetical scenarios (N = 91 participants)CAKUT diagnosis (disease severity)ECMO offeredECMO mode: VA or VVCommon rationale for offering ECMOCommon rationale for withholding ECMON (%)N (%)N (%)N (%)Case scenario #1PUV (moderate)Yes83 (91)No8 (9)VA33 (40)VV40 (48)Defer to ICU10 (12%)Respiratory status and hemodynamics may improve after ECMOTendency to oxygenate and ventilate75 (90)66 (80)Respiratorydisease is likely irreversibleFailed tendency to oxygenate and ventilate5 (63)5 (63)Case scenario #2Bilateral renal agenesis (severe)Yes10 (11)No81 (89)VA4 (40)VV2 (20)Defer to ICU4 (40%)Opportunity for fluid removalRespiratory status and hemodynamics may improve after ECMO8 (80)6 (60)Respiratory disease is likely irreversibleFailed tendency to oxygenate and ventilate65 (80)64 (79)Case scenario #3Bilateral cystic kidneys (moderate)Yes77 (85)No14 (15)VA33 (43)VV33 (43)Defer to ICU11 (14%)Respiratory status and hemodynamics may improve after ECMOOpportunity for fluid removal69 (90)63 (82)Respiratory disease is likely irreversibleUnstable/worsening cardiorespiratory conditionLack of ECMO survivability data for patients with bilateral cystic kidneys10 (71)9 (64)9 (64)Case scenario #4UPJ obstruction + VSD + trisomy 21 (moderate–severe)Yes43 (47)No48 (53)VA25 (58)VV10 (23)Defer to ICU8 (19%)Opportunity for fluid removalObstructive renal disease may have improved survival benefit35(81)31(72)Failed tendency to oxygenate or ventilateRespiratory disease is irreversibleUnstable/worsening cardiorespiratory conditions43 (90)28 (58)Abbreviations: CAKUT congenital anomalies of the kidney and urinary tract, ECMO extracorporeal membrane oxygenation, VA venoarterial, VV venovenous, PUV posterior urethral valves, UPJ ureteropelvic junction, VSD ventricular septal defect Summary of responses for hypothetical scenarios (N = 91 participants) Yes 83 (91) No 8 (9) VA 33 (40) VV 40 (48) Defer to ICU 10 (12%) Respiratory status and hemodynamics may improve after ECMO Tendency to oxygenate and ventilate 75 (90) 66 (80) Respiratory disease is likely irreversible Failed tendency to oxygenate and ventilate 5 (63) 5 (63) Yes 10 (11) No 81 (89) VA 4 (40) VV 2 (20) Defer to ICU 4 (40%) Opportunity for fluid removal Respiratory status and hemodynamics may improve after ECMO 8 (80) 6 (60) Respiratory disease is likely irreversible Failed tendency to oxygenate and ventilate 65 (80) 64 (79) Yes 77 (85) No 14 (15) VA 33 (43) VV 33 (43) Defer to ICU 11 (14%) Respiratory status and hemodynamics may improve after ECMO Opportunity for fluid removal 69 (90) 63 (82) Respiratory disease is likely irreversible Unstable/worsening cardiorespiratory condition Lack of ECMO survivability data for patients with bilateral cystic kidneys 10 (71) 9 (64) 9 (64) Yes 43 (47) No 48 (53) VA 25 (58) VV 10 (23) Defer to ICU 8 (19%) Opportunity for fluid removal Obstructive renal disease may have improved survival benefit 35 (81) 31 (72) Failed tendency to oxygenate or ventilate Respiratory disease is irreversible Unstable/worsening cardiorespiratory conditions 43 (90) 28 (58) Abbreviations: CAKUT congenital anomalies of the kidney and urinary tract, ECMO extracorporeal membrane oxygenation, VA venoarterial, VV venovenous, PUV posterior urethral valves, UPJ ureteropelvic junction, VSD ventricular septal defect Discussion This multicenter survey demonstrates heterogeneity in ECMO candidacy criteria, prenatal prognostic considerations, and peri-ECMO renal management for neonates with CAKUT across CHNC and non-CHNC institutions. The criteria for ECMO candidacy and initiation can vary significantly across NICUs and depend on institutional protocols, clinical training, and judgement. Some experienced physicians and high-volume ECMO centers may utilize ECMO more generously for CAKUT-related complications than other centers that adhere to stricter guidelines. This may present large inconsistencies in patient selection impacting outcomes as some patients who might benefit from ECMO may be undertreated in more conservative practices, and vice versa. Neonatal ECMO utilization and survival for congenital diaphragmatic hernia and other neonatal respiratory diseases have been widely studied, but data on ECMO use in CAKUT anomalies is lacking [5]. This knowledge gap and variability in ECMO practices for neonates with CAKUT across various facilities underscores the need for standardization and collaboration. Prenatal markers and measurements may help prognosticate the severity of pulmonary hypoplasia, renal disease, and postnatal survival in patients with CAKUT [5, 10]. In our survey study, anhydramnios and observed-to-expected lung volume ratios were the most common prenatal findings reported to influence ECMO candidacy decisions. However, neither was considered to be an absolute contraindication. In fact, only about 11% (10 out of 95 participants) would utilize detection of anhydramnios as an absolute contraindication for ECMO. Furthermore, prenatal markers were not considered at all for ECMO decision-making by 31 out of 95 (33%) participants. This variability reflects both limited published outcomes and the complex interplay between pulmonary prognosis and renal function in this patient population [1, 2, 6]. A recent retrospective study by Vincent et al. described a cohort of 23 neonates, of whom 7 (30%) required ECMO. Of these ECMO-managed neonates, 86% (p = 0.026) had prenatal findings of anhydramnios at a median gestational age of 26 weeks, and 29% of those treated with ECMO survived to discharge [5]. In another retrospective study, Short et al. described a cohort of five patients with CAKUT requiring ECMO support, reporting an 80% survival. The single nonsurvivor had documented anhydramnios prior to 28 weeks’ gestation and refractory pulmonary hypertension [4]. Similarly, Baudin et al. reported that a diagnosis of oligohydramnios before 32 weeks’ gestation was strongly associated with increased mortality in patients with congenital kidney disease [10]. While prenatal parameters such as lung volume, amniotic fluid status, and gestational age at the onset of anhydramnios may influence ECMO decision-making in some centers, only a minority of survey participants in our study considered gestational age or amniotic fluid status as part of ECMO candidacy, and the reported gestational age thresholds influencing clinical decision-making varied widely. This lack of consensus emphasizes the need for further studies to better evaluate prenatal markers for predicting survival and outcomes for neonates with CAKUT and perhaps developing a stratification model for predicting outcomes and counselling parents. Bilateral renal agenesis was the most common renal diagnosis for ECMO exclusion in our study. Other renal anomalies were decided in an individualized, multidisciplinary decision-making approach. Surprisingly, 41% of physicians in this survey would consider ECMO for all renal anomalies, including bilateral renal agenesis. This may suggest that more experienced, higher volume ECMO centers are more likely to consider these anomalies. Multiple studies have shown that obstructive uropathies may confer a survival benefit compared with intrinsic renal disease [2, 6, 10]. However, studies have also shown poor prognosis and higher mortality in infants with intrinsic renal disease and especially those carrying the diagnosis of bilateral renal agenesis [10, 11]. In our HC component of the survey, only 10 (11%) of participants would trial ECMO for case scenario 2 that presented the diagnosis of anhydramnios and bilateral renal agenesis. Nonetheless, consensus on ECMO candidacy for CAKUT anomalies remains diverse among physicians and centers due to limited survivability data, uncertainty in prognosis, ethical dilemmas, and lack of clear guidelines, as evidenced by the responses in our survey. Postnatal evaluation of patients with CAKUT may have a greater influence on determining ECMO candidacy than prenatal markers. These may include (and are not limited to) birthweight, respiratory sufficiency, and renal clinical parameters. Bagdure et al. reported that birthweight at the time of ECMO cannulation was the only independent predictor of survival in patients with CAKUT [6]. Our findings revealed significant variability in birthweight cutoffs for CAKUT-ECMO exclusion among centers and participants. While a birthweight of < 2.0 kg was the most common response, this varied widely from < 1.6 up to < 2.5 kg as limits for ECMO exclusion. Responses for respiratory clinical parameters also varied in our study. However, most participants agreed that patients with the best achieved PaCO2 levels of > 100 mmHg, and best achieved preductal oxygen saturation of < 70% would not be strong candidates for ECMO. These clinical selection criteria may be shaped by retrospective studies of infants with congenital diaphragmatic hernia indicating lower survival despite ECMO support in those who failed to achieve lower respiratory laboratory values [12, 13]. The role of pediatric nephrology in early decision-making also varied significantly. While some centers routinely involved nephrology within 24–48 h of life, others engage them only after ECMO initiation. This discrepancy likely impacts both peri-ECMO fluid management and post-ECMO KRT planning. Notably, early nephrology involvement is perceived to improve outcomes, yet formalized protocols for CKRT initiation, anticoagulation strategies, and decannulation planning are not yet present in many institutions. Variations in operational and ethical considerations were also evident. Institutions without defined protocols often rely on case-by-case deliberations, which may lead to inconsistency in care, prolonged decision-making, and inequities in access to ECMO as a therapeutic modality. Ethical challenges—particularly surrounding perceived futility, parental requests for maximal intervention, and anticipated long-term morbidity—were common themes. The hypothetical case scenarios in our study provided an opportunity to assess participants’ responses regarding ECMO candidacy based on diagnosis, presentation severity, and objective clinical data. Physicians were more likely to initiate ECMO for obstructive renal disease than for intrinsic renal disease with similar clinical presentations (case 1 vs. case 3). Our survey results demonstrated that even among those physicians who would opt for comfort care discussions with family, most of them would still consider a trial of ECMO if parents insisted on maximal intervention or if there is an absence of other genetic syndromes and anomalies. This discrepancy likely reflects ethical dilemmas and the paucity of survivability data, which complicates physicians’ ability to counsel families regarding survival outcomes for patients with CAKUT supported on ECMO. However, for those same physicians who opted for comfort care discussions with family, the consideration of offering ECMO decreased as the clinical diagnosis changed from obstructive renal disease (case 1) to intrinsic renal disease (case 3) and further declined when a genetic syndrome (trisomy 21) was added (case 4). Expectedly, very few physicians would consider ECMO for bilateral renal agenesis with severe pulmonary hypertension. Furthermore, providers who offered ECMO to case 2 (bilateral renal agenesis with severe pulmonary hypertension) and case 4 (UPJ obstruction with moderate pulmonary hypertension and trisomy 21) were from institutes that offer ECMO to all patients with CAKUT anomalies. The survey did not collect data on annual ECMO case volumes for each center; therefore, responses could not be stratified by center volume. There were no major differences in the choice of ECMO mode among participants, which may be due to institutional and personal practice preferences. The most common reasons to offer ECMO for the hypothetical cases included opportunity for fluid removal, possible improvement in respiratory and hemodynamics with ECMO support, and that obstructive uropathies may have a survival benefit in neonates with CAKUT. There is undoubtedly an interest among experts in offering at least a trial of ECMO support to various CAKUT phenotypes, often in the setting of individualized, multidisciplinary deliberation. However, this interest is accompanied by substantial variability in clinical thresholds, risk tolerance, and institutional policies, highlighting the absence of a unified framework to guide decision-making. Perhaps a structured approach, such as a modified Delphi process involving multidisciplinary experts across participating centers, could be used to identify areas of agreement, define best practices, and prioritize research gaps. Integration of survey findings with available outcome data from multicenter cohorts would further inform evidence-based consensus recommendations and, ultimately, development of standardized practice guidelines. Study limitations Despite these insights, several limitations of this study should be acknowledged. Because the survey was distributed via listservs without access to the full recipient list, the denominator of eligible participants was unknown, and a response rate could not be calculated. Additionally, voluntary participation introduces the possibility of selection bias, as centers with greater experience or interest in CAKUT–ECMO may have been more likely to respond. Conclusion Collectively, these findings underscore the paucity of data for ECMO use in neonates with CAKUT anomalies and the urgent need for standardized, evidence-informed guidelines for ECMO utilization in neonates with CAKUT. Such guidelines could help better define prenatal pulmonary and renal prognostic data, enhance ability to counsel parents, clarify exclusion criteria, outline optimal timing and modality for KRT, and establish a structured role for pediatric nephrology in ECMO decision-making. Multicenter outcome studies—particularly those that stratify results by CAKUT subtype, severity of pulmonary hypoplasia, and renal prognosis—would be essential to inform on these recommendations. Supplementary Information Below is the link to the electronic supplementary material. Graphical abstract (PPTX 80.8 KB) Supplementary file 1 (PDF 80.4 KB) Supplementary file 2 (DOCX 45.1 KB) Acknowledgements The authors would like to thank the University of Texas Southwestern Medical Center for institutional support and resources that contributed to the completion of this study. We also acknowledge our colleagues and collaborators from Children's Hospital Neonatal Consortium who assisted with survey development, survey distribution and data collection. Abbreviations Autosomal recessive polycystic kidney disease Congenital anomalies of the kidney and urinary tract Children’s Hospital Neonatal Consortium Continuous kidney replacement therapy Extracorporeal membrane oxygenation Extracorporeal Life Support Organization Gestational age Peritoneal dialysis Posterior urethral valves Uretero-pelvic junction Venoarterial or venovenous extracorporeal membrane oxygenation Author contribution Conceptualization: SA and AM. Methodology: SA, AM, NR, and RD. Survey development: SA, AM, NR, RD, BM, JD, SJ, AS, JH, KD, KS, KM, KB, LL, MR, PT, RS, SP, SK, SG, SH, TM, TB, and TN. Data collection: SA. Formal analysis: SA and AM. Data interpretation: SA, AM, NR, RD, JD, and BM. Writing—original draft: SA and AM. Writing—review and editing: SA, AM, NR, RD, BM, JD, SJ, AS, JI, JH, KD, KS, KM, KB, LL, MR, PT, RS, SP, SK, SG, SH, TM, TB, and TN. Supervision: SA, AM, BM, RD, and NR. Project administration: SA and AM. Senior author: AM. Funding This study was supported by departmental and institutional resources. No external funding was received. Data availability De-identified data supporting the findings of this study are available from the corresponding author upon request and with appropriate institutional approvals. Declarations Conflict of interest The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Graphical abstract (PPTX 80.8 KB) Supplementary file 1 (PDF 80.4 KB) Supplementary file 2 (DOCX 45.1 KB) Data Availability Statement De-identified data supporting the findings of this study are available from the corresponding author upon request and with appropriate institutional approvals.