Early outcomes after post-cardiotomy extracorporeal membrane oxygenation in paediatric patients: a contemporary, binational cohort study

Abstract

Open in new tabDownload slide

OBJECTIVES

The aim of this study was to assess the early outcomes and risk factors of paediatric patients requiring extracorporeal membrane oxygenation after cardiac surgery (post-cardiotomy).

METHODS

Retrospective binational cohort study from the Australia and New Zealand Congenital Outcomes Registry for Surgery database. All patients younger than 18 years of age who underwent a paediatric cardiac surgical procedure from 1 January 2013 to 31 December 2021 and required post-cardiotomy extracorporeal membrane oxygenation (PC-ECMO) in the same hospital admission were included in the study.

RESULTS

Of the 12 290 patients included in the study, 376 patients required post-cardiotomy ECMO (3%). Amongst these patients, hospital mortality was 35.6% and two-thirds of patients experienced a major complication. Hypoplastic left heart syndrome was the most common diagnosis (17%). The Norwood procedure and modified Blalock–Taussig shunts had the highest incidence of requiring PC-ECMO (odds ratio of 10 and 6.8 respectively). Predictors of hospital mortality after PC-ECMO included single-ventricle physiology, intracranial haemorrhage and chylothorax.

CONCLUSIONS

In the current era, one-third of patients who required PC-ECMO after paediatric cardiac surgery in Australia and New Zealand did not survive to hospital discharge. The Norwood procedure and isolated modified Blalock–Taussig shunt had the highest incidence of requiring PC-ECMO. Patients undergoing the Norwood procedure had the highest mortality (48%). Two-thirds of patients on PC-ECMO developed a major complication.

INTRODUCTION

In the contemporary era, outcomes after paediatric cardiac surgery are excellent [1]. However, a small proportion of patients require extracorporeal membrane oxygenation (ECMO) after surgery. This may be due to failure to be weaned from cardiopulmonary bypass (CPB), post-operative low cardiac output state (LCOS), refractory cardiac arrest or respiratory failure [2]. Paediatric post-cardiac surgical ECMO [post-cardiotomy ECMO (PC-ECMO)] was first described in 1973, for the management of post-operative LCOS after repair of tetralogy of Fallot [3]. Since then, many publications have reported single-centre retrospective survival data and predictors of outcome, in addition to reports from registries such as the Extracorporeal Life Support Organisation (ELSO) database [4–9].

The requirement for PC-ECMO predicts greater mortality and morbidity in both adult and paediatric populations. Factors previously reported to predict survival in children requiring PC-ECMO include increasing complexity of their underlying congenital heart disease, longer duration of CPB, or myocardial ischaemia times, longer ECMO runs, the development of renal failure, and the requirement for a second run on ECMO [4, 8, 10]. The Norwood operation and single-ventricle physiology have been studied more extensively than other diagnostic and procedural groups, with comparatively poorer survival outcomes [11–14]. Neonates, particularly low weight or premature have also been shown to have poorer survival and a higher risk of morbidity, particularly intracranial haemorrhage and neuro-developmental delay [5, 15, 16].

However, most of the existing literature on this subject relates to patients who were operated >15 years ago and may not reflect contemporary paediatric cardiac surgical practice. Furthermore, most studies were either small single-centre retrospective studies, or registry studies which lacked granular detail regarding anatomical and surgical factors. The aim of this study was to evaluate early outcomes of paediatric patients requiring PC-ECMO in a large bi-national patient population and identify potential predictors of adverse outcomes.

METHODS

Sample

All patients who were 18 years or younger and underwent a paediatric cardiac surgical procedure in Australia and New Zealand between 1 January 2013 and 31 December 2021 were identified from the Australia and New Zealand Congenital Outcomes Registry for Surgery (ANZCORS) database [1]. Patients who required ECMO post-operatively in the same hospital admission were eligible for inclusion. Data on each patient were retrospectively extracted from the ANZCORS database and from the patient’s hospital records.

Ethical statement

ANZCORS includes perioperative data on every paediatric cardiac surgical procedure performed in all 5 centres in Australia and New Zealand (Auckland, Brisbane, Melbourne, Perth, Sydney). Details about ANZCORS have been previously published [1]. Binational ethical approval for this study was obtained and the requirement for consent was waived (HREC/21/QCHQ/80891, approved 26 October 2021).

Endpoints

The primary end point was hospital mortality, from the date of the patient’s index cardiac surgical procedure. ‘Death on ECMO’ included patients who died on ECMO and those who were decannulated but expected to die within 24 h after decannulation. Secondary end points included 30-day mortality, procedure-specific mortality, reoperation for revision of cardiac repair or correction of residual haemodynamic defects and major morbidity. Major morbidity included ischaemic and haemorrhagic strokes, re-exploration for bleeding, chylothorax, sepsis, pneumonia, renal replacement therapy, necrotizing enterocolitis (NEC), requirement for ECMO circuit exchanges, and the requirement for a second run on ECMO.

Statistical analysis

The baseline characteristics of the PC-ECMO and non-ECMO cohorts were compared using the Wilcoxson rank-sum test and Fisher’s exact test. Survival was depicted using Kaplan–Meier survival curves. The Mantel–Cox rank-sum test was utilized to compare survival between groups. Univariable Cox regression was used to explore associations between risk factors and the risk of hospital mortality. Risk factors which were significant in the univariable analysis (P < 0.001) were entered into a multivariable Cox regression, with collinearity assessed by calculating the variance inflation factor and the proportional hazards assumption assessed using the scaled Schoenfeld residuals. Statistical and graphical analysis was performed using R (R Foundation for Statistical Computing, Vienna, Austria) and GraphPad Prism 9 (Dotmatics, Boston, USA).

RESULTS

Between 1 January 2013 and 31 December 2021, 12 290 patients underwent a paediatric cardiac surgical procedure in Australia and New Zealand. A total of 376 patients (3%) required PC-ECMO. A total of 11 914 (97%) patients did not require PC-ECMO (non-ECMO cohort).

Baseline characteristics

Baseline patient characteristics for the PC-ECMO and non-ECMO cohorts are outlined in Table 1. Patients in the PC-ECMO cohort were younger with lower weight at surgery and higher Society for Thoracic Surgeons-European Association for Cardio-Thoracic Surgery (STAT) mortality scores for their primary cardiac surgical procedure. They had longer CPB and myocardial ischaemia times than patients in the non-ECMO cohort. Amongst the patients who required PC-ECMO, 350 (93.1%) procedures were performed using CPB while 26 (6.9%) were non-CPB procedures. In the non-ECMO cohort, 9404 (78.9%) procedures were performed using CPB while 2510 (21.1%) procedures were non-CPB procedures.

The median age at surgery in the PC-ECMO cohort was 22 days [interquartile range (IQR) 5–124.25 days]. Neonates represented more than half of the PC-ECMO cohort with an additional 30% of patients between 1 month and 1 year of age. Thus, 80% of patients were under 1 year of age. 208 patients (55%) were neonates; 57 (15.1%) patients had a gestation of 37 weeks or less, 45 (12%) had a birthweight of 2500 g or less and 52.4% had an antenatal diagnosis. Non-cardiac anomalies were also more common amongst the PC-ECMO cohort (14.4% vs 10.9%, P = 0.044) (see Supplementary Material, Table S1).

Table 2 and Supplementary Material, Table S2 outline the most common primary diagnoses and surgical procedures in the PC-ECMO cohort. A total of 250 patients (66.5%) had a biventricular circulation, while 126 patients (33.5%) had single-ventricle physiology. The most common diagnosis amongst patients requiring PC-ECMO was hypoplastic left heart syndrome, with the most common operation being the Norwood procedure.

Timing of initiation of post-cardiotomy extracorporeal membrane oxygenation

The timing of PC-ECMO initiation and its relation to mortality are summarized in Supplementary Material, Table S3. Nearly 90% of patients had PC-ECMO instituted within the initial 48 h postoperatively. There was a trend towards higher mortality amongst patients who were placed on ECMO >48 h postoperatively. Mortality was highest (65%) when PC-ECMO was instituted at >7 days postoperatively; amongst such cases, the majority (75%) were placed on ECMO during resuscitation for refractory cardiac arrest.

Society for Thoracic Surgeons-European Association for Cardio-Thoracic Surgery mortality category

The majority of patients requiring PC-ECMO had a STAT mortality score of 4 or 5 (n = 225, 59.8%). Supplementary Material, Table S4 gives the stratification of the PC-ECMO cohort by STAT category. However, 85 (22.6%) patients who required PC-ECMO had undergone a STAT category 1 or 2 operation, with most of these cases being in STAT category 2 (18.8%). A review of these patients’ medical records suggests that there were, in general, other contributing factors for their deterioration and the requirement for PC-ECMO. Of the 14 patients in STAT category 1 who required PC-ECMO, 7 (50%) had a major non-cardiac anomaly. One patient underwent a redo sternotomy for conduit replacement on a background of congenitally corrected transposition of the great arteries, 1 patient was intubated and ventilated with acute cardiac failure on a background of rheumatic aortic valve disease, and 1 patient was a low birthweight neonate. Amongst the 71 patients in STAT 2 who required PC-ECMO, 19 (27%) had a pre-existing major non-cardiac anomaly, 18 (25%) were ventilated preoperatively, 6 (8%) were placed on ECMO prior to surgery and 5 (7%) patients had received cardiopulmonary resuscitation (CPR) for a preoperative cardiac arrest.

Extracorporeal membrane oxygenation-specific details

The indications for ECMO, type of ECMO, site of cannulation, left heart decompression and weaning strategies from ECMO are outlined in Table 3. The most common indication for initiation of PC-ECMO was failure to wean from CPBypass (n = 171, 45.2%). The trans-mediastinal approach was used for cannulation in 343 patients (91.2%). In patients with a biventricular circulation, left heart decompression was performed in 38 patients (10.1% overall, 15.2% of biventricular hearts), most commonly via direct cannulation of the left atrium, or via the left atrial appendage. Decompression of the left heart was not an independent predictor of hospital mortality on multivariable analysis.

A total of 322 (85.6%) patients were successfully weaned from PC-ECMO. Weaning strategies were decided at the discretion of the treating teams, as outlined in Table 3. 259 (69%) patients were weaned using reduction of circuit flows, with smaller cohorts of patients being weaned via insertion of an arterio-venous bridge (n = 23, 6%) or via pump-controlled retrograde trial off ECMO (n = 19, 5%).

Of the 376 patients who were placed on PC-ECMO, there were 52 (14.3%) deaths on ECMO. Two patients were transitioned to long-term ventricular assist devices. No patient underwent transplantation as an exit strategy from ECMO or during the same hospital admission. 322 (85.6%) patients were successfully weaned from PC-ECMO after a median duration of 4 days (range 0–43 days). Patients who survived to hospital discharge had a shorter duration of PC-ECMO (4 vs 5 days, P < 0.001). 45 (12%) patients had a second run on ECMO with a 30-day mortality of 35.6% (compared to 21.0% in patients with only 1 ECMO run, P = 0.037) and hospital mortality of 66.7% (compared to 31.8% in patients with only 1 ECMO run, P < 0.001). There was a trend towards a higher mortality in patients for whom ECMO was commenced in >7 days postoperatively, while patients who were placed on ECMO in the operating theatre had the lowest mortality (see Supplementary Material, Table S3).

Primary end point (hospital mortality)

Hospital mortality was 35.6% in the PC-ECMO cohort (Fig. 1). The median time to death amongst patients who died in hospital was 21 days (IQR 8.25—55.75). Hospital mortality was 28.8% amongst patients with biventricular hearts versus 48.8% for those with a single ventricle (P = 0.002) (Fig. 2 and Supplementary Material, Table S5). Supplementary Material, Table S6 summarizes hospital mortality in the PC-ECMO cohort stratified by age; neonates had the highest hospital mortality (41.8%).

Predictors of hospital mortality

Multivariable regression analysis for predictors of hospital mortality for patients who required PC-ECMO is detailed in Table 4 (univariable analysis is detailed in Supplementary Material, Table S7). On multivariable analysis, the only factors predictive of hospital mortality included single-ventricle status and intracranial haemorrhage.

Secondary end points

Thirty-day mortality

A total of 85 (22.6%) patients in the PC-ECMO cohort died within 30 days (Fig. 3). The median time to death was 11.5 days (IQR 7–18.25). In patients with a biventricular circulation, 30-day mortality was 16.8% (n = 42) compared to 31.0% (n = 39) in the single-ventricle cohort (Supplementary Material, Table S5).

Procedure-specific mortality

Procedure-specific mortality is presented in Table 2. Amongst patients who required PC-ECMO, patients undergoing the Norwood procedure had the highest hospital mortality (48.3%). The Norwood procedure and isolated modified Blalock–Taussig shunt (mBTS) had the highest incidence of requiring PC-ECMO (univariable odds ratio of 10.0 and 6.8, respectively).

Reintervention and residual haemodynamic lesions

Diagnostic cardiac catheterization was performed in 32 (8.5%) patients, with a further 12 (3.2%) patients undergoing an unplanned interventional cardiology procedure whilst on PC-ECMO. Eighty-four (22.3%) patients had their index surgical repair revised, while an additional 12 patients (3.2%) underwent surgery for repair of a newly diagnosed structural lesion whilst on PC-ECMO. Amongst patients who underwent a revision of their index surgical repair, survival to decannulation from ECMO was 84.5% (n = 71) with a hospital mortality of 39.3% (n = 33).

Morbidity

Major complications after ECMO are summarized in Supplementary Material, Table S8. 248 (65.9%) patients who required PC-ECMO had 1 or more major complication. A total of 129 patients (34.1%) required re-exploration, with the most common site being the mediastinum. Patients who required re-exploration for bleeding did not have higher hospital mortality (38.0% versus 34.8%, P = 0.43). Twenty-two (5.8%) patients developed intracranial haemorrhages on ECMO, while 25 (6.6%) patients had a stroke. Patients who developed intracranial haemorrhage on PC-ECMO had a statistically significant higher mortality (P = 0.022) (Table 4). A total of 122 (32.5%) patients who required PC-ECMO required renal replacement therapy, primarily peritoneal dialysis. The requirement for renal replacement therapy was a significant predictor of hospital mortality (52.0% vs 28.4%, P < 0.001) (Table 4). Thirty-six (9.6%) patients developed NEC whilst on ECMO, with 5 (1.3%) patients undergoing a laparotomy. Patients who developed NEC on ECMO did not have a higher hospital mortality (47.2% vs 34.8%, P = 0.41). Patients undergoing a laparotomy on ECMO for NEC had in-hospital mortality of 80.0% (4 of 5 patients) vs 41.9% for those in whom the NEC was medically managed (15 of 36 patients).

DISCUSSION

Our study shows that ∼3% of patients undergoing paediatric cardiac surgery in Australia and New Zealand in the current era require PC-ECMO. The incidence of PC-ECMO in the literature ranges from 0.5% to 15% and is influenced by myriad of factors, including surgical era, biventricular versus single-ventricle circulation and local experience [2]. The most common diagnosis in our experience was hypoplastic left heart syndrome. Patients who required PC-ECMO had high 30-day and hospital mortality, especially those who underwent a Norwood procedure. Patients undergoing the Norwood operation or isolated mBTS were at the greatest risk of requiring PC-ECMO.

Most PC-ECMO recipients in this study had a STAT mortality category of 4 or 5 (59.8%). Interestingly, there were still patients in STAT categories 1 (3.8%) and 2 (18.8%) who required PC-ECMO In all these patients there were generally other non-cardiac or patient factors that contributed to an increased perioperative risk such as major non-cardiac anomalies, low birthweight, and preoperative acute cardiac failure or cardiac arrest.

The majority of patients who required PC-ECMO in our experience were placed on veno-arterial ECMO. This is not unexpected in the context of failure to wean from CPB which was the commonest indication for PC-ECMO in our study (45.5%) and given that 88.8% of patients were cannulated within 48 h of their index operation. The transmediastinal route was the most common strategy for cannulation, which we presume was selected because all these patients had had a median sternotomy. However, this decision was at the discretion of the operation surgery and the rationale behind the decision could not be reliably ascertained from the medical records given the retrospective nature of the study.

The indication for PC-ECMO was not a predictor of outcomes in our study. The distinction between PC-ECMO for failure to wean from CPB and for post-operative LCOS has been variably reported in the literature. Chaturvedi et al. [6], in a two-centre experience of children with biventricular hearts demonstrated a higher survival (64% vs 29%) in those commenced on ECMO in the operating theatre than in the intensive care unit (ICU). However, similar to our study, reports from Aharon et al. [7], Huang et al. [10], Alsoufi et al. [8] and Morris et al. [17] did not identify a difference in mortality based on the indication for PC-ECMO [7, 8, 10, 17]. There was, however, a trend towards lower hospital mortality amongst patients who were cannulated for PC-ECMO either in the operating theatre during their initial operation or in the first 48 h of their postoperative recovery in our study. Hospital mortality was 32.7% amongst those patients cannulated in the operating theatre prior to returning to the ICU, compared to 65.0% in patients cannulated after 7 days postoperatively. Seventy-five percent of patients cannulated for ECMO >7 days after surgery had ECMO instituted during resuscitation (extracorporeal CPR). On multivariable Cox regression, extracorporeal CPR approached but did not reach statistical significance as a predictor of hospital mortality.

The role of left heart decompression in paediatric ECMO in patients with a biventricular heart remains unclear. The theoretical benefit of left heart venting in these patients physiologically relates to the prevention of left ventricular distension, which may improve myocardial perfusion and reduce pulmonary congestion. Left heart decompression was performed in 38 of 212 patients (15%) with a biventricular circulation and was not predictive of lower hospital mortality. The decision to vent the left heart was at the discretion of the operating surgeon and the rationale behind the decision could not be reliably ascertained from the medical records given the retrospective nature of the study. Sperotto et al. [19] published a propensity-weighted analysis from the ELSO registry in 2022, which suggested that left atrial decompression was associated with reduced hospital mortality [19]. Interestingly, patients who received a left heart vent in the ELSO registry had longer myocardial ischaemia times and longer duration of preoperative ventilation. This may reflect an inherent selection bias in which treating teams preferentially vented the left heart in certain populations of patients.

Residual haemodynamic lesions are an established determinant of survival in patients requiring PC-ECMO [18]. Our group has previously published our experience with cardiac catheterization on PC-ECMO and recommended that every attempt be made to identify and rectify residual lesions on ECMO [21]. Amongst the 22% of patients who underwent revision of their index surgical repair whilst on ECMO in our study, survival to decannulation and hospital mortality were similar to the overall study cohort. While we cannot be sure that these patients would not have survived without correction of the residual defect, it appears that reoperating patients on ECMO does not result in a higher mortality.

86% of patients were successfully weaned from PC-ECMO in our study. The majority of patients were weaned via reduction of circuit flows, with smaller cohorts of patients being weaned using insertion of an arterio-venous bridge or via a pump-controlled retrograde trial off ECMO (PCTRO). In our unit, we prefer to use PCRTO and our experience with this technique has been previously reported [22].

There is an important distinction between our experience in Australia and New Zealand and that of many previous reports from North American centres regarding the use of long-term mechanical circulatory support and heart transplantation in patients who require ECMO after paediatric cardiac surgery. Only 2 patients in our series were transitioned to long-term VAD, and none underwent transplantation while on PC-ECMO. Australia and New Zealand have a dominantly public health system for paediatric cardiac surgery. Long-term VAD support (either as bridge to recovery or transplant) is only funded at the Royal Children’s Hospital, Melbourne in Australia and at the Starship Children’s Hospital, Auckland in New Zealand. Thus, access to durable MCS programs is comparatively restrictive (patients weighing < 5 kg are generally ineligible and patients between 5-10kg are reviewed on a case-by-case basis). Furthermore, organ availability is also limited, particularly for neonates and infants. This is particularly relevant given the high representation of neonates and infants in our study cohort. Alsoufi et al. reported improved survival amongst patients undergoing PC-ECMO with heart transplantation used as an exit strategy, similar to the report from the ELSO registry [4, 8]. One in 6 patients received a heart transplant while on PC-ECMO with a survival to decannulation in 61% of patients. Interestingly, survival to decannulation was higher in our experience (85.6%) which could reflect differences in patient selection, particularly amongst high-risk neonates.

Mortality in our study was lower than reported in a recent multicentre retrospective study from the ELSO registry, which described a hospital mortality of 55% [4]. Other single-centre reports have also demonstrated similarly higher hospital mortality [8, 10, 17, 18]. On multivariable analysis, the only factors predictive of hospital mortality in our study were single-ventricle status, intracranial haemorrhage and chylothorax. This is different to the findings reported from the ELSO registry which suggested that younger age at surgery, longer ECMO duration and greater surgical complexity was associated with higher mortality [4].

In our study, patients with a single-ventricle circulation had a higher 30-day and hospital mortality than those with a biventricular circulation, which is similar to previous reports [12–15]. Single-ventricle status was also a significant predictor of hospital mortality on multivariable analysis. For patients undergoing the Norwood procedure the high hospital mortality (48.3% hospital mortality after PC-ECMO) is comparable to previous reports. Pizarro et al. [12] reported 50% survival to discharge after stage 1 palliation in patients who required PC-ECMO and Sperotto et al. [4] reported a hospital mortality of 58% from the ELSO Registry in a similar group of patients. Contributing factors to the higher mortality after PC-ECMO in this patient cohort could relate to the complexity of balancing the shunted physiology on ECMO, underlying physiological fragility of the heart in the early post-stage 1 recovery phase and the decision whether to occlude the shunt delivering pulmonary blood flow.

Overall, two-thirds of patients who required PC-ECMO had a major postoperative complication in our study. One-third of patients required re-exploration for bleeding. However, reoperation for bleeding on PC-ECMO was not associated with higher mortality. Our findings regarding the prognostic significance of renal failure for hospital mortality on ECMO are consistent with previous reports [4, 8]. Patients who required renal replacement therapy for renal failure on ECMO had a 2.7 times higher probability of hospital mortality in our series.

Nearly 10% of patients developed NEC whilst on ECMO in our study. 30-day and particularly hospital mortality was higher amongst patients who developed NEC. Four of the 5 patients who required a laparotomy for NEC whilst on ECMO did not survive indicative of the severity of this complication. We believe that the incidence of NEC in this patient cohort is likely to be underestimated, as the diagnostic criteria of NEC are varied and the diagnosis itself can be challenging in the setting of a postoperative patient on ECMO [20].

Similar to previous reports, patients undergoing a second run on ECMO after paediatric cardiac surgery fared poorly in our study [4, 8]. A second run on ECMO conferred a greater than four-fold increase in the risk of hospital mortality, with two-thirds of the patients dying in hospital. Alsoufi et al. [8] reported only 1 survivor amongst 18 patients undergoing a second run on ECMO.

Strengths and limitations

Our study represents one of the largest cohorts of patients who required PC-ECMO after paediatric cardiac surgery in the current era. Our study carries all the drawbacks of a retrospective study. Our findings are subject to selection bias, as decisions regarding candidacy for PC-ECMO are inherently at the discretion of the treating team and differences in survival could be related to institutional and individual surgeon thresholds for the institution of PC-ECMO. Consequently, it is possible that not all patients who required ECMO were placed on ECMO and that some patients who were placed on ECMO did not need ECMO. Furthermore, patients undergoing paediatric cardiac surgery are a heterogenous population and the influence of institutional and surgeon preference on the conduct and technical aspects of the operation cannot be ascertained. Finally, postoperative ICU strategies are different in different institutions with implications for the timing, management and outcomes of patients placed on PC-ECMO.

CONCLUSIONS

In the current era, one-third of patients who require PC-ECMO after paediatric cardiac surgery in Australia and New Zealand did not survive to hospital discharge The Norwood procedure and isolated mBTS had the highest incidence of requiring PC-ECMO. Patients undergoing the Norwood procedure had the highest mortality (48%). Two-thirds of patients on PC-ECMO developed a major complication. Single-ventricle status and intracranial haemorrhage predicted hospital mortality in patients who required PC-ECMO.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

ACKNOWLEDGEMENTS

The collaborators in the ANZCORS Collaborative are as follows: David Andrews (Perth Children’s Hospital, Perth, Australia); John Artrip (Perth Children’s Hospital, Perth, Australia); Johann Brink (Starship Children’s Hospital, Auckland, New Zealand); Christian Brizard (Royal Children’s Hospital, Melbourne, Australia); Ben Davies (Royal Children’s Hospital, Melbourne, Australia); Kirsten Finucane (Starship Children’s Hospital, Auckland, New Zealand); Janelle Johnson (Queensland Children’s Hospital, Brisbane, Australia); Matt Liava’a (The Children’s Hospital at Westmead, Sydney, Australia); Ian Nicholson (The Children’s Hospital at Westmead, Sydney, Australia); and Aditya Patukale (Perth Children’s Hospital, Perth, Australia).

Conflict of interest: none declared.

DATA AVAILABILITY

The data underlying this article cannot be shared publicly due to privacy restrictions and as a condition of ethical approval. The data will be shared on reasonable request to the corresponding author.

Author contributions

Lachlan Crawford: Data curation; Formal analysis; Methodology; Writing—original draft; Writing—review & editing. Supreet P. Marathe: Conceptualization; Data curation; Methodology; Supervision; Writing—review & editing. Kim S. Betts: Formal analysis. Tom R. Karl: Writing—review & editing. Adrian Mattke: Writing—review & editing. Sarfaraz Rahiman: Writing—review & editing. Isobella Campbell: Data curation. Takamichi Inoue: Data curation. Harikrishnan Nair: Data curation. Ajay Iyengar: Data curation; Resources; Writing—review & editing. Igor E. Konstantinov: Data curation; Resources; Writing—review & editing. Prem Venugopal: Conceptualization; Data curation; Methodology; Resources; Supervision; Validation; Writing—review & editing. Nelson Alphonso: Conceptualization; Data curation; Methodology; Project administration; Resources; Supervision; Validation; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Katrien Francois, Aditya K. Kaza and Awais Ashfaq for their contribution to the peer review process of this article.

Presented at the 37th Annual Meeting of the European Association for Cardio-Thoracic Surgery, Vienna, Austria, 6 October 2023.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• ANZCORS

  Australia and New Zealand Congenital Outcomes Registry for Surgery

 
• CPB

  Cardiopulmonary bypass

 
• CPR

  Cardiopulmonary resuscitation

 
• ECMO

  Extracorporeal membrane oxygenation

 
• ELSO

  Extracorporeal Life Support Organisation

 
• ICU

  Intensive care unit

 
• IQR

  Interquartile range

 
• LCOS

  Low cardiac output state

 
• mBTS

  Modified Blalock–Taussig shunt

 
• NEC

  Necrotizing enterocolitis

 
• PC-ECMO

  Post-cardiotomy extracorporeal membrane oxygenation

 
• STAT

  Society for Thoracic Surgeons-European Association for Cardio-Thoracic Surgery

Author notes