WHAT IS ALREADY KNOWN ON THIS TOPIC

3:1 Compression:ventilation (C:V) ratio is recommended during neonatal cardiopulmonary resuscitation (CPR).

The optimal compression and ventilation approach is unknown.

Continuous chest compression (CC) superimposed by a sustained inflation (SI) significantly reduced time to return of spontaneous circulation (ROSC) in animal studies and a pilot trial in preterm infants <32 weeks’ gestation compared with 3:1 C:V.

WHAT THIS STUDY ADDS

There was no statistical significant difference in time to ROSC between continuous CC superimposed by an SI and 3:1 C:V.

There was no statistical significant difference in survival between continuous CC superimposed by an SI and 3:1 C:V.

The trial suffered inherent difficulties as several institutional review boards postponed their approval until the first interim safety analysis and obtaining clinical trials insurance for a trial addressing neonatal CPR was nearly impossible.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICYIntroduction

About 0.1% of term and up to 15% of preterm infants receive chest compression (CC) in the delivery room (DR),1 which is associated with high mortality and short- and long-term neurological morbidity.2 3

Neonatal resuscitation guidelines recommend a 3:1 compression:ventilation (C:V) ratio based on expert opinion, consensus and extrapolation from animal data, rather than strong scientific evidence from clinical studies.1 4 Animal studies have compared C:V ratios (eg, 2:1, 4:1, 9:3 or 15:2) or continuous CC with asynchronised ventilation5–12; however, neither resulted in shorter time to return of spontaneous circulation (ROSC) or reduced mortality. Over the last decade, we have examined continuous CC superimposed by a high distending pressure (termed sustained inflation, SI). When CC+SI was compared with 3:1 C:V in a post-transitional piglet model, regional and systemic haemodynamic and respiratory parameters were improved, with significantly reduced time to ROSC and mortality.13 14 In a pilot trial in infants <32 weeks’ gestation, mean time to ROSC with CC+SI was significantly reduced.15 With these encouraging results, a larger trial comparing CC+SI with 3:1 C:V during neonatal resuscitation in asphyxiated infants at birth was warranted.16 We hypothesised that in newborn infants requiring cardiopulmonary resuscitation (CPR) in the DR, CC+SI compared with 3:1 C:V will decrease time to ROSC.

MethodsStudy design and participants

The SURV1VE-Trial was an international, multicenter, prospective, cluster cross-over randomised controlled trial in asphyxiated infants at birth conducted at four level III neonatal intensive care units (NICUs) (two each in Austria and Canada). Research ethics committees approved the trial (Edmonton: #Pro00066739, Graz: #30-368ex17/18, Vienna: #1750/2018, Halifax: #1024910). All sites had approval for deferred consent; some (Edmonton, Vienna) had a waiver of consent if the infant died in the DR. Research staff approached parents after intervention for written informed consent for data collection. The protocol appears in online supplemental file 1, the statistical analysis plan in online supplemental file 2. The trial is reported according to Consort 2010 statement: extension to cluster randomised trials,17 and Neonatal Utstein Style (online supplemental file 3).18

Participants

The initial protocol included term and preterm infants16; however, after the Sustained Aeration of Infant Lungs (SAIL) trial,19 inclusion criteria were changed to infants >28 weeks’ requiring CC in the DR. Exclusion criteria were congenital abnormalities with adverse effect on breathing or ventilation (eg, congenital diaphragmatic hernia), congenital heart disease requiring intervention, or if parents did not provide consent.

Randomisation and cross-over

Hospitals were randomised 1:1 (computer-generated allocation sequence) to CC+SI or 3:1 C:V for Period 1 (12 months), and then crossed over to the other intervention during Period 2 (12 months). A 2-month washout period occurred after Period 1, to allow for personnel retraining and assessment of adherence to the new intervention.

Blinding

Owing to the nature of the intervention, only the statistician was blinded by reporting data as groups A and B until the data were locked.

Sample size and power calculation

Primary outcome was time to ROSC. We hypothesised that time of CC to ROSC would be reduced with CC+SI compared with 3:1 C:V. A review of data from Edmonton over a 2-year period identified a mean time to ROSC of 420 s. We calculated a sample size of 208 infants (104/group) would detect a 33% reduction (282 vs 420 s) in time to ROSC using Cox proportional hazards regression with 80% power and a two-tailed alpha error of 0.05. To account for cluster randomisation, the sample size was multiplied by design effect of 1.045 (intracluster correlation coefficient, from pilot data), with a total sample size of 218 infants (109/group).

Intervention

CC was performed using the two-thumb encircling hands technique at 90/min, using a compression depth of 1/3 of the anterior–posterior chest diameter.20–22 CPR quality metrics were not assessed, as there is currently no available technology for newborn infants.

Study interventions3:1 C:V group

In the 3:1 C:V group, infants received CC at 90/min and ventilations at 30/min in a 3:1 C:V ratio with heart rate (HR) re-evaluated every 60 s.20–22

CC+SI group

In the CC+SI group, infants received SIs with a peak inflation pressure (PIP) of 25–30 cmH2O during continuous CC at 90/min. SI was delivered over 20 s, followed by a positive end expiratory pressure (PEEP) of 5–8 cmH2O for 1 s. Then the next 20 s SI was started while CCs continued, then PEEP for 1 s, then another SI for 20 s. After 3×20 s CC+SI (a total of 60 s), HR was assessed. If HR remained <60/min, CC+SI was continued in 60-second intervals (3×20 s CC+SI) with HR assessment every 60 s.

Revert to standard of care rule for CC+SI

If CPR was ongoing for 5 min with CC+SI, the clinical team converted to 3:1 C:V ratio.

OutcomesPrimary outcome

The primary outcome was time to ROSC, defined as duration of CC until HR increased >60/min, determined by auscultation, which was maintained for 60 s.

Secondary outcomes

Secondary outcomes among others included neonatal mortality (death <28 days), brain injury rates (reported on MRI or head ultrasound), DR interventions, therapeutic hypothermia, pneumothorax, infection/sepsis, intraventricular haemorrhages (IVHs) and bronchopulmonary dysplasia.

Data collection and statistical analysis

Study data were collected and managed using REDCap electronic data capture tools hosted at the University of Alberta.23 24 A study specific DR record was used to record all interventions including duration of CC. Data were analysed on an intention-to treat basis and included all randomised participants. A survival analysis was done to analyse the difference in time to ROSC between groups. To account for cluster randomisation, Cox proportional hazards regression with time to ROSC as an outcome and allocation group as an independent variable was created. Centres were entered as clusters in the model and statistical significance of allocation group variable was evaluated. Clinical characteristics and outcomes were compared using Student’s t-test for parametric and Mann-Whitney U-test for non-parametric comparisons of continuous variables, and χ2 for categorical variables. For safety and selected secondary outcomes, OR and 95% CIs were estimated. Data are presented as mean (SD) for normally distributed continuous variables and median (IQR) when the distribution was skewed. All p-values are two-sided and p<0.05 was considered statistically significant. Statistical analysis was performed using SAS V.9.4 (SAS Institute) or later.

Data and Safety Monitoring Board

Prior to trial commencement, a Data and Safety Monitoring Board (DSMB) was appointed, consisting of two neonatologists with resuscitation expertise. A priori stopping rules were established and interim safety analyses were planned at 10%, 25% and 50% of enrolment. Stopping rules included: (1) 25% increase in mortality, pneumothorax, or IVH with CC+SI and (2) Bayesian posterior probability of CC+SI being better <0.5 or >0.98. If this probability was <0.5, the trial would be stopped for futility. If the posterior probability was >0.98, the DSMB could consider stopping for superiority.

Results

Patient recruitment occurred between 19 October 2017 and 22 September 2022 and initially included infants born between 23 and 42 weeks’ gestation. However, when the steering committee became aware of the SAIL-trial results,19 the SURV1VE-trial was temporarily stopped, to adjust inclusion criteria in discussion between steering committee and DSMB. The trial was restarted with adjusted inclusion criteria (≥28 weeks’ gestation) and finally stopped due to funding constraints. At stoppage, 12 infants in the CC+SI group and 15 infants in the 3:1 group (figure 1) were included (Edmonton n=14, Graz n=5, Halifax n=2, Vienna n=4.) Parents of one infant in both groups declined consent. The demographics of both study groups are presented in table 1.

Figure 1

CONSORT diagram: screening, eligibility and randomisation.

Primary outcome

All 11 infants in the CC+SI group achieved ROSC, while 2/14 infants in the 3:1 C:V group did not achieve ROSC and died in the DR. The CC times of these two infants were 13 min and 33 min. Two infants in the CC+SI group had CC ongoing for >5 min and were switched to 3:1 C:V per protocol, but analysed within the CC+SI group. Three infants (one infant with CC+SI and two infants with 3:1 C:V) achieved ROSC, but had a second episode of CC, before achieving ROSC and admission to the NICU.

The median (IQR) time to ROSC with CC+SI was (174–780) s with 3:1 C:V (p=0.0502, log rank, and p=0.16, cox proportional hazards regression). Figure 2 shows the Kaplan-Meier curve of ROSC; in both groups, the two infants who did not achieve ROSC were censored at the end of their CC time only for the Kaplan-Meier curve.

Figure 2

Kaplan-Meier graph of time to return of spontaneous circulation. CC+SI, chest compression+sustained inflation; 3:1 C:V, 3:1 compression:ventilation ratio; ROSC, return of spontaneous circulation.

Secondary outcomes

Exploratory secondary outcomes included DR interventions (table 2), NICU outcomes (table 3) and initial neurological examination at 1–6 hours after birth (online supplemental table 4). Twelve MRI exams were performed (six per group), which showed injuries related to hypoxic ischaemic encephalopathy (HIE). No infant was diagnosed with infection/sepsis, necrotising enterocolitis, bronchopulmonary dysplasia or retinopathy of prematurity.

Table 2

Delivery room interventions

Table 3

Outcomes during NICU administration

Safety outcomes

The main safety outcome was mortality, which was 2/11 (18%) with CC+SI versus 8/14 (57%) with 3:1 C:V, p=0.10 (Fisher’s exact test) and OR (95% CI) 0.17 (0.03 to 1.07). In the CC+SI group, two infants died after NICU admission (care withdrawal for severe HIE (n=1) and severe IVH (n=1)). In the 3:1 C:V group, two infants died in the DR and six further died after NICU admission (care withdrawal for severe HIE (n=4), acute myocardial ischemia (n=1) and cardiorespiratory failure (n=1)). Care was withdrawn between 0 and 8 days after birth in both groups.

Additional safety outcomes included IVH and air leaks. IVH occurred with CC+SI (1/11, Grade 4 IVH) and with 3:1 C:V (2/14, 1× Grade 4 and 1× Grade 3) (OR (95% CI) 0.60 (0.05 to 7.63)). Pneumothorax was diagnosed in 1/11 with CC+SI, and 2/14 (one pneumothorax and one pneumomediastinum) with 3:1 C:V, neither of which required treatment (OR (95% CI) 0.60 (0.05 to 7.63)). There were no tension pneumothoraces.

Discussion

Annually, ~3 million infants worldwide receive CPR with devastating outcomes including mortality and severe neurological disabilities. The approach of 3:1 C:V ratio during neonatal CPR is based on expert consensus rather than clinical evidence.1 4 Over the last decade, we have shown that CC+SI significantly improved haemodynamic and respiratory parameters, resulting in significantly reduced time to ROSC and mortality.13–15 In this phase-2 cluster trial, we compared CC+SI with 3:1 C:V in infants >28 weeks requiring CC in the DR. The results can be summarised as follows: CC+SI reduced time to ROSC and improved survival, although both were not statistically significant, and CC+SI was not associated with adverse events.

To achieve ROSC, healthcare professionals must provide high quality CC to optimise cardiac output while providing adequate ventilation to maintain lung aeration to optimise oxygen delivery. Antegrade blood flow during CPR is achieved by direct cardiac compression or by increasing intrathoracic pressure.25 While continuous CC without rescue breaths improves ROSC and survival after adult cardiac arrest,26 in newborns cardiac arrest is due to asphyxia with insufficiently aerated lungs, therefore a combination of ventilation and CC is necessary for ROSC. Chandra et al used a high PIP (60 cmH2O) during continuous CC in an animal model and reported improved carotid flow, without compromising oxygenation.27 Similarly, Sobotka et al demonstrated that SI given immediately after birth increases intrathoracic pressure without impeding blood flow.28 CC+SI combines these interventions, resulting in improved pulmonary artery and carotid artery blood flow.13 14

Furthermore, the downward force applied during CC results in forced expiration while chest recoil air re-enters the lung. Tsui et al applied a relatively low downward force (0.16 kg) onto the chest of anaesthetised infants and reported that the tidal volumes forced out of the lungs were greater than the dead space, while the amount of air entering during recoil with a PEEP of 5 cmH2O was minimal.29 Similarly, in neonatal piglets, Li et al reported that because the air forced out during CPR is larger than the air delivered, a loss of functional residual capacity (FRC) results, which could lead to atelectasis, reduced oxygen delivery and delayed ROSC. In comparison during CC+SI, the constant high distending pressure prevents decruitment, thereby maintaining FRC,30 improving oxygen delivery and reducing time to ROSC.

In post-transitional piglets with asphyxia cardiac arrest, CC+SI resulted in improved pulmonary and carotid blood flow, improved minute ventilation from passive ventilation, and significantly reduced time to ROSC.13 Another animal study in post-transitional piglets with asphyxia cardiac arrest also reported reduced time to ROSC.31 This was confirmed in a pilot trial comparing CC+SI with 3:1 C:V in infants <32 weeks’ gestation: The time to ROSC was 31 (9) s with CC+SI compared with 138 (72) s with 3:1 (p=0.011).15 In the current trial, the median (IQR) time to ROSC was not statistically significant different between CC+SI with 90 (60–270) s and 615 (174–780) s with 3:1 C:V. Furthermore, in animal studies CC+SI improved survival (7/8 (87.5%) vs 3/8 (37.5%),(p=0.038)13); in this trial, survival was 82% with CC+SI and 43% with 3:1 C:V, not reaching statistical significance. The survival with 3:1 C:V was not different to neonatal registries, which reported survival between 34% and 60%.2 32 Overall, enrolled patients number was too low; therefore, no conclusion should be drawn nor should CC+SI be used outside of research settings.

Any DR trial faces the uncertainty of patient recruitment with either antenatal/deferred consent or waiver of consent, and uncertainty of when a potential eligible newborn will be delivered.33 Clinical trials examining CC have additional challenges. The rate of neonatal CPR is ~1–3/1000 births in high-risk delivery centres,34 which suggests that many sites have three to five CC events annually. Given the low number of CC events, many sites withdrew their commitment before the trial started due to lack of potential infants and the fear that the healthcare professionals would forget about the study intervention.

While the trial was approved by four institutional review boards (IRBs), other IRBs unexpectedly postponed their approval until the first interim safety analysis. The DSMB had no safety concerns at the first interim analysis, however, as the trial was closed a few months later, we do not know if these IRBs would have approved the study.

In Canada, all clinical trials need clinical trials insurance to comply with Guideline for Good Clinical Practice regarding Investigator/Institution insurance against trial related claims. While there is an existing clinical trials insurance agreement between Canadian universities, insurance had to be purchased for international sites. Obtaining insurance for a neonatal CPR trial was nearly impossible. Risk management at the University of Alberta negotiated for 3.5 years to secure clinical trial insurance for other sites. This duration is unacceptable and raises the question if a similar delay would occur if the trial examined CC in children or adults.

Limitations

The trial was stopped early, due to funding constraints and recruitment delays due to insurance and the COVID-19 pandemic; thus the proposed sample size was not obtained. Sustained inflation has come under scrutiny in preterm infants 23–26 weeks’ gestation,19 which led to the trial being paused to adjust the inclusion criteria to infants >28 weeks’ gestation. Recruitment over 5 years was low (n=27, 25 enrolled and 2 declined) reflecting the limited available patients.

Conclusion

There were no statistical differences in time to ROSC and mortality between CC+SI and 3:1 C:V. The number of enrolled patients was too low, therefore no conclusion should be drawn nor should CC:SI be used outside of research settings. A larger randomised trial comparing CC+SI with 3:1 C:V is warranted.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. Data used to generate the results reported in this study will be made available following publication to researchers who provide a methodologically sound proposal. Data will only be made available if approval is granted from the Human Research Ethics Committee Board, University of Alberta, Edmonton, Canada. Furthermore, all requesters will need to sign a data transfer agreement. Requests should be directed to the corresponding author.

Ethics statementsPatient consent for publication

Consent obtained from parent(s)/guardian(s).

Ethics approval

The research ethics committees approved the trial (Edmonton: #Pro00066739, Graz: #30-368 ex17/18, Vienna: #1750/2018, Halifax: #1024910). Participants gave informed consent to participate in the study before taking part.

Acknowledgments

We thank the parents, who agreed for their infants to take part in the trial and the public for donating money to our funding agencies.

We thank the Independent Data Monitoring and Safety Committee: Myra Wyckoff (professor of neonatology, UT Southwestern Medical Center, Dallas, USA) and Neil Finer (emeritus professor).

We would like to thank the Research Coordinators at Edmonton (Sylvia van Os and Caroline Fray), Halifax (Tara Hatfield) and Graz (Bernhard Schwaberger, Lukas Mileder, Nariae Baik-Schneditz and Christina Wolfsberger).

The study guarantor (GMS) affirms that the manuscript is an honest, accurate and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as originally planned (and, if relevant, registered) have been explained.