Introduction

The first minutes of life and delivery room management of newborns have a fundamental impact on neonatal mortality and morbidity. Establishing interventions and protocols for resuscitation in the delivery room poses a challenge due to the heterogeneous population as term infants differ from preterm infants who were born at the limit of viability. Recommendations and guidelines on stabilization procedures in the delivery room are constantly changing and evolving by integrating new procedures based on recent evidence (e.g., application of sustained inflations and continuous positive airway pressure (CPAP)).1,2,3

Over the past decades, a number of research studies have examined the physiology and outcomes of delayed cord clamping (DCC); thereby, its numerous beneficial effects on infants have been proven, e.g., improved transitional circulation and iron stores, increased hemoglobin level at birth, decreased need for blood transfusion and lower incidence of intraventricular hemorrhage (IVH) and necrotizing enterocolitis (NEC).4,5,6

Therefore, on the basis of the guidelines influenced by the International Liaison Committee on Resuscitation (ILCOR) recommendations, it is suggested to delay cord clamping (CC) by at least 60 s, ideally after ensuring adequate lung aeration.7 However, the timing of CC and the steps of resuscitation are still not synchronized. Unfortunately, non-vigorous and non-breathing infants needing immediate interventions for stabilization are usually clamped immediately and excluded from most studies on different umbilical cord management. In consequence, for preterm and term infants who require resuscitation after birth, we still have insufficient evidence on the optimal time of CC.4,8,9 The purpose of our study was to collect all available data on this most vulnerable newborn population in order to discover the effect of intact cords when initiating resuscitation.

Methods

Our systematic review and meta-analysis is reported based on the recommendations of PRISMA 2020 guideline10 (Supplementary Table 1), and as methodological guidance, the Cochrane Handbook11 was followed. The prestudy protocol was registered in advance on the International Prospective Register of Systematic Reviews with registration number CRD42022370338. A deviation from the protocol occurred as we also conducted subgroup analyses based on the usage of special resuscitation trolleys. Ethical approval was not required due to the inherent design of the systematic review.

Eligibility criteria

To address our research question, we included randomized clinical trials (RCTs) and observational studies comparing the initiation of neonatal resuscitation (airway opening maneuvers, positive pressure ventilation, chest compression, etc. except drying and stimulation only) before (intact cord resuscitation (ICR)) and after CC. Conference abstracts and case reports were excluded. Exclusion criteria for the examined population were monochorionic twins, triplets or higher-order multiple pregnancies, major congenital malformations, fetal hydrops, twin-to-twin transfusion syndrome, placental abruption, and placenta previa.

We defined our primary outcomes in advance as in-hospital mortality; presence of IVH (all grades and severe (≥grade 3)), periventricular leukomalacia, and cerebral palsy. Our secondary outcomes included delivery room parameters, early complications of prematurity and maternal outcomes, etc. (see Supplementary Material, sections 2 and 3 and Supplementary Table 2).

Search strategy and selection process

Our systematic search was conducted in four main databases: MEDLINE (via PubMed), Embase, CENTRAL (the Cochrane Central Register of Controlled Trials), and Web of Science on March 1, 2024, using a predefined search key (see in Supplementary Material, section 4). During the search, no filters or language restrictions were applied. Reference and citation lists of the included studies were examined for further eligible articles using the citationchaser.12 After duplicates were removed both automatically and manually, two independent review authors (G.Sz.M. and V.U.) performed the selection process separately via Rayyan (Rayyan Systems, Cambridge, Massachusetts, USA; Qatar Computing Research Institute, Doha, Qatar)13 and Endnote 20 (Clarivate Analytics, Philadelphia, Pennsylvania, USA)14 reference manager programs. Publications were screened according to the eligibility criteria by title and abstract first and then by full text. Disagreements were resolved by involving the corresponding author (Á.G.).

Data were collected from the eligible articles by two authors (G.Sz.M. and V.U.) independently, using a standardized data collection sheet which was created based on the consensus of clinical and methodological experts. The following data were extracted: title, first author, year of publication, countries, number of centers, study period, DOI (digital object identifier), study design, study population, patient demographics, inclusion and exclusion criteria, interventions, and outcomes measured.

Study risk of bias (RoB) assessment

The RoB assessment was performed by two authors (G.Sz.M. and V.U.) separately based on the recommendations of the Cochrane Collaboration, using the Cochrane risk-of-bias tool for randomized trials (RoB2) (Cochrane Bias Methods Group, Cochrane Collaboration).15 The corresponding author (Á.G.) resolved any occurring disagreements.

Synthesis methods

As we assumed considerable between-study heterogeneity in all cases, a random-effects model was used to pool effect sizes.

For binary outcomes, risk ratios (RRs) with a 95% confidence interval (CI) were used for the effect size measure. To calculate the study RRs and the pooled RRs, the total number of patients and those with an event of interest in each group were separately extracted from the studies. We reported the results as the risk of an event of interest in the experimental group versus the risk of an event of interest in the control group. For continuous outcomes, differences between the means (MD) with 95% CI were used for effect size measure. To calculate the study MDs and pooled MDs, the sample size, the mean, and the corresponding standard deviation (SD) were extracted from each study. For the Apgar score, the quartiles were reported in most cases (instead of mean and SD). Therefore, differences between group medians (MedD) were used as effect size measures with 95% CI as recommended by McGrath et al. 16 We reported the results as an experimental group minus control group values.

Results were considered statistically significant if the pooled CI did not contain the null value. We summarized the findings of the meta-analysis in forest plots. As the study number was small, we did not report the prediction intervals (i.e., the expected range of effects of future studies) of results. Between-study heterogeneity was also described by Higgins&Thompson’s I2 statistics.17

Subgroup analysis was performed based on gestational age (GA), usage of special resuscitation trolleys, and the type of intervention in the control group.

Small-study publication bias was assessed by visual inspection of funnel plots and calculating Harbord (modified Egger’s) test p-value18 for RR effect size and classical Egger’s test p-value19 for MD effect size. Unfortunately, the number of studies was too low; therefore, these assessments were meaningless and were not reported.

Potential outlier publications were planned to be explored using different influence measures and plots following the recommendations of Harrer et al. 20. However, the study number was limited; therefore, it was ineffectual to conclude and report.

All statistical analyses were calculated by R software (v4.3.0; R Development Core Team)21 using the meta (v6.5.0)22 package for basic meta-analysis calculations and plots, and dmetar (v0.0.9000)23 package for additional influential analysis calculations and plots.

For additional details see the Supplementary Material, section 5.

Assessing the level of evidence

To evaluate the quality of evidence, we followed the recommendations of the “Grades of Recommendation, Assessment, Development, and Evaluation (GRADE)” workgroup.24

Results

Systematic search and selection

The systematic search yielded 17,141 articles. Following duplicate removal and selection processes, we identified 11 eligible studies and an additional one by reference and citation search (Supplementary Fig. 1). In a subset of the qualified papers,25,26,27,28,29 the study population included infants for whom it was either uncertain or unnecessary to administer any form of resuscitation following birth. After contacting the corresponding authors for more detailed data, we included one more study29 in the analysis. In one article,30 we successfully obtained the necessary information needed to determine the total sample size for in-hospital mortality. In conclusion, a total of six RCTs29,30,31,32,33,34 were included in our quantitative analysis, and ten RCTs25,26,27,28,29,30,31,32,33,34, and two observational studies35,36 in the systematic review. Details and results of the studies included in the systematic review are summarized in Supplementary Table 3.

Study characteristics

The baseline characteristics of studies included in the meta-analysis can be found in Table 1, and the detailed interventions of the examined groups and the respiratory status at the time of CC are summarized in Table 2. We analyzed six RCTs including 610 preterm29,31,32,33,34 and term30,34 infants. The timing of CC and the rates of different resuscitation procedures were heterogeneous among the studies. In the intervention group, newborns received DCC at various time points (either at 50 s,32 at 60 s,29,31 at least 180 s30,34) or physiological-based cord clamping (PBCC).33 In the control group, immediate/early cord clamping (ICC/ECC) (<60 s)29,30,34 or DCC (>30–60 s)31,32,33 were performed. Therefore, besides the GA, we also performed subgroup analyses based on the type of intervention in the control group and the usage of a special resuscitation trolley (see in the Supplementary Material, Figs. 16–31)

Table 1 Baseline characteristics of studies included in the meta-analysis.
Full size table
Table 2 Detailed interventions and respiratory status of the infants at the time of CC in the examined groups.
Full size table

In-hospital mortality

Analysis of in-hospital mortality included five RCTs30,31,32,33,34 and involved 584 patients (Fig. 1). This event occurred in 11 out of 297 patients assigned to the ICR group and 13 out of 287 patients assigned to the CC prior to resuscitation group, which indicates no significant difference between the examined groups (RR 0.89, 95% CI [0.24, 3.36]).

Fig. 1: Forest plot representing the risk ratio of in-hospital mortality in infants who received ICR or CC prior to resuscitation after birth.
figure 1

RR risk ratio, 95% CI 95% confidence interval, CC cord clamping.

Full size image

No significant differences were found between the groups either in our subgroup analyses (Supplementary Figs. 2–4), or in any studies included in our systematic review and examined this outcome25,28,35,36 (Supplementary Table 3).

Delivery room parameters (oxygen saturation level, Apgar score) and temperature at NICU admission

The pooled analysis of three RCTs30,32,34 with 392 patients showed higher mean SpO2 at 5 min (MD 6.67%, 95% CI [−1.16%, 14.50%]) and 10 min (MD 2.87%, 95% CI [−5.53%, 11.28%]) after birth in the ICR group; however, the difference was not statistically significant (Fig. 2a, b).

Fig. 2: Forest plot representing the mean difference of SpO2 after birth in infants who received ICR or CC prior to resuscitation after birth.
figure 2

a Forest plot representing the mean difference of SpO2 at 5 min after birth in infants who received ICR or CC prior to resuscitation after birth. b Forest plot representing the mean difference of SpO2 at 10 min after birth in infants who received ICR or CC prior to resuscitation after birth. SpO2 oxygen saturation level by pulse oximetry, min minutes, MD mean difference, SD standard deviation, 95% CI 95% confidence interval, CC cord clamping.

Full size image

We found no significant difference between the examined groups in terms of Apgar score at 1 min and 5 min after birth (1 min MedD –0.09, 95% CI [−0.55, 0.36]29,31,32,33,34 and 5 min MedD –0.03, 95% CI [−0.36, 0.29]31,32,33,34) (Supplementary Figs. 5 and 6) and temperature at admission to the NICU (MD −0.04 °C, 95% CI [−0.20 °C, 0.12 °C]) (Fig. 3).

Fig. 3: Forest plot representing the mean difference of temperature at admission to the NICU in infants who received ICR or CC prior to resuscitation after birth.
figure 3

NICU neonatal intensive care unit, MD mean difference, SD standard deviation, 95% CI 95% confidence interval, CC cord clamping.

Full size image

In the articles included in our systematic review that examined Apgar scores,26,27,28,35,36 SpO2 shortly after birth27,28 and temperature at NICU admission,25,28,35 no significant difference was found between the intervention and control groups (Supplementary Table 3).

In terms of heart rate (HR), Badurdeen et al. 26 observed that PBCC resulted in a similar mean HR between 60 s to 120 s after birth compared to infants receiving ECC. Andersson et al. 30 reported significantly lower HR values in the ICR group than in the ECC group at 1 min and 5 min after birth. Raina et al. 34 also found significantly lower HRs in the ICR group than in the ECC resuscitation group at 5 and 10 min. In the study of Hoeller et al. 36, the PBCC group had lower HRs during the first 72 h of life than those who underwent standard DCC, reaching significance by 10 h of monitoring (Supplementary Table 3).

Early complications of prematurity (IVH, NEC, ROP, BPD, PDA, and LOS)

For the analysis of all grades of IVH, we had three articles31,32,33 involving 300 patients, and 55 infants with IVH (Fig. 4a). The overall effect size was RR of 1.25, 95% CI [0.77, 2.00]. Regarding severe ( ≥ grade 3) IVH, we had four studies29,31,32,33 covering 326 patients with an event number of 21 patients (Fig. 4b). The rate of severe IVH showed no significant difference between the groups (RR 0.96, 95% CI [0.30, 3.01]), although we found a RR of 0.75, 95% CI [0.06, 10.11] when we examined only those studies that used specialized resuscitation trolleys (Supplementary Fig. 7).

Fig. 4: Forest plot representing the risk ratio of IVH in infants who received ICR or CC prior to resuscitation after birth.
figure 4

a Forest plot representing the risk ratio of all grades of IVH in infants who received ICR or CC prior to resuscitation after birth. b Forest plot representing the risk ratio of severe IVH (≥grade 3) in infants who received ICR or CC prior to resuscitation after birth. IVH intraventricular hemorrhage, RR risk ratio, 95% CI 95% confidence interval, CC cord clamping.

Full size image

No significant differences were found between the intervention and control groups in articles included in our systematic review that examined all grades25,36 and severe25,28 IVH, with the exception of the article by Hocq et al. 35 who reported that the incidence of all grades of IVH decreased during the ICR implementation period (Supplementary Table 3).

The analysis of four studies29,31,32,33 involving 326 patients and 14 outcome events resulted in an RR of 2.05, 95% CI [0.34, 12.30] for NEC ≥ stage 2, although it did not reach significance level (Fig. 5). When we analyzed only studies that compared ICR to DCC prior to resuscitation,31,32,33 the RR for NEC ≥ stage 2 was 2.89, 95% CI [0.51, 16.40]; however, this result was not statistically significant (Supplementary Fig. 8). Analyzing studies29,31,33 where newborn stabilization was performed with a special resuscitation trolley, we found an RR of 1.22, 95% CI [0.12, 12.85] but it did not reach a significance level either (Supplementary Fig. 9).

Fig. 5: Forest plot representing the risk ratio of NEC ≥ stage 2 in infants who received ICR or CC prior to resuscitation after birth.
figure 5

NEC necrotizing enterocolitis, RR risk ratio, 95% CI 95% confidence interval, CC cord clamping.

Full size image

Hocq et al. 35 reported a lower incidence of NEC ≥ stage 2 during the ICR implementation period. Further studies examining this outcome25,28,36 did not find any significant differences between the examined groups (Supplementary Table 3).

ROP requiring treatment analysis resulted in a RR of 1.60, 95% CI [0.50, 5.13]29,31,32 (Fig. 6), and for BPD, we found a RR of 1.20, 95% CI [0.83, 1.76]29,31,32,33 (Supplementary Fig. 10), these findings did not reach significance level. In PDA requiring treatment (RR 0.88, 95% CI [0.47, 1.66])31,32,33 (Supplementary Fig. 11) and LOS (RR 0.91, 95% CI [0.44, 1.87])29,32,33 (Supplementary Fig. 12), there were no clinically or statistically significant differences between the two groups.

Fig. 6: Forest plot representing the risk ratio of ROP requiring treatment in infants who received ICR or CC prior to resuscitation after birth.
figure 6

ROP retinopathy of the prematurity, RR risk ratio, 95% CI 95% confidence interval, CC cord clamping.

Full size image

No significant differences were found between the intervention and control groups in articles included in our systematic review that examined ROP,25,28,35,36 BPD,25,35,36 PDA25,35 and LOS28,36 (Supplementary Table 3).

Need for blood transfusion and surfactant therapy in NICU

We found no significant differences between the groups in the need for blood transfusion (RR 0.95, 95% CI [0.73, 1.25])29,31,32,33 (Supplementary Fig. 13) and need for surfactant therapy (RR 0.96, 95% CI [0.75, 1.22])29,31,32,33 (Supplementary Fig. 14).

No significant differences were observed between the groups in any of the articles included in our systematic review that examined the need for blood transfusion25,35 and surfactant therapy35 (Supplementary Table 3).

Safety parameters

In our study, we included the outcomes of the need for phototherapy, hypothermia (<36.0 °C) at NICU admission, and maternal outcomes including maternal blood loss, pp. hemorrhage, and pp. infection as safety parameters.

We found no significant differences between the intervention and control groups in the need for phototherapy (RR 1.10, 95% CI [0.92, 1.30])29,32,33,34 (Supplementary Fig. 15) and studies25,26,35 included in our systematic review that examined this outcome found no differences either (Supplementary Table 3).

For hypothermia25,28,33,35 and maternal outcomes (blood loss,26,32,33 pp. hemorrhage,25,26,28,32,33 pp. infection25,26,28,32) no significant differences were found between the examined groups (Supplementary Table 3).

RoB assessment and quality of evidence

The RoB assessment is summarized in Supplementary Tables 4 and 5. Four trials29,30,31,33 had a high RoB due to deviations from the intended interventions. The level of evidence is presented in Supplementary Table 6, for most of our outcomes the GRADE assessment resulted in low certainty due to serious RoB and imprecision, but certainty had to be downgraded to very low due to serious RoB and very serious imprecision in some cases. The assessment of small study bias was meaningless as we had only a few studies.

Discussion

The present systematic review and meta-analysis examined the optimal umbilical cord management during neonatal resuscitation. According to the resuscitation guidelines influenced by the ILCOR recommendations, delaying CC by a minimum of 60 s is recommended, preferably following adequate lung aeration.7 A recent systematic review and network meta-analysis with individual participant data on preterm infants found that the highest reduction in mortality occurred when CC was deferred for at least 120 s. Furthermore, the study suggests that resuscitation with an intact cord might be beneficial, but more evidence is needed to support this practice.37 The major limitations of the first quantitative analysis examining this question conducted by Avinash et al. 38 were the limited number of studies available for inclusion and their relatively small sample sizes.

Currently, there are well-defined protocols for cases where CC precedes newborn resuscitation. In cases where initial resuscitation interventions and appropriate thermal care can be safely performed with an intact cord without compromising the newborn, CC may be delayed during these interventions.7,8 Therefore, it is essential to establish standardized procedures and equipment for ICR. However, this concept lacks robust evidence; consequently, explicit protocols and equipment for ICR have not yet been defined.

Main findings

In our study, no statistically significant differences were found in terms of in-hospital mortality, delivery room parameters, and early complications of prematurity. Intact cord management during resuscitation appears to be safe and may improve initial oxygenation, although this is in conflict with current standards for delivery room resuscitation.

In some of the examined outcomes, we observed the following findings which might be relevant in patient care.

Delivery room parameters

Although our results did not reach the statistically significant level, they suggest a possible beneficial effect of ICR on oxygenation. This finding was also noted in another study27 that was excluded from our analysis because of the population examined. In addition to CC timing, FiO2 is another critical factor in neonatal stabilization. Current international guidelines recommend starting resuscitation with 21–30% FiO2 to mitigate the potential for hyperoxia-induced tissue damage.39 All included studies reporting on this aspect30,31,32,33 used an initial FiO2 < 40%. Nevertheless, findings from both animal and human studies indicate the potential benefits of starting resuscitation with 100% FiO2.40,41,42 As further larger investigations are necessary for conclusive evidence, the ongoing DOXIE trial was conducted to directly compare the use of 30% FiO2 to 100% FiO2 during ICR.43

Due to a lack of data, we could not perform an HR analysis, which is an important delivery room parameter; however, studies reporting data on this aspect26,30,34,36 found lower HR values after birth in the ICR group which is hypothesized to be a result of the increased blood volume following ICR.36

Early complications in preterm infants

However, the results of early preterm complications did not show statistically significant differences between the groups, interestingly, the risk of NEC ≥ stage 2 and treatment-requiring ROP seemed to be higher in the ICR group. Examining studies that were excluded from the analysis because of the population or the study design, Deng et al. 28 also reported higher rates of NEC ≥ stage 2 and ROP ≥ phase 2 in the DCC + nCPAP group compared to the DCC-only group, but these results did not reach statistical significance either. In contrast, Hocq et al. 35 found a lower incidence of NEC ≥ stage 2 following ICR implementation in their hospital protocol. Free radicals and hyperoxia might play a role in the development of NEC and ROP.44,45,46

Resuscitation trolleys

To provide continuous placenta-newborn connection, while allowing an immediate stabilization of non-vigorous newborns, different resuscitation platforms were developed such as LifeStart Trolley (Inspiration HealthCare Group PLC, Croydon, UK), Concord Birth Trolley (Leiden University Medical Center, Leiden, Netherlands), NOOMA cart (Maternal Life, LLC, Palo Alto, California, USA) and INSPiRe Trolley (Integrated Neonatal Support on Placental Circulation with Resuscitation, Alberta Health Services, Edmonton, Alberta, Canada).47 Although resuscitation with an intact cord can also be achieved with standard equipment, certain challenges such as the warming device to prevent hypothermia and the availability of sufficient respiratory support may persist.48 Among the RCTs we included in our meta-analysis, ICR has performed with29,31,33,34 and without30,32 the use of specialized resuscitation trolleys as well. A subgroup analysis revealed that employing specialized trolleys might improve the impact of ICR on early complications associated with prematurity: although the rate of severe IVH did not significantly differ between the groups in the analysis of all studies included, when examining specifically those using special equipment, there appeared to be a lower risk of severe IVH in the ICR group compared to the CC prior to the resuscitation group. For NEC ≥ stage 2, although the risk was still higher in the ICR group, it was almost halved when a special trolley was used. However, conducting a subgroup analysis resulted in an even smaller sample size and none of these results reached significance.

Common concerns about DCC

For DCC, safety parameters include hypothermia, the necessity for phototherapy, maternal blood loss, and pp. infection. In our analysis, we did not observe significant differences in the need for phototherapy. Although we did not have sufficient data to perform a statistical analysis for the rest of the outcomes, studies25,26,28,32,33 reporting data on these did not find any significant difference between the groups, except Knol et al. 33 and Hocq et al. 35 who found moderate hypothermia on admission to NICU in a higher proportion of very preterms receiving ICR than the control group.

Strengths and limitations

In this systematic review and meta-analysis, we aimed to achieve the highest level of evidence available; therefore, we followed our pre-registered protocol. Studies were included only if they provided explicit information indicating that all or nearly all newborns in both arms had received resuscitation after birth. We conducted a quantitative analysis of the eligible RCTs and included all studies examining this specific question in the systematic review part. As we had a broad population (preterm and term infants) and heterogeneous interventions to examine a wide range of outcomes, we conducted subgroup analyses where feasible to mitigate the effect of these factors.

Due to the inclusion criteria and the early stage of ICR implementation in clinical practice, a limited number of studies were eligible, posing limitations to our analysis. The generalizability of our findings is challenged by the small sample size and number of events (sometimes zero). In addition, protocols of interventions and definitions of outcomes were heterogeneous or even missing in some cases. Devices used for bedside newborn stabilization can be crucial; however, different resuscitation platforms were used in the studies included. Another limitation was the presence of a moderate and high RoB in most domains. Therefore, caution is needed when interpreting our results.

Implications for clinical practice and future research

The translation of scientific findings into daily practice plays a key role, highlighting that effective implementation significantly improves the quality and cost-efficiency of healthcare. This process is essential to ensure that advancements in medical research directly benefit patient care and public health.49,50

The potential beneficial effect of ICR on oxygenation in the population of term infants suggests that recommending ICR may be justified. Nevertheless, ICR should be applied in preterm infants with caution, and we recommend that this practice be performed only in specialized centers with appropriate expertise, protocols, and equipment, given the potential for complications. Based on our results the use of special resuscitation trolleys appears to be beneficial.

There is still no clear and strong evidence for optimal cord management during neonatal resuscitation. There are ongoing multicenter RCTs51,52,53,54 to examine this question and we encourage researchers to conduct further high-quality RCTs with large sample sizes, homogeneous intervention protocols, and outcome definitions. In addition, it is essential to differentiate outcome data between infants requiring post-birth stabilization measures and those not requiring them. This approach increases the representativeness of the results.

Conclusion

Intact cord management during resuscitation appears to be safe and may improve initial oxygenation, although this is in conflict with current standards for delivery room resuscitation. The early complications of prematurity remain unclear. The use of specialized resuscitation trolleys seems promising to reduce the risk of IVH. There is an urgent need for further high-quality RCTs with larger patient numbers, especially with specialized resuscitation trolleys and physiological-based CC.