Cord Management at Birth for Preterm Infants (NLS # 5051) TF SR

Conflict of Interest Declaration

The ILCOR Continuous Evidence Evaluation process is guided by a rigorous ILCOR Conflict of Interest policy. Members of the review group declared the following conflicts of interest:

Walid El-Naggar:

• member of the iCOMP collaborative group
• received NICHD grant as a co-investigator of the Umbilical Cord Milking in Non-Vigorous Infants (MiNVI Trial)
• received a grant from IWK Research as the principal investigator of the MoCC trial
• received a grant from Nova Scotia Health Research Foundation (NSHRF) as a principal investigator of the study: The effect of umbilical cord milking on hemodynamic status of preterm infants: a randomized controlled trial
• received a grant from National Health and Medical Research Council (NHMRC) as a co-investigator of the Australian Placental Transfusion study (APTS)

Peter Davis

• member of the iCOMP collaborative group
• received Australian NHMRC funding for BabyDUCC trial

Justin Josephsen

• member of the iCOMP collaborative group
• published an UCM trial that was considered for inclusion in iCOMP,
• received NICHD funding as co-investigator of the VentFirst trial.

Lene Seidler

• lead of iCOMP collaborative group
• received a grant from the Australian National Health and Medical Research Council (NHMRC) to support iCOMP.

Daniela Costa-Nobre, Tetsuya Isayama, Roger Soll and Keith Couper have no relevant COI.

Helen Liley (Task Force Chair)

• member of the iCOMP panel of expert advisors
• Site investigator for the APTS study which was included in iCOMP analyses
• Received a grant from the Australian National Blood Authority to study mediators of red cell transfusions in the APTS study.

On the basis of the above conflicts: Gary Weiner (Task Force Deputy Chair) made decisions about recusal on the basis of conflict of interest and determined that no review group members needed to be recused. However, review group members who had no COI undertook completion of the ILCOR checklist for decisions to utilise results of the iCOMP analyses for development of Task Force Consensus on the Science of Resuscitation and Treatment Recommendations.

CoSTR Citation

El-Naggar W, Davis PG, Josephsen J, Seidler L, Soll R, Costa-Nobre D, Isayama T, Couper K, Schmölzer G, Weiner G, Liley HG on behalf of the Neonatal Life Support Task Force* [Internet] Brussels, Belgium: International Liaison Committee on Resuscitation (ILCOR) Advanced Life Support Task Force, 2023 December xx. Available from: http://ilcor.org

* Task Force members (in addition to Drs El-Naggar, Davis, Josephsen, Costa-Nobre, S, Madar, Schmölzer, Weiner and Liley); Dawson JA, de Almeida MF, Fabres G, Fawke J, Foglia E, Guinsburg R, Kawakami M, Lee H, Madar J, McKinlay C, Siguira T, Solevåg AL, Rüdiger M, Trevisanuto D, Wyckoff MH

Methodological Preamble (and Link to Published Systematic Review)

The ILCOR Adolopment process has been used to guide the adaptation of the newly conducted systematic review and network meta-analysis with individual participant data (IPD) on cord management at preterm birth (iCOMP study). {Seidler 2023 2209} and {Seidler 2023 2223}

The iCOMP collaboration conducted the only IPD meta-analyses of cord-management of preterm infants to date. It includes two meta-analyses:

• Deferred cord clamping, cord milking, and immediate cord clamping at preterm birth: a systematic review and individual participant data meta-analysis {Seidler, 2023 2209)
• Short, medium, and long deferral of umbilical cord clamping compared with umbilical cord milking and immediate clamping at preterm birth: a systematic review and network meta-analysis with individual participant data {Seidler, 2023 2223}

Pairwise meta-analysis (meta-analysis of original trial parallel group comparisons) using individual participant data rather than trial level data, allows for optimal assessment of data integrity, risk of bias and adjusted treatment effects, including the influence of baseline and population characteristics (subgroup analyses). The IPD network meta-analysis (NMA) estimates comparative effectiveness of different cord management strategies using both direct and indirect evidence. Combining IPD with NMA can improve precision, increase information, and reduce bias. Combining the results of IPD NWM with the pairwise IPD meta-analysis can help define the best cord management strategy for preterm infants at birth. The IPD pairwise and network meta-analyses were assessed by a member of our content expert team, who was not involved in the iCOMP study, using the AMSTAR2 critical appraisal tool. {Shea 2017 j4008} The pooled analysis was deemed to be of high quality.

The Task Force notes that ILCOR systematic reviews typically exclude unpublished studies (e.g. conference abstracts, trial protocols), while the iCOMP analysis includes such studies. However, we note the statement by the authors of the iCOMP papers; “We conducted extensive data processing, quality, and integrity checks of all included data”. {Seidler 2023 2209} The Task Force concluded that these steps ensured a sufficient level of integrity that is not usually available for unpublished data and that the reduction in publication bias from including unpublished studies was advantageous. {Hunter 2022 e068791}

Cord Management at Birth for Preterm Infants PICOST

A- FOR IPD PAIRWISE META-ANALYSIS:

Population:

Preterm infants born at <37⁺⁰ weeks’ gestation and their mothers.

Interventions:

1- Deferred (delayed/later) cord clamping (DCC)

2- Umbilical cord milking (UCM)

Comparisons:

1- Immediate (early) cord clamping (ICC)- compared to each of the above interventions.

2- Between-intervention comparisons (i.e. DCC vs. UCM).

Definitions used in PROSPERO submission:

Deferred cord clamping (DCC)- defined as cord clamping without milking, with or without respiratory support at >15 seconds after birth.

Umbilical cord milking (UCM)- defined as either a- repeated compression of the cord from the placental side toward the infant with the connection to the placenta intact immediately after birth (intact-cord milking) or b- drainage of a long segment of the cord by compression from the cut end toward the infant with or without respiratory support after clamping and cutting of the long segment after birth (cut-cord milking).

Immediate cord clamping (ICC)- defined as cord clamping, without milking or initiation of respiratory support, at ≤15 seconds after birth of the infant or as defined by trialist.

Outcomes:

A) Short-term:

• 9 (Critical) Infant’s mortality before hospital discharge (primary)
• 7 (Critical) Infant’s inpatient morbidities (e.g., intraventricular hemorrhage (IVH), necrotizing enterocolitis, retinopathy of prematurity, bronchopulmonary dysplasia) for preterm infants <32 weeks’ gestation
• 9 (Critical) Maternal mortality
• 7 (Critical) Maternal complications (post-partum hemorrhage and infection)
• 6 (Important):
  • Resuscitation and stabilization interventions (e.g. receiving positive pressure ventilation ± intubation ± chest compressions ± medications);
  • Blood transfusion
  • Hematologic and cardiovascular status (in-hospital)
  • Hyperbilirubinemia treated with phototherapy

B) Long-term:

• 8 (Critical) Neurodevelopmental outcomes (with age-appropriate validated tools)
• 6 (Important) Hematologic status (in infancy)

Study Designs:

The pairwise meta-analysis with individual participant data (IPD) on cord management at preterm birth (iCOMP study) included randomized trials comparing umbilical cord management strategies. Trials with missing data, integrity issues and those not fitting intervention categories were excluded. Cluster- and quasi-randomized trials were also excluded.

Timeframe:

All years and all languages were included. Medical databases including MEDLINE, Embase and CENTRAL, and clinical trial registries, including ClinicalTrials.gov were originally searched up to Feb 2022, and WHO ICTRP up to March 2022. The search was updated on June 6th, 2023 and no additional eligible studies were identified.

A priori subgroups examined:

Subgroup analyses were conducted for the primary outcome of death (prior to hospital discharge), and two key secondary outcomes (intraventricular hemorrhage (IVH) any grade and any receipt of blood transfusion). Subgroup analyses were conducted by including a participant-level or study-level (population characteristics) interaction term in the mixed-effects logistic regression model. The influence of the subgroup characteristic on treatment effect is expressed as OR with 95% CI for the interaction term.

Subgroups were defined according to:

• Gestation at birth by week

• Type of pregnancy (singleton, multiple)

• Mode of birth: (vaginal, cesarean)

• Resuscitation provided at bedside with cord intact

• Planned position of the baby relative to the placenta

• Onset of labor (e.g. spontaneous onset or spontaneous pre-labor ruptured membranes; not

spontaneous onset or spontaneous pre-labor ruptured membranes)

• Perinatal mortality rate (for country)

• Study year

Details of the planned analyses can be found in the study protocol {Seidler 2020 e034595}and in more detail in the time-stamped statistical analysis plan.

PROSPERO registration:

The iCOMP review was registered with PROSPERO CRD42019136640 in 2019.

Selecting outcomes

The choices of outcomes using the GRADE classifications of “critical, important, or less important” were agreed upon by consensus of the Task Force and were similar to those used in previous reviews. {Strand F328} Highest priority was given to the critical outcomes of mortality, neurodevelopmental impairment, severe intraventricular hemorrhage, morbidities (including necrotizing enterocolitis, bronchopulmonary dysplasia, and retinopathy of prematurity), and maternal mortality. Additional analyses were done to address the important outcomes of resuscitation and hematologic and cardiovascular status.

Potential subgroups were defined a priori.

Although, long-term neurodevelopmental and hematologic outcomes were identified by the Task Force to be critical and important, they were not available in the current iCOMP study.

Some of the outcomes reported in this document have not been published elsewhere and/or have not undergone independent peer review. These post-hoc analyses are identified in the text.

Risk of Bias:

All data were checked, harmonised, re-coded, and assessed for risk of bias following prespecified criteria. Risk of bias was assessed for all studies using Cochrane criteria adapted for IPD. A separate risk of bias assessment was conducted for death before discharge, delivery room, and post-delivery room outcomes since levels of blinding and missing data varied across these outcomes. Integrity and data quality were assessed following a pre-specified checklist collated from previous tools, including items such as ethics approvals, incomplete data, implausible values, and publication retraction notices or expressions of concern. {Seidler 2023 2209}

The majority of studies were at low risk of selection bias. All included studies were assessed by the iCOMP Collaborative to be at high risk of performance bias due to the nature of the intervention. While risk of bias was high for delivery room outcomes due to lack of blinding of caregivers, it was low for most studies for other outcomes. Most studies were at low risk of attrition bias. Evidence Profile Tables were collated for primary and key secondary outcomes applying the GRADE framework. These include details on risk of bias. {Seidler 2023 2209}

Bias Assessments: NLS 5051 Bias assessments Reprinted from The Lancet, vol. 402 Seidler, A. L., Aberoumand, M., Hunter, K. E., Barba, A., Libesman, S. et al. pages 2209-2222 Copyright 2023, with permission from Elsevier.

GRADE evaluations:

Certainty of evidence was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach for the primary and key secondary outcomes, and an adaptation of the GRADE approach for subgroup analyses. All of these assessments were performed in duplicate, with conflicts resolved by a third reviewer in the iCOMP team.

Post-hoc GRADE evaluations:

Because three comparisons were included (deferred cord clamping vs immediate cord clamping, umbilical cord milking vs immediate cord clamping, umbilical cord milking vs deferred cord clamping) the number of outcomes was very high. A methodological decision was made to omit the GRADE evaluations of some comparisons for the iCOMP publication. However, the NLS Task Force decided that a GRADE evaluation of some additional comparisons might influence the strength of our final recommendation and therefore these were performed post-hoc. This applies to the following outcomes for IPD analysis (both for infants <32 weeks’ gestation and for those ≥32 weeks’ gestation) of Comparison 1; DCC compared to ICC for;

• hemoglobin concentrations (g/dL) and hematocrit values (%) within the first 24 hours after birth,
• body temperature on admission
• and for the maternal outcome of postpartum infection

It also applied to the following outcomes (both for infants <32 weeks’ gestation and for those ≥32 weeks’ gestation) for Comparison 2; UCM compared to ICC:

• hemoglobin concentrations (g/dL) and hematocrit values (%) within the first 24 hours after birth,
• body temperature on admission
• receiving inotropic support for hypotension

It also applied to the following outcomes (both for infants <32 weeks’ gestation and for those ≥32 weeks’ gestation) for Comparison 3; UCM compared to ICC:

• hemoglobin concentrations (g/dL) and hematocrit values (%) within the first 24 hours after birth

Consensus on Science (1)NLS 5051 Consensus on Science 1

Consensus on Science (2)NLS 5051 Consensus on Science 2

Treatment Recommendations

A- In preterm infants born at less than 37 weeks’ gestational age who are deemed not to require immediate resuscitation at birth, we recommend deferring clamping of the umbilical cord for at least 60 seconds. (Strong recommendation, high-certainty evidence).

Justification: In making a strong recommendation for deferring cord clamping for at least 60 seconds in preterm infants <37 weeks’ gestation who are deemed not to require resuscitation at birth, the NLS Task Force took into consideration the following points:

1. There is high-certainty evidence from 20 trials including 3260 infants for reduced mortality after deferred cord clamping compared to immediate cord clamping (OR 0.68, 95% CI 0.51 to 0.91) with the number needed to treat for benefit (NNTB) 40, 95% CI 143 to 26.

The reduction in mortality was robust across several participant-level and trial-level subgroups (including gestational age at birth, mode of birth, multiple birth, sex, trial year, and perinatal mortality rate).

1. The evidence is derived from individual patient data (IPD) rather than trial level data which increases the quality of the meta-analysis. The IPD analysis included comprehensive checking of data integrity and provides more robust evidence overall and for secondary analyses.

1. The meta-analysis results were consistent across all pre-specified sensitivity analyses, including combining individual participant data with aggregate data (from trials not providing individual participant data), different outcome definitions, excluding trials with high risk of bias, and different analysis methods (e.g., two-stage model).

1. The evidence is supported by the IPD network meta-analysis.

1. There is moderate-certainty evidence of fewer blood transfusions (OR 0.59, 95% CI 0.47 to 0.73) from 13 trials including 1929 infants <32 weeks’ gestation, and of higher hemoglobin concentrations within the first 24 hours after birth (mean difference (MD)= 1.26 g/dL, 95% CI 0.72 to 1.80) from 8 trials including 523 infants after deferred cord clamping compared to immediate cord clamping. The higher hemoglobin concentration was also replicated in preterm infants ≥32 weeks’ gestation.

1. In choosing 60 seconds or more as the recommended interval for deferred cord clamping, we took into consideration that the recommendation is for infants who are deemed not to require resuscitation at birth and that most included infants (80%) in the deferred clamping arm received cord clamping after 60 seconds or more. The OR for reducing mortality after DCC≥ 60 seconds vs. immediate cord clamping = 0.63 (95% CI 0.44-0.88, p = 0.01) (post-hoc analysis).

1. Deferral of cord clamping for 120 seconds, or more (long DCC) was associated with the greatest reduction in mortality compared to immediate cord clamping in the IPD network meta-analysis (OR 0·31; 95% CrI 0·11-0·80). However, this evidence was based on 5 small trials including small numbers of extremely preterm infants (<121). Two of the trials excluded infants requiring resuscitation. {Rana 2018 655, Ranjit 2015 29}

The reported adherence to long DCC was the lowest at 67% (compared to about 80% for medium deferral and 95% for immediate clamping, milking and short deferral).

• While it is reasonable to consider this approach, the task force cannot recommend the long deferral for all infants based on this evidence. Instead, the long deferral could be considered only if there is no contraindication and if appropriate newborn stabilization can be provided on the intact cord (skilled team, proper training, appropriate equipment, enough space and ability to provide thermal management).
• More evidence is needed before recommending long DCC. Practicality, feasibility, cost-effectiveness, and equity issues need to be addressed.

1. The Task Force noted that there was moderate-certainty evidence, from 8 trials including 1995 infants <32 weeks' gestation showing more hypothermia (body temperature <36·5°C) on admission after deferred cord clamping compared to immediate cord clamping (OR 1.28, 95% CI 1.06 to 1.56) with the number need to treat to harm (NNTH) 16, 95% CI 9 to 71. There are no important differences in rates of hypothermia in infants ≥32 weeks’ gestation. Nevertheless, it is important that measures should be taken to maintain normal temperatures in all preterm infants when practising deferred cord clamping. Refer to ILCOR recommendations regarding maintaining normal immediately after birth in preterm infants. {Berg 2023 e187}

1. Parents report that deferred cord clamping provides a positive experience with the mothers feeling closer and more attached to their infants. {Bradshaw 2019 225}

B- In preterm infants born at 28⁺⁰ to 36⁺⁶ weeks’ gestational age who do not receive deferred cord clamping, we suggest umbilical cord milking as a reasonable alternative to immediate cord clamping to improve infant hematologic outcomes. Individual maternal and infant circumstances should be taken into account. (Conditional recommendation, low certainty evidence).

Justification: Despite the low-certainty evidence from 18 trials including 1565 infants that umbilical cord milking may not reduce the critical outcome of death before discharge compared to immediate cord clamping (OR 0.73, 95% CI 0.44 to 1.20), there are hematologic effects that favour umbilical cord milking compared to immediate cord clamping:

1. There is low certainty evidence from 12 trials including 944 infants showing higher hemoglobin concentrations (g/dL) within the first 24 hours after birth after umbilical cord milking compared to immediate cord clamping (MD 0.45 g/dL, 95% CI 0.17 to 0.73 g/dL) in infants <32 weeks’ gestation. The finding is replicated in infants ≥32 weeks’ gestation.

1. There is moderate certainty evidence from 15 trials including 1163 infants showing fewer red cell transfusions after UCM compared to ICC (OR 0.69, 95% CI 0.51 to 0.93; I² = 20%; NNTB 10, 95% CI 5 to 55) in infants <32 weeks’ gestation. The finding is replicated in infants ≥32 weeks’ gestation.

1. There is no evidence of increased rates of adverse effects in preterm infants <37 weeks’ gestation or their mothers after umbilical cord milking compared to immediate cord clamping.

1. There are no important differences in the critical or important outcomes after umbilical cord milking compared to deferred cord clamping in the preterm infants born at 28-36⁺⁶ weeks’ gestation.

1. The IPD meta-analyses did not distinguish between the two methods of cord milking (intact-cord and cut-cord). The intact-cord was milked (2-4 times) in the majority of trials, whilst few trials milked the cut-cord once, therefore no specific recommendations are made for each method.

C- We suggest against intact cord milking for infants born at less than 28 weeks’ gestation. (Weak recommendation; low certainty evidence). There is insufficient evidence to make a recommendation regarding cut-cord milking in this gestational age group.

Justification: In suggesting against against intact umbilical cord milking in infants <28 weeks’ gestation, the task force took into consideration that there is low-certainty evidence from 7 trials including 860 infants <32 weeks’ gestation that the critical outcome of severe intraventricular hemorrhage is increased after intact-cord milking compared to deferred cord clamping (OR 2.20, 95% CI 1.13 to 4.31), however:

1. The evidence from the IPD pairwise meta-analysis is driven by an RCT that was stopped prematurely because of increased rates of severe intraventricular hemorrhage in the pre-specified subgroup of preterm infants <28 weeks’ gestation. {Katheria 2019 1877}

1. The following report from the same RCT that compared the outcomes of umbilical cord milking and deferred cord clamping in the other pre-specified subgroup of preterm infants born at 28-32 weeks’ gestation did not find any evidence of increased severe intraventricular hemorrhage, mortality, or other clinical outcomes after umbilical cord milking compared to deferred cord clamping. {Katheria 2023 217.e1} This study was not included in the analysis as it was published after the iCOMP meta-analysis was completed and the CoSTR development process was started.

D- In preterm infants born at less than 37 weeks’ gestational age who are deemed to require immediate resuscitation at birth, there is insufficient evidence to make a recommendation with respect to cord management. (Weak recommendation; low certainty evidence).

Justification: We could not make a recommendation regarding cord management of preterm infants who are deemed to require resuscitation at birth.

1. The pairwise IPD review reported that adherence to deferred cord clamping was low (<75% in those trials reporting adherence), mostly because of the preference of health care providers to practice immediate cord clamping or cord milking when the infant was judged to require resuscitation at birth. In other studies, the adherence was not reported. These 2 factors limit the generalizability of the meta-analysis findings and limit extension of our recommendation to non-vigorous infants and those who are deemed to require resuscitation at birth.

1. Nevertheless, there is growing evidence from animal studies and feasibility studies in human infants that supports stabilization of the infant while deferring cord clamping (resuscitation with intact cord/physiologic cord clamping/baby-directed cord clamping). This approach is also supported by sound physiological principles. {Bhatt 2013 2113, Crossley 2009 4695, Duley 2018 F6, Hooper 2015 608}

1. We are awaiting the results of studies currently underway that evaluate the resuscitation /stabilization of infants with the cord intact. These studies are expected to help us define the best way to manage these infants. Questions related to the practicality, feasibility, cost-effectiveness, and equity will need to be addressed.

E- There is insufficient evidence to make recommendations on cord management for maternal, fetal, or placental conditions that were considered exclusion criteria in many studies (monochorionic multiple fetuses, congenital anomalies, placental abnormalities, alloimmunization and/or fetal anemia, fetal compromise, and maternal illness). In these situations, we suggest individualized decisions based on severity of the condition and assessment of maternal and neonatal risk. (Weak recommendation; very low certainty of evidence).

Justification: There is uncertainty regarding the optimal cord management strategy in deliveries complicated by monochorionic multiple pregnancies, infants who have major congenital abnormalities, fetal anemia, or other conditions that may impact maternal or fetal well-being at the time of birth as these conditions were largely excluded from most clinical trials. We were also unable to draw conclusions regarding optimal cord management in the setting of placental problems including abruption, incision through an anterior placenta, placenta previa, or abnormalities of placental vasculature or insertion. Until more data are available for specific situations such as these, decisions about cord management in the presence of maternal, placental, or fetal complications need to be individualized, based on severity of presentation and clinical assessment of risk to the mother or baby.

F- Whenever circumstances allow, the plan for umbilical cord management should be discussed between maternity and neonatal providers and parents before delivery, and should take into account individual maternal and infant circumstances. (Good practice point)

Knowledge Gaps:

• There are insufficient data on long-term neurodevelopment outcomes, or other post-discharge outcomes following different cord management strategies. We expect long-term data to become available from some trials currently underway.
• There are insufficient data on optimising cord management as a public health strategy to improve child health and development.
• There are insufficient data for cord management of preterm infants who are judged to require immediate resuscitation.
• There are insufficient data for cord management of preterm infants born under specific conditions, including fetal congenital anomalies, placental abnormalities, monochorionic multiple gestation, alloimmunization and/or fetal anemia, fetal compromise, maternal general anesthesia, and maternal illness.

• Further evaluation of measures to prevent hypothermia during deferred cord clamping is required.

• The optimal duration of deferred cord clamping remains uncertain. It is unclear whether it should vary with different maternal or fetal conditions.
• There are few studies of cut-cord milking as a management strategy.
• The impact of cord management on vertical transmission of infectious diseases is uncertain.
• There is a need for widely agreed nomenclature and definition of different interventions including “delayed”, “deferred”, “later”, “optimal”, and “physiologic” cord clamping, as well as “milking”, “stripping”, “intact-cord”, and “cut-cord”.

EtD: NLS 5051 Preterm Cord Management Et D table

References: NLS 5051 References