Delayed Cord Clamping in Preterm Infants: Is it Time to Become Standard Practice?
Manuscript Citation: Tarnow-Mordi W, Morris J, Kirby A, et al. Delayed Cord Clamping in Preterm Infants. New England Journal of Medicine 2017; 377 (25): 2445-2455. PMID: 29081267
Elizabeth Sager MD1,2, Joseph Hagan ScD1, Shweta Parmekar MD1,2
1Texas Children's Newborn Center, Texas Children's Hospital, Houston, TX
2Department of Pediatrics, Section of Neonatology, Baylor College of Medicine, Houston, TX
Elizabeth Sager, MD
Neonatology Fellow, Baylor College of Medicine
Texas Children's Hospital
6621 Fannin St. WT 6-104
Houston, TX 77030
The authors declare no conflicts of interest.
Keywords: immediate cord clamping, delayed cord clamping, composite outcome
Abbreviations: CLD- chronic lung disease, IVH- intraventricular hemorrhage, LOS- late onset sepsis, NEC- necrotizing enterocolitis, PMA- postmenstrual age, ROP- retinopathy of prematurity
Type of Investigation: Prevention
Do preterm infants (< 30 weeks), who had delayed cord clamping have decreased risk of mortality or major morbidity (severe brain injury, late-onset sepsis (LOS), severe retinopathy of prematurity (ROP), and necrotizing enterocolitis (NEC)) compared to preterm infants who had immediate cord clamping, by 36 weeks postmenstrual age (PMA)?
Design: Multicenter, multi-national, randomized controlled trial
Allocation: Infants were randomly assigned to group centrally when delivery became imminent with the use of an interactive voice-response system. There was stratification according to gestational age (< 27 weeks versus ≥ 27 weeks), center and multiple birth status.
Blinding: Unblinded (patients, clinicians) due to the nature of the intervention.
Follow-up period: All infants were followed until 36 weeks PMA. Follow-up in childhood is planned.
Setting: 14 in Australia, 5 in New Zealand, 1 in Canada, 1 in France, 2 in North Ireland, 1 in Pakistan, and 2 centers in the United States participated for a total of 25 centers.
Patients: 1,566 neonates, average 28±2 weeks gestational age, 65.7% born by cesarean delivery and 24.9% were of multiple births. 782 neonates were randomly allocated to immediate cord clamping group and 784 neonates were allocated to delayed cord clamping group with an average birth weight of 1000 ±269 g. Inclusion criteria: Delivery prior to 30 weeks. Exclusion criteria: fetal hemolytic disease, hydrops fetalis, twin-twin transfusion, genetic syndromes, and potentially lethal malformations.
Intervention: Infants were randomized to immediate cord clamping, defined as clamping within 10 seconds after delivery, or delayed cord clamping, defined as clamping 60 seconds or more after delivery.
Outcomes: The primary outcome was a composite of death or major morbidity, excluding chronic lung disease (CLD), and prior to an amendment, a composite of death or major morbidity, including CLD. The decision to amend the primary outcome was made because more infants than expected met the criteria for chronic lung disease. Other major morbidities included severe brain injury of grade 3 or 4 intraventricular hemorrhage (IVH), late cerebral abnormality assessed by ultrasonography, length of stay (LOS), necrotizing enterocolitis (NEC), or severe retinopathy of prematurity (ROP). The secondary outcomes were death, severe brain injury, a composite of death or severe brain injury, and for infants who survived to 36 weeks PMA, late cerebral abnormality of ultrasonography, grade 3 or 4 IVH, LOS, NEC, severe ROP, CLD, and patent ductus arteriosus requiring treatment.
Statistical Analysis and Sample Size: A sample size of 1600 infants was estimated to achieve 90% power to detect a difference of 30% vs. 22% incidence in the primary outcome in the immediate-clamping versus delayed-clamping groups, respectively. Data were analyzed according to the intention-to-treat principle.
Patient follow-up: Of the1634 infants randomized, 64 were excluded due to delivery after 30 weeks, still birth or consent withdrawal. Of those included in the intention-to-treat population, 69 infants had incomplete data for > 1 component of the primary outcome and were therefore not included in the primary analysis. The rate of adherence was 73.2% in the delayed cord clamping group and 94.9% in the immediate cord clamping group. The primary reasons for nonadherence in delayed and immediate cord clamping were clinical concern about the infant and implementation issues, respectively.
There was no difference in the composite outcomes of death or major morbidity, with inclusion or exclusion of CLD between the two groups (p= 0.96 with CLD and 0.45 without CLD). There was a significantly decreased rate of mortality in infants who had delayed cord clamping (p=0.03) before post hoc adjustment for multiple outcomes, however the difference was not significant after adjustment (p=0.39). There were no significant differences in the other secondary outcomes.
In this randomized controlled trial, delayed cord clamping did not result in a lower incidence of the a priori specified primary outcome of death or major morbidity by 36 weeks PMA compared to immediate cord clamping.
Delayed cord clamping allows for the transfer of residual placental blood and stem cells to the baby. It is estimated in term infants that by 1 minute of life, over 80% of the blood is redistributed to the infant, which helps to increase both the fetal hemoglobin oxygen content and cardiac output . Additionally, delayed cord clamping may be a more physiologic transition, as it provides time for the infant to aerate the lungs and increase pulmonary blood flow before venous return from the placental circulation is lost . In fact, previous studies have shown harm when clamping occurred before spontaneous respirations .
In 2012, a Cochrane meta-analysis evaluated over 35 short-term and long-term effects of early versus delayed cord clamping in infants born at < 37 weeks . The analysis concluded that there were insufficient data to support reliable conclusions about the effect on death or neurosensory disability at age 2-3 years, grade 3 or 4 IVH, or periventricular leukomalacia. Delayed cord clamping for 30-120 seconds was associated a decreased need for transfusion and rate of total IVH and improved circulatory stability. However, there were several limitations of the Cochrane analysis including small sample sizes and 23 of the 35 outcomes were reported by three or fewer trials.
Since then, larger studies have been conducted to target more premature infants than those included in the Cochrane analysis. A recent systemic review and meta-analysis of 531 infants born at < 32 weeks showed improved blood pressures at 4 hours, decreased rates of transfusions, total IVH, sepsis, and mortality, and a higher initial hematocrit with delayed cord clamping . As the body of evidence supporting delayed umbilical cord clamping increased, many committees published statements endorsing the practice of delayed umbilical cord clamping ranging from 30 seconds to 180 seconds, for vigorous preterm infants, including the American College of Obstetricians and Gynecologists, American College of Nurse-Midwives, World Health Organization, and the American Academy of Pediatrics . Despite these recommendations for vigorous preterm infants, majority of hospitals in the United States lack delayed umbilical cord clamping policies, primarily due to concerns about delayed resuscitation .
This study was unable to demonstrate a difference in the composite outcome of death or major morbidity in preterm infants. However, there was decreased mortality by 36 weeks PMA for those who received delayed cord clamping (relative risk 0.29, 95% CI 0.49 to 0.97, P=0.03). The importance of this finding is strengthened in the context of other randomized controlled trials comparing delayed versus immediate cord clamping in preterm infants. Fogarty et al. performed a meta-analysis which combined this trial's results with those of 17 other relevant, randomized controlled trials, totaling 2,834 infants. Delayed cord clamping significantly reduced hospital mortality compared to early clamping in preterm infants (relative risk 0.69, 95% CI 0.52-0.91, p=0.009). The reduced hospital mortality remained significant in a subgroup analysis restricted to premature infants < 29 weeks (relative risk 0.70, 95% CI 0.51-0.95, p=0.02). Furthermore, a sensitivity analysis showed that 18 additional studies of the average size showing no significant difference between delayed cord clamping compared to immediate cord clamping would be needed to create a non-significant effect for hospital mortality (p>0.05) .
Given these results, why are providers hesitant to implement delayed cord clamping policies? Fear of adverse consequences for the ill neonate. Such fear is exemplified in this study by the lower rate of protocol adherence in the delayed cord clamping group. The primary reason for nonadherence was clinical concern for the infant, which occurred most commonly in the more premature infant and those with lower Apgar scores. A systematically lower protocol adherence for these infants introduces the possibility of sample bias. Multiple studies have evaluated the safety of delayed cord clamping and have showed that delayed cord clamping had no effect on maternal hemorrhage or blood transfusion, low Apgar scores at 1 or 5 minutes, need for neonatal resuscitation or endotracheal intubation at delivery, or temperature on admission to NICU [4,5,8]. Immediate resuscitation may not be necessary when the umbilical cord is intact. Studies comparing ventilation during delayed cord clamping to delayed cord clamping alone demonstrate no difference in mean time to establish a breath, delivery room interventions or early hemodynamics . Other limitations to implementing delayed cord clamping policies include determination of the optimal duration of delayed cord clamping and the role of umbilical cord milking as an alternative to delayed cord clamping.
This large multi-centered randomized controlled trial in conjunction with the meta-analysis demonstrates that delayed cord clamping has substantial benefits. Healthcare providers should consider implementing policies to promote delayed cord clamping, especially in stable preterm infants.
EBM Lesson: Adjustment for Multiple Comparisons -
In this study, the investigators chose to analyze the secondary outcomes by applying an adjustment for multiple comparisons (the Benjamani-Hochberg procedure) because comparisons of 13 different secondary outcomes were made. This adjustment increased the p-value for mortality by a factor of 13, changing the results from statistically significant to non-significant. The Benjamani-Hochberg procedure is a method to decrease the Type I error rate (i.e., false discovery rate) by adjusting the p-value for the number of multiple comparisons being made. The adjustment does not alter the reality of whether or not delayed clamping reduces mortality. Rather, the interpretation of the data analysis results is what was altered by the a priori decision to treat death as one of many secondary outcomes. This decision yielded Tarnow-Mordi et al.'s interpretation that "the difference [in mortality] was not significant (P=0.39)," as opposed to the interpretation of the reported unadjusted p-value that would have corresponded to an a priori decision to make death the primary study outcome (p=0.03).
Adjustment for multiple comparisons is used to control the Type I error rate. In other words, the adjustment is used to reduce the probability of finding a statistically significant difference in any outcome when looking across all comparisons, if there is no such difference in reality. Unfortunately, this reduction in the Type I error rate comes with a corresponding increase in the probability of making a Type II error (i.e., failing to find statistical significance for a difference that really exists). While the theoretical basis for routine adjustment for multiple comparisons is sound, in practice it may preclude important discoveries . So, when are adjustments for multiple comparisons appropriate? Ideally adjustment for multiple comparisons can serve a purpose in confirmatory studies when results of multiple hypothesis tests are needed to reach a definitive conclusion, but care should be taken to examine and not automatically discard non-significant findings after adjustment for multiple comparisons.
Acknowledgement: The Journal Club is a collaboration between the American Academy of Pediatrics - Section of Neonatal Perinatal Medicine and the International Society for Evidence-Based Neonatology (EBNEO.org).
- Yao AC, Moinian M, Lind J. Distribution of blood between infant and placenta after birth. The Lancet. 1969; 7626: 871-873.
- Katheria AC, Lakshminrusimha S, Rabe H, McAdam R, Mercer JS. Placental transfusion: a review. Journal of Perinatology. 2017; 37(2): 105-111.
- Niermeyer S. Physiologic approach to cord clamping: Clinical issues. Maternal Health, Neonatology and Perinatology. 2015; 1 (21): 1-13.
- Rabe H, Diaz-Rossello JL, Duley L, Dowswell T. Effect of timing of umbilical cord clamping and other strategies to influence placental transfusion at preterm birth on maternal and infant outcomes. Cochrane Database of Systematic Reviews. 2012; 8: CD003248
- Backes, CH, Rivera BK, Haque U, Bridge JA, Smith CV, et al. Placental transfusion strategies in very preterm neonates: a systematic review and meta-analysis. Obstetrics and Gynecology. 2014; 124(1): 47-56.
- Committee on Obstetricians and Gynecologist. Committee opinion no.543: timing of umbilical cord clamping after birht. OBstetrics and Gynecology. 2012; 120: 1522-1526.
- Jelin AC, Kupperman M, Erickson K, Clyman R, Schulkin J Obstetricians' attitudes and beliefs regarding umbilical cord clamping. Journal of Maternal- Fetal Neonatal Medicine. 2014; 27(14): 1457-61.
- Fogarty M, Osborn DA, Askie L, Seidler AL, Hunter K, et al. Delayed vs early umbilical cord clamping for preterm infants: a systematic review and meta-analysis. American Journal Obstetrics & Gynecology. 2018; 218(1): 1-18.
- Katheria AC, Brown MK, Rich W, Arnell K. Providing a Placental Transfusion in Newborns Who Need Resuscitation. Frontiers Pediatriacs. 2017; 5: 1-7.
- Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology. 1990; 1 (1): 43-6.