Does umbilical cord milking result in higher measures of systemic blood flow in preterm infants?

December 07, 2015


Katheria AC, Truong G, Cousins L, et al. Umbilical cord milking versus delayed cord clamping in preterm infants. Pediatrics. 2015 July;136(1):61-69; doi: 10.1542/peds.2015-0368. PMID: 26122803


Leah Yieh, MD, MPH
Neonatal-Perinatal Fellow
Oregon Health & Science University

Jamie B. Warren, MD, MPH
Assistant Professor of Pediatrics, Division of Neonatology
Oregon Health & Science University




(P) In infants < 32 weeks’ gestation born by Cesarean delivery, (I) does umbilical cord milking as (C) compared to delayed cord clamping, (O) improve systemic blood flow as measured by superior vena cava flow and right ventricular output (T) in the first 12 hours of life?


  • Design: Randomized controlled trial
  • Allocation: Infants were randomly assigned by opaque, sealed envelopes prior to delivery. Computer-generated randomization was stratified by gestational age and mode of delivery. Obstetricians were made aware of the randomization immediately before delivery. Multiples received the same assignment.
  • Blinding: Only the delivery nurse and obstetrician performing the intervention were aware of the assigned groups. Physicians, nurses, and respiratory therapists attending the delivery stayed in a resuscitation suite adjacent to the delivery room. No documentation of the intervention was made in the physician or nursing notes. Blinded echocardiograms and head ultrasounds were performed mostly by the principal investigator. Other investigators performing echocardiograms were not involved in the randomization or present for the intervention. Images were analyzed without knowledge of the assigned group by the principal investigator.
  • Follow-up period: Hemodynamic recordings, echocardiograms, and head ultrasounds were performed within the first 12 hours of life. A single study-related blood draw to measure hematocrit was done at 12 hours of life. Electrical cardiometry was used to continuously measure cardiac output via surface EKG electrodes that detect changes in thoracic impedance with the application of a low-voltage current. In addition, cerebral saturations, as measured by near-infrared spectroscopy, were monitored for the first 24 hours of life. Routine head ultrasounds were done on the second or third day of life according to unit protocol.
  • Setting: 2 tertiary care centers (Sharp Mary Birch Hospital for Women and Newborns, Loma Linda University Medical Center)
  • Patients: Entry criteria included infants with gestational age 23 0/7 to 31 6/7 weeks. Exclusion criteria included monochorionic multiples, incarcerated mothers, placenta previa, concern for abruptions, Rh sensitization, hydrops, congenital anomalies, or the obstetrician declining to perform the intervention. Delayed consent was used in this study as both interventions were considered to pose minimal risk, and antenatal consent would potentially exclude the most critically ill neonates. Therefore, parents were approached immediately after the delivery to provide written consent for continued data collection. Data was destroyed for any infants whose parents did not want to enroll in the study.
  • Intervention: Umbilical cord milking (UCM) was performed by holding the infant at or 20 cm below the level of the placenta. The cord was pinched close to the placenta and milked toward the infant over 2 seconds before releasing the cord and allowing it to refill with blood for 1-2 seconds. This stripping mechanism was repeated for a total of 4 times. Delayed cord clamping (DCC) was performed by holding the infant at or 20 cm below the level of the placenta for at least 45 seconds before clamping the cord.
  • Outcomes:
    • Primary outcome: Superior vena cava (SVC) flow within the first 12 hours of life
    • Secondary outcomes: Incidence of intraventricular hemorrhage (IVH)
  • Analysis and Sample Size: Based on an initial pilot study to determine the feasibility of the study, an initial sample size calculation determined that 40 or more infants were needed in each arm to demonstrate at least a 25% difference in SVC flow at 12 hours of life assuming a type 1 error of 0.05 and 80% power. In order to detect a change in the incidence of IVH, a second site, which could not measure hemodynamics, was added. Therefore, the sample size was increased to 600. Based on a prior study, the authors estimated the baseline incidence of IVH at 22%. Interim analysis after enrolling 197 infants showed a statistical difference in the primary outcome of SVC flow, but lower than expected incidences of IVH of 13% (DCC group) and 7% (UCM group). At these frequencies, the study would have needed at least 780 patients in each arm to detect a statistical difference in the incidence of IVH, which was estimated to prolong the study to 7 years; therefore, the trial was stopped due to challenges of recruiting enough subjects.
  • Patient follow-up: Of 342 infants assessed for eligibility, 197 were randomized (August 2013–August 2014). 154 (78%) were delivered by Cesarean section; 75 randomized to UCM and 79 to DCC. 14 infants in the DCC group were deemed too unstable by the obstetrician and underwent immediate cord clamping; 1 infant in the DCC group received UCM. 2 neonates in the UCM group underwent immediate clamping. 140 infants (70 in each group) had complete hemodynamics measured. 10 had partial or no hemodynamic data collected at site 1 due to staff limitations. Only clinical outcomes were collected at site 2.


Perinatal and Neonatal Outcomes for Infants Delivered by Cesarean Section

Umbilical cord milking (n = 75) Delayed cord clamping

(n = 79)

Gestational age (weeks) 28 +/- 2 28 +/- 2
Birth weight (grams) 1255 +/- 413 1132 +/- 392
Female 46 48
Diabetes 11 15
Chorioamnionitis 20 20
Pregnancy-induced hypertension 22 25
Narcotics prior to delivery 16 19
Duration of rupture of membranes (hours) 63 +/- 169 67 +/- 147
General anesthesia 6 2
Antenatal steroids 69 75
Antenatal magnesium 72 64
Measured outcomes
Time clamp cord (seconds)* 20 +/- 10 42 +/- 12
Delivery room temperature (Celsius) 36.8 +/- 0.4 36.6 +/- 0.4
Birth hemoglobin (g/dL)** 16.3 +/- 2.4 15.6 +/- 2.2
Urine output in first 24 hours (ml/kg/hr)** 4.42 +/- 1.3 3.99 +/- 1.2
SVC flow (ml/kg/min)** 93 +/- 24 81 +/- 29
RVO (ml/kg/min)* 261 +/- 80 216 +/- 73
Any IVH 5 10
Severe IVH (grade 3 or higher) 3 3

Data presented as mean +/- standard deviation
* p < .001
** p < .05

There were no significant demographic differences between the two groups. Infants who underwent UCM after Cesarean delivery had higher birth hemoglobin (p<.05), higher urine output in the first 24 hours of life (p<.05), higher blood pressure in the first 15 hours of life (p<.05), and higher measures of SVC flow (p<.05) and right ventricular output (RVO) (p<.001). There were no differences in left ventricular output (LVO) by echocardiography or cardiac output measured by electrical impedance. In addition, there were no differences in cerebral saturation, pulse oxygen saturation, or heart rate over 24 hours. No differences were detected between the two arms in the infants delivered vaginally. There were fewer absolute numbers of neonates with total IVH in the UCM group, but the study was not powered to detect a statistically significant difference in this outcome.


The authors conclude that UCM in preterm infants delivered by Cesarean section provides greater placental transfusion as demonstrated by higher initial hemoglobin and improved SBF compared to DCC. The authors also suggest that these findings indicate improvement in organ perfusion in the first 24 hours of life, as evidenced by both higher blood pressures and urine output during that time frame.

Visit Acta to access a pdf copy of this EBNEO commentary!


The American College of Obstetricians and Gynecologists recommends a 30-60 second delay in umbilical cord clamping for all preterm deliveries, though limited data regarding outcomes by mode of delivery exists.1 UCM is a method to auto-infuse blood into the neonate in a shorter amount of time. A meta-analysis published in the Journal of the American Medical Association analyzed 7 randomized controlled trials comparing UCM with immediate cord clamping in infants < 33 weeks’ gestation and found that those undergoing UCM had a higher hemoglobin and decreased risk for IVH and need for oxygen at 36 weeks.2 Additionally, a retrospective study of preterm infants < 30 weeks’ gestation who underwent UCM demonstrated an associated decrease in IVH, necrotizing enterocolitis, and death.3

A study in 2011 comparing UCM with DCC in preterm infants < 33 weeks’ gestation found no major clinical differences between the groups though the authors of the present study point out that the trial did not stratify outcomes by mode of delivery.4 The reviewed study (Katheria et al) is the first trial to compare UCM and DCC at Cesarean delivery and to evaluate DCC for more than 30 seconds in preterm infants. The authors hypothesized that UCM after Cesarean delivery would result in improved systemic blood flow and be associated with decreased morbidities compared with DCC.

The study was unique in its use of delayed consent in which parents were informed of the intervention and consented after the delivery. The authors argue that “antenatal consent was not practical because it would exclude the potentially sickest newborns.” In addition, they justify their use of delayed consent as UCM and DCC are “both standard practices” at the institutions involved and are of “equivalent low risk.” This method of consent, however, may raise safety and ethical concerns for some. On the other hand, in a commentary published in the same issue of Pediatrics, this “innovative approach to obtaining consent” is praised as a strategy to increase enrollment and obtain a more representative sample of high risk infants.5 Although there appears to be no immediate risks associated with DCC and UCM, there is limited data on long-term outcomes in this population.6 In regards to concerns for enrolling preterm infants in clinical trials, a retrospective study by Foglia et al examined in hospital outcomes including death, bronchopulmonary dysplasia, grade 3 or higher IVH, and severe retinopathy of prematurity for more than 5000 extremely preterm infants. The authors found no difference between those enrolled in randomized controlled trials compared to those who were eligible but not enrolled.7

The study was adequately blinded and used advanced techniques, such as electrical cardiometry, to measure neonatal hemodynamics. Though there is limited data on the routine use of electrical cardiometry in premature infants, its validity was evaluated by comparing measurements of RVO, LVO, and SVC flow with echocardiogram. The primary outcome of SVC flow was chosen as it is an important marker of neonatal transition and is not affected by fetal shunts. RVO was also noted to be lower in the DCC group, which previous studies have shown to be associated with increased oxygen requirement, severe IVH, and death.8 There was no difference in LVO, but the authors note that LVO may be confounded by a left to right shunt across a patent ductus arteriosus and thereby overestimate systemic blood flow.

The authors argue that given the markers suggestive of improved organ perfusion, UCM may stabilize fluctuations in systemic blood flow and therefore prevent IVH, although the study did not have adequate power to assess this outcome. As to whether or not these surrogate markers of perfusion translate into other clinically meaningful outcomes, such as reduced risk of respiratory distress syndrome, bronchopulmonary dysplasia, necrotizing enterocolitis, or retinopathy of prematurity, remains to be seen. Despite not being statistically significant, the different rates of IVH between the two interventions (5/75 infants who underwent UCM compared to 10/79 in the DCC group) appear to be clinically relevant. With an absolute risk difference of 6/100, it is difficult to dismiss the potential decreased risk of IVH associated with UCM. Accordingly, further investigation is warranted in future trials. Another limitation of the study is that there was lack of a control group to undergo immediate cord clamping. As there are known improved clinical outcomes associated with placental transfusion, the authors state that they did not have clinical equipoise to assign neonates to undergo immediate cord clamping; thus, assignment to such a control group would have been unethical. Lastly, the study was limited by only including 94/197 infants who were < 29 weeks’ gestational age.

 Due to a shorter amount of time required to perform UCM, it may be advantageous compared to DCC in critically ill newborns who need immediate resuscitation. It remains uncertain if and how the present study will influence clinical practice. Larger trials involving smaller, more premature neonates are needed in addition to long-term follow up to assess neurodevelopmental implications of these interventions. This randomized-controlled trial provides an optimal design and consent model upon which future, larger studies can build upon. The authors confidently end their manuscript with the statement that “UCM should no longer be considered experimental…[but] should be considered as a beneficial option for preterm infants delivered by Cesarean delivery.” With the growing body of evidence supporting the safety and use of UCM in premature newborns, we concur with the authors in advocating for this intervention.


  1. Committee Opinion No.543: Timing of umbilical cord clamping after birth. Obstet Gynecol 2012;120:1522-6.
  2. Al-Wassia H, Shah PS. Efficacy and safety of umbilical cord milking at birth: a systematic review and meta-analysis. JAMA Pediatr 2015;169:18-25.
  3. Patel S, Clark EA, Rodriguez CE, Metz TD, Abbaszadeh M, Yoder BA. Effect of umbilical cord milking on morbidity and survival in extremely low gestational age neonates. Am J Obstet Gynecol 2014;211:519 e1-7.
  4. Rabe H, Jewison A, Alvarez RF, et al. Milking compared with delayed cord clamping to increase placental transfusion in preterm neonates: a randomized controlled trial. Obstet Gynecol 2011;117:205-11.
  5. Soll RF, Tarnow-Mordi WO. Optimizing Placental Transfusion for Preterm Infants. Pediatrics 2015;136:177-9.
  6. Ghavam S, Batra D, Mercer J, et al. Effects of placental transfusion in extremely low birthweight infants: meta-analysis of long- and short-term outcomes. Transfusion 2014;54:1192-8.
  7. Foglia EE, Nolen TL, DeMauro SB, et al. Short-term Outcomes of Infants Enrolled in Randomized Clinical Trials vs Those Eligible but Not Enrolled. JAMA 2015;313:2377-9.
  8. Rich WD, Auten KJ, Gantz MG, et al. Antenatal consent in the SUPPORT trial: challenges, costs, and representative enrollment. Pediatrics 2010;126:e215-21.