Secondary Logo

Journal Logo

Original Research

Milking Compared With Delayed Cord Clamping to Increase Placental Transfusion in Preterm Neonates

A Randomized Controlled Trial

Rabe, Heike MD, PhD; Jewison, Amanda MBBS; Fernandez Alvarez, Ramon MD; Crook, David PhD; Stilton, Denise RCN; Bradley, Robert MBBS; Holden, Desmond MBBS, PhD for the Brighton Perinatal Study Group

Author Information
doi: 10.1097/AOG.0b013e3181fe46ff

Anemia of prematurity, a common problem seen in almost all preterm neonates younger than 33 weeks of gestation, often requires transfusion of donor red blood cells. The placenta is a reservoir of fetal blood, which could be equally useful to the neonate. Consequently, the timing of the clamping of the cord has been the subject of much debate. Recent reviews1,2 of the data from 10 randomized trials involving a total of 454 preterm neonates found that a delay of cord clamping time of at least 30 seconds stabilizes the circulatory system of neonates during the first day of life, leading to less requirement for volume therapy, transfusion, and inotropic support, reduced the need for donor red cell transfusions, decreased the incidence of intraventricular hemorrhage, and improved neurodevelopmental outcome.1–4

After a Cochrane review,1 a slight delay of 30 seconds before clamping the cord has been introduced as the routine care for the delivery of preterm neonates in our own hospitals. In the present study, we investigated whether the more rapid strategy of milking the cord four times could achieve some of these positive benefits for preterm babies compared with our routine practice of waiting for 30 seconds before clamping the cord.

Studies on harvesting of placental blood indicate that the placenta can contain up to 40% of the total circulating fetal blood volume, with perhaps 15 to 20 mL being situated in the cord vein (5and T. Brune, personal communication). In a recent randomized study of 30 to 120 seconds of delayed cord clamping in 46 preterm neonates younger than 33 weeks of gestation,6 the blood volume was increased by delayed clamping (mean 74.4 mL/kg; range 45–103 mL/kg) compared with immediate clamping (62.7 mL/kg; range 47–77 mL/kg), corresponding to an 18% difference in blood volume (P<.001). Neonates born by cesarean delivery with delayed cord clamping had a higher blood volume compared with immediate clamping (mean 70.4 mL/kg; range 45–83 mL/kg; compared with mean 64 mL/kg; range 48–77 mL/kg) but this difference did not reach statistical significance (P=.10). The authors concluded that a cord clamping time of 40 seconds would achieve euvolemia (defined as a blood volume of approximately 75–100 mL/kg) in preterm neonates.

Milking the cord, an alternative way of achieving placento-fetal transfusion at the time of delivery, has been recently evaluated.7,8 Forty preterm neonates younger than 29 weeks of gestation were randomized to either immediate clamping of the cord or two to three times of milking the cord after delivery. The authors found higher hemoglobin values after birth (mean hemoglobin 141 g/L [range 122–169] compared with mean Hemoglobin 165 g/L [range 137–196]; P<.01), higher blood pressure (mean blood pressure 28±8 mm Hg compared with 34±9 mm Hg; P<.03), and less need for blood transfusions (mean number 4.0±4.2 compared with 1.7±3; P<.02) in the group allocated to milking.

Jones et al9 propose that the optimal circulating blood volume of preterm neonates should be in the range of 75 to 100 mL/kg body weight. The study by Aladangady6 suggests that this can be achieved by a slight delay in cord clamping time of 30 to 40 seconds. Their study also demonstrates that after immediate cord clamping, the circulating blood volume is approximately 47 to 77 mL/kg body weight.

To achieve a blood volume of at least 75 to 100 mL/kg body weight, 30 to 40 mL/kg has to be transferred into the neonate. In one study of harvested placental blood, a residual volume of approximately 15 to 20 mL blood could often be obtained from the cord vein itself (5and T. Brune, personal communication).


We conducted a randomized, comparative trial at a single tertiary care center (the Royal Sussex County Hospital in Brighton, UK). Our aim was to compare the two types of intervention regarding the clamping of the umbilical cord after delivery of a preterm neonate younger than 33 weeks of gestation, 30 seconds of waiting before clamping the cord compared with milking the cord four times. We hypothesized that there would be a difference in hemoglobin and hematocrit values between delayed cord clamping and milking the cord before cord clamping.

Preterm neonates between 24 0/7 and 32 6/7 completed weeks of gestation were included if antenatal informed consent could be obtained from the parents before delivery. Exclusion criteria were multiple pregnancies (twins and more), fetal hydrops, Rhesus sensitization, or known major congenital abnormalities. Ethical approval was obtained from Brighton and East Sussex Ethics Committee (reference 06/Q1905/15), and the study was approved and supervised by the Hospital Research and Development Directorate. The study was conducted according to the principles of European Union Good Clinical Practice.

When a preterm neonate is delivered, the obstetrician or midwife will not have access to the full length of the cord because part of it still remains inside the womb. We assumed that at least half to two thirds of the cord length would be available for milking, thus representing 7 to 10 mL of blood. If the cord were to be milked four times, then approximately 30 to 40 mL of blood would be transferred into the neonate. This was calculated under the assumption that the cord vein would be rapidly refilling itself because of its large diameter and the resulting negative pressure gradient compared with the placenta from the previous milking. When piloting the procedure, we were able to complete the milking within 10 to 12 seconds.

Before delivery, the fetuses were randomized into two groups. Those randomized to the clamping group were positioned 20 cm below the level of the placenta, between their mother's thighs (vaginal delivery) or to the mother's side (cesarean delivery). The umbilical cord then was clamped at 30 seconds. In the milking group, the neonates were positioned 20 cm below the level of the placenta, between the mother's thighs (vaginal delivery) or to the mother's side (cesarean delivery), with the cord being milked toward the neonate four times at a speed of 20 cm/2 seconds. This involved holding the cord at the introitus or cesarean delivery wound with one hand and milking the umbilical cord for its remaining accessible whole length toward the neonate four times (Fig. 1). The cord was clamped after the fourth milking. The neonates in both groups were placed immediately in plastic bags to maintain their temperature. The 30 seconds of cord clamping time was measured by using the wall-mounted clocks in each delivery suite.

Fig. 1.:
Milking of the cord.Rabe. Milking Compared With Delay in Cord Clamping. Obstet Gynecol 2011.

According to our local protocol, after vaginal delivery, women received a combination of oxytocin and ergometrin by intramuscular injection (unless the mother had hypertension, in which case oxytocin alone was administered) and, after cesarean delivery, intravenous oxytocin. Gestational age was determined by the obstetrical team before enrollment, by date of last menstruation if available, and by early booking ultrasound scan (usually week 8). Demographic and other maternal data were collected from the mother's notes at the time of enrollment or after delivery. Neonate data were obtained from clinical notes.

Randomization was based on computer-created tables performed by a person not involved in the trial. The randomization was stratified by gestational age, 24 0/7 to 27 6/7 completed weeks of gestation and 28 0/7 to 32 6/7 weeks of gestation.

The randomization allocation cards were kept on the labor ward in sealed opaque envelopes and consecutively numbered. The attending midwife opened the envelope before delivery. Blinding of the clinicians was not possible because of the nature of the interventions and because of the routine practice that the neonatal team is directly present in the delivery room.

The primary outcomes were neonatal blood hematocrit and hemoglobin at 1 hour after birth. Secondary outcomes included cord blood pH; Apgar scores at 5 and 10 minutes; temperature on admission to the neonatal unit; blood pressure at 4 hours of age; blood sugar on admission; maximum serum bilirubin and duration of phototherapy; hematocrit and hemoglobin at 24 hours, day 3, day 7, and weekly thereafter; number of blood transfusions in first 42 days of life; intraventricular hemorrhage (staging according to Papile10); number of septic episodes in first 42 days of life; death of newborn or mother; days requiring ventilation; number of surfactant treatments; days requiring oxygen; bronchopulmonary dysplasia defined as oxygen requirement at 36 weeks of corrected age; retinopathy of prematurity; necrotizing enterocolitis (staging according to Bell11); and length of hospital stay.

Departmental policy aimed at preventing anemia of prematurity is to introduce protein via parenteral nutrition on day 1 of life and oral iron of 6 mg/kg body weight if 60 mL/kg per 24 hours of milk feeds are tolerated12 (Box 1).

Box 1.:
Local Unit Guidelines for Prevention of Anemia of Prematurity Including Thresholds for Donor Blood Transfusions

The sample size calculation was based on the study by McDonnell,14 who used the same procedure described in our clamping group; the mean hematocrit at 1 hour was 55%±7%. To detect a 10% difference at α level of 0.05 and 80% power for a two-sided test, each group therefore would require 26 neonates ( Allowing for a 20% drop-out increased the total number of neonates required to 58 (29 in each group).

Statistical analysis of the results was performed using Microsoft Excel and SPSS 16.0. Variables were assessed for normality. For normally distributed variables, we calculated means and standard deviations, whereas for nonnormally distributed data we calculated medians and ranges. Student t test was used for analyzing differences in normally distributed data; the Mann-Whitney test was used when distributions were not normal. Repeated-measures analysis of variation was used to compare time courses within the two groups. Finally, we used χ2 and Fisher exact test to analyze categorical data. Data were analyzed on an intention-to-treat basis.


The study team recruited 58 neonates over an 18-month period from 212 women who were potentially eligible participants (Fig. 2). Maternal demographic data for both groups are shown in Table 1. None of the mothers died. We intended to measure the exact timing of when the cord was clamped by using a stop clock. However, because of unavailability of research staff during off hours, these data were not obtained in most cases.

Fig. 2.:
Flow chart of assessing and including eligible participants in the trial.Rabe. Milking Compared With Delay in Cord Clamping. Obstet Gynecol 2011.
Table 1:
Maternal Demographic and Clinical Data

Table 2 lists the neonates' characteristics and clinical data on admission, which were not significantly different. There were similar outcomes for both groups with regard to comorbidity, mortality, and the other secondary outcomes listed in Table 3. We focused the analysis on achieved hemoglobin values during the first 6 weeks of life. Mean hemoglobin values for each group at 1 hour after birth were 17.3 g/L for clamping and 17.5 g/L for milking (P.=71). Figure 3 illustrates no difference for hemoglobin values except for the last dataset with higher hemoglobin values in the milking group on day 42, which were confirmed by repeated-measure analysis of variance (P=.001). There were no statistically significant differences in leukocyte or platelet counts in the observation period (data not shown). Hemoglobin values were only analyzed until the first blood transfusion was administered (Fig. 4). Sixteen neonates in the clamping group and 10 in the milking group remained transfusion-free during this time period. One neonate in the milking group was born with an extremely low birth weight of 440 g. This neonate required excessive blood transfusions for replacing iatrogenic blood loss, had severe necrotizing enterocolitis, and later died after a prolonged hospital stay. Therefore, we reported the mean number of blood transfusions and other outcome measures with and without accounting for this neonate in Table 3.

Table 2:
Neonatal Characteristics and Admission Data
Table 3:
Neonatal Morbidities
Fig. 3.:
Hemoglobin values during the first 6 weeks of life. Values are analyzed only until the neonate's first transfusion (if any). No significant differences were identified between the groups except for infants aged 42 days (P=.01), which was confirmed by repeated-measures analysis of variance.Rabe. Milking Compared With Delay in Cord Clamping. Obstet Gynecol 2011.
Fig. 4.:
Kaplan-Meier graph for number of infants with delayed cord clamping or milking but without transfusions in the first 6 weeks of life.Rabe. Milking Compared With Delay in Cord Clamping. Obstet Gynecol 2011.


This study has demonstrated that milking the cord four times had similar effects as waiting for 30 seconds before clamping the cord in preterm neonates. A structured search performed by the Cochrane Pregnancy and Child Birth Group in February 2010 did not reveal any other studies of milking the cord four times (unpublished data). We also searched the Cochrane Neonatal Group trials register (January 16, 2008), The Cochrane Central Register of Controlled Trials (The Cochrane Library, Issue 4, 2007), PubMed (1966–January 16, 2008), and EMBASE (1974–January 16, 2008) using the following terms: umbilical cord AND clamp AND (preterm OR premature OR neonate, premature). Both groups in our study achieved higher mean hemoglobin values (clamping, 17.3 g/L; milking, 17.5 g/L; P=.71) compared with the milking group (16.5 g/L) in the trial by Hosono et al8 directly after birth. Our results show that our calculation of using milking the cord four times instead of 30 seconds of cord clamping time was correct. The inclusion of another control group of neonates randomized to immediate cord clamping would have been useful but was not ethically acceptable in a hospital where clamping the cord after 30 seconds is standard procedure.

It would have been preferable to use a more accurate measure of placental transfusion other than hemoglobin and hematocrit after birth, for example, by measuring the circulating blood volume or red cell mass.6,15 However, the available method of biotin labeling of red cells is time-consuming and difficult to perform on a large number of neonates (N. Aladangady, personal communication). The technique of using diluted adult hemoglobin compared with circulating fetal hemoglobin15 is limited to neonates who require a donor blood transfusion within the first few days after birth, a treatment we are trying to avoid.

Seventy-eight percent of neonates in the milking group were delivered by cesarean delivery compared with 58% in the clamping group, although this difference was not statistically significant (P=.16). Aladangady6 reported lower circulating red cell volume with a wider range in neonates born by this delivery mode in the delayed cord clamping group. One could speculate whether relatively more blood is pooled in the placenta when delivering a neonate by cesarean delivery as the active contraction of the uterine muscles to expel the placenta is interfered with by the surgical intervention. In our study, the neonates allocated to milking had similar hemoglobin values after birth compared with our clamping group, indicating a similar amount of blood transfer. There was a significantly higher hemoglobin value in the milking group after 42 days, which might indicate that the procedure of milking four times guarantees the transfer of blood from the placenta into the neonate, particularly in cesarean deliveries. One could speculate whether more stem cells have been transferred into the neonate as well.

Even though neonates with extremely low birth weight (less than 600 g birth weight) might not benefit as much from placental transfusion with regard to preventing blood transfusions because of the relatively large iatrogenic blood loss for blood tests, they can still benefit from the immediate postpartum effects on extra-uterine adaptation. There were no significant differences with regard to immediate measures, such as Apgar scores, temperature on admission, or mean arterial blood pressure. Because of the small sample size of this trial, subgroup analysis for mode of delivery or gestational age was not possible and not intended. We wanted to answer the question whether milking the cord four times could be an alternative method to waiting for 30 seconds before clamping the cord.

Ventilatory support after birth is now heavily influenced by the changing policies of attempting nasal continuous positive airway pressure ventilation rather than intubation directly after birth, or intubation and administration of surfactant with immediate extubation onto continuous positive airway pressure. Future trials in this area will need to take these new policies into consideration.

We recommend that future studies of various modes of clamping the cord carefully define the interventions to be used, eg, to use only one well-defined mode of (repeated) milking or a well-defined time interval before clamping the cord. This will enhance precise interpretation of the results, supports future meta-analysis, and will eventually lead to the definition of best practice in this difficult area.


1. Rabe H, Reynolds G, Diaz-Rossello J. Early versus delayed umbilical cord clamping in preterm infants. The Cochrane Database of Systematic Reviews 2004, Issue 4. Art: No: CD003248. DOI: 10.1002/14651858.CD003248.pub2.
2. Rabe H, Reynolds G, Diaz-Rossello J. A systematic review and meta-analysis of a brief delay in clamping the umbilical cord of preterm infants. Neonatology 2008;93:138–44.
3. Mercer JS, Vohr BR, McGrath M, Padbury JF, Wallach M, Oh W. Delayed cord clamping in very preterm infants reduces the incidence of intraventricular hemorrhage and late onset sepsis: a randomized, controlled trial. Pediatrics 2006;117:1235–42.
4. Mercer JS, Vohr BR, Erickson-Owens DA, Padbury JF, Oh W. Seven-month developmental outcomes of very low birth weight infants enrolled in a randomized controlled trial of delayed versus immediate cord clamping. J Perinatol 2010;30:11–6.
5. Brune T, Garritsen H, Witteler R, Schlake A, Wüllenweber J, Louwen F, et al. Autologous placental blood transfusion for the therapy of anaemic neonates. Biol Neonate 2002;81:236–43.
6. Aladangady N, McHugh S, Aitchison TC, Wardrop CA, Holland BM. Infants' blood volume in a controlled trial of placental transfusion at preterm delivery. Pediatrics 2006;117:93–8.
7. Hosono S, Mugishima H, Fujita H, Hosono A, Minato M, Okada T, et al. Umbilical cord milking reduces the need for red cell transfusions and improves neonatal adaptation in infants born less than 29 weeks' gestation: a randomised controlled trial. Arch Dis Child Fetal Neonatal Ed 2008;93:F14–9.
8. Hosono S, Mugishima H, Fujita H, Hosono A, Okada T, Takahashi S, et al. Blood pressure and urine output during the first 120 h of life in infants born at less than 29 weeks' gestation related to umbilical cord milking. Arch Dis Child Fetal Neonatal Ed 2009;94:F328–31.
9. Jones JG, Holland BM, Hudson IR, Wardrop CA. Total circulating red blood cells versus haematocrit as the primary descriptor of oxygen transport by the blood. Br J Haematol 1990;76:288–94.
10. Papile LA, Burstein J, Burstein R, Koffler H. Incidence and evolution of subependymal and intraventricular hemorrhage: a study of infants with birth weights less than 1,500 gm. J Pediatr 1978;92:529–34.
11. Bell MJ, Ternberg JL, Feigin RD, Keating JP, Marshall R, Barton L, et al. Neonatal necrotizing enterocolitis. Therapeutic decisions based upon clinical staging. Ann Surg 1978;187:1–7.
12. Rabe H, Fernandez Alvarez JR, Seddon P, Lawn C, Amess PN. A management guideline to reduce the frequency of blood transfusion in very low birth weight infants. Am J Perinat 2009;26:179–83.
13. Kirpalani H, Whyte RK, Andersen C, Asztalos EV, Heddle N, Blajchman MA, et al. The Premature Infants in Need of Transfusion (PINT) study: a randomized, controlled trial of a restrictive (low) versus with liberal (high) transfusion threshold for extremely low birth weight infants. J Pediatr 2006;149:301–7.
    14. McDonnell M, Henderson-Smart DJ. Delayed umbilical cord clamping in preterm infants: a feasibility study. J Paediatr Child Health 1997;33:308–10.
    15. Strauss RG, Mock DM, Johnson KJ, Cress GA, Burmeister LF, Zimmerman MB, et al. A randomized clinical trial comparing immediate versus delayed clamping of the umbilical cord in preterm infants: short-term clinical and laboratory endpoints. Transfusion 2008;48:658–65.

    Cited By

    This article has been cited 1 time(s).

    Obstetrics & Gynecology
    Increasing the Placental Transfusion for Preterm Infants
    Bell, E
    Obstetrics & Gynecology, 117(2 Part 1): 203-204.
    PDF (110) | CrossRef

    Supplemental Digital Content

    © 2011 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.