Research
Original Investigation
Efficacy and Safety of Umbilical Cord Milking at Birth
A Systematic Review and Meta-analysis
Heidi Al-Wassia, MD; Prakesh S. Shah, MD, MSc
Editorial
IMPORTANCE Umbilical cord milking (UCM) is suggested to improve neonatal outcomes.
OBJECTIVES To perform a systematic review and meta-analysis of the efficacy and safety of
UCM in full-term and preterm neonates.
Journal Club Slides and
Supplemental content at
jamapediatrics.com
DATA SOURCES A systematic search of MEDLINE, EMBASE, CINAHL, the Cochrane Database
of Clinical Trials, the clinicaltrails.gov database, and the reference list of retrieved articles from
1940 to 2014.
STUDY SELECTION Randomized clinical trials comparing UCM with other strategies of
handling the umbilical cord at birth in full-term and preterm infants. Seven of the 18 initially
identified studies were selected.
DATA EXTRACTION AND SYNTHESIS Two reviewers independently extracted data and
assessed the risk for bias in included trials using the criteria outlined in the Cochrane
Handbook for Systematic Reviews of Interventions.
MAIN OUTCOMES AND MEASURES Neonatal mortality before discharge from the hospital.
RESULTS We included 7 randomized clinical trials involving 501 infants. Infants with a
gestational age of less than 33 weeks allocated to UCM compared with control conditions
showed no difference in the risk for mortality (risk ratio [RR], 0.75 [95% CI, 0.35 to 1.64]; risk
difference [RD], −0.02 [95% CI, −0.09 to 0.04]), hypotension requiring volume expanders
(RR, 0.71 [95% CI, 0.41 to 1.25]; RD, −0.09 [95% CI, −0.22 to 0.05]), or inotrope support (RR,
0.77 [95% CI, 0.51 to 1.17]; RD, −0.10 [95% CI, −0.25 to 0.05]). Higher initial levels of
hemoglobin (mean difference, 2.0 [95% CI, 1.3-2.7] g/dL) and hematocrit (mean difference,
4.5% [95% CI, 1.5%-7.4%]) were identified in the UCM groups. We found a reduced risk for
oxygen requirement at 36 weeks (RR, 0.42 [95% CI, 0.21 to 0.83]; RD, −0.14 [95% CI, −0.25
to −0.04]) and for intraventricular hemorrhage of all grades (RR, 0.62 [95% CI, 0.41 to 0.93];
RD, −0.12 [95% CI, −0.22 to −0.02]) in infants assigned to UCM. Among infants with a
gestational age of at least 33 weeks, UCM was associated with higher hemoglobin levels in the
first 48 hours in 224 infants (mean difference, 1.2 [95% CI, 0.8-1.5] g/dL) and at 6 weeks of
life in 170 infants (mean difference, 1.1 [95% CI, 0.7-1.5] g/dL).
CONCLUSIONS AND RELEVANCE Umbilical cord milking was associated with some benefits and
no adverse effects in the immediate postnatal period in preterm infants (gestational age, <33
weeks); however, further studies are warranted to assess the effect of UCM on neonatal and
long-term outcomes.
Author Affiliations: Department of
Pediatrics, King Abdulaziz University,
Jeddah, Saudi Arabia (Al-Wassia);
Department of Pediatrics, Mt Sinai
Hospital, Toronto, Ontario, Canada
(Shah); Institute of Health Policy,
Management, and Evaluation,
University of Toronto, Toronto,
Ontario, Canada (Shah).
JAMA Pediatr. doi:10.1001/jamapediatrics.2014.1906
Published online November 3, 2014.
Corresponding Author: Heidi
Al-Wassia, MD, Department of
Pediatrics, King Abdulaziz University,
PO Box 80215, Jeddah 21589, Saudi
Arabia ([email protected]).
E1
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ on 11/07/2014
Research Original Investigation
Efficacy and Safety of Umbilical Cord Milking
I
n 1949, McCausland et al 1 surveyed members of the
American Board of Obstetrics and Gynecology and
reported no uniformity of practice in their management
of umbilical cord and placental blood. The benefits of
delayed cord clamping (DCC) and other strategies to influence placental transfusion at birth have been under investigation for decades.2,3 Recently, interest in the evidently old
procedures of transferring residual blood from the placenta
to the infant by means of DCC or umbilical cord milking
(UCM) has shown a resurgence. However, practice among
obstetricians varies 7 decades later.
A recent Cochrane review reported that DCC in preterm
infants was associated with fewer transfusions of packed
red blood cells and a lower risk for intraventricular hemorrhage (IVH) and necrotizing enterocolitis compared with
immediate cord clamping (ICC).4 Concerns about polycythemia, hyperbilirubinemia, and delay in transition were not
sustained in the review. Because of the level of heterogeneity between trials included in that review, no clear effect
on other outcomes could be concluded. For full-term
infants, DCC is suggested; however, to our knowledge, longterm outcomes of DCC have not been investigated in detail.5
Furthermore, the optimal timing for DCC has yet to be
determined because studies varied in the duration, from
less than 30 seconds to 180 seconds in preterm infants4 and
as long as 300 seconds in full-term infants.5
The stripping of blood from the umbilical cord, or UCM,
was pondered for years and suspected to be beneficial.1,6-10
Nevertheless, methodologic limitations of older studies hindered the adoption of UCM as a standard of care. A more
recent series of studies assessed the safety and efficacy of
UCM.11-17 The key difference between DCC and UCM is the
mechanism of cord blood transfer to the infant. In DCC, a
passive transfer of additional blood volume occurs at a slow
rate, mostly by uterine contractions, whereas in UCM an
active transfer of additional blood volume occurs at a rapid
rate and within a short time, which may or may not be beneficial to neonates, especially preterm neonates. Our objective was to perform a systematic review and meta-analysis
of the efficacy and safety of UCM in full-term and preterm
infants.
Type of Studies
We included randomized clinical trials (RCTs) published in
the form of original research articles in peer-reviewed journals. Quasi-RCTs, observational studies, narrative reviews,
letters, editorials, abstracts, and commentaries were
excluded, as were duplicate reports that lacked additional
information. Case reports and series, qualitative studies,
review articles, and studies that did not report UCM methods were read to identify other potential studies but
excluded from the meta-analysis.
Type of Intervention
We included studies that investigated UCM vs a control intervention (other strategies of handling the umbilical cord at birth,
including ICC, DCC, or no intervention) and reported any outcomes of interest. The investigators must have reported the
UCM procedure in detail so that it is replicable.
Outcomes
Studies were included if they reported a primary outcome of
neonatal mortality before discharge from the hospital or any
1 of the following secondary outcomes:
1. Condition at birth (ie, cord arterial pH, Apgar scores at 1 and
5 minutes);
2. Hematological variables, including the first hematocrit and
hemoglobin levels measured within 48 hours of birth, the
need for transfusion of packed red blood cells before discharge, peak serum bilirubin level before discharge, hyperbilirubinemia requiring phototherapy, polycythemia (venous hematocrit level, >65% [to convert to a proportion of
1.0, multiply by 0.01]) at any time during admission, and levels of hemoglobin and ferritin at 3 to 6 months of age;
3. Short-term morbidities, including respiratory distress syndrome, hypotension in the first 24 hours of birth requiring
volume or inotrope support, IVH of all grades,19 severe IVH
(grade III or IV), oxygen dependency at 28 days and 36 weeks
of corrected GA,20 stage II or III necrotizing entercolitis,21
late-onset sepsis, retinopathy of prematurity, patent ductus arteriosus, and duration of hospital stay; and
4. Neurodevelopmental outcomes at 18 and 24 months.
Review Methods
Methods
We followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines18 for the conduct and reporting of this review. Both of us independently assessed and identified articles for eligibility and collected data
for this study. All discrepancies encountered in the review process were resolved by consensus.
Search Strategy
Published RCTs were identified using manual and electronic
search strategies. This search was applied to MEDLINE (1946
to April 2014), EMBASE (1980 to April 17, 2014), CINAHL (1982
to April 2014), and the Cochrane Central Register of Controlled Trials (April 2014) (eMethods in the Supplement). We
also searched the online meta-register of Current Controlled
Trials for relevant ongoing clinical trials (April 2014). Additional citations were sought by hand searching the reference
list of the retrieved articles.
Criteria for Considering Studies for Review
Study Population
We included studies of full-term and preterm infants. We made
an a priori decision to compare the 2 populations (gestational
age [GA], <33 and ≥33 weeks) separately because effects and
outcomes of interest would be distinct for these two groups.
E2
Data Extraction
Data extraction was performed using a standardized data collection form. Any discrepancies in extracted data were resolved by consensus. We contacted authors to obtain additional or missing data.
JAMA Pediatrics Published online November 3, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ on 11/07/2014
jamapediatrics.com
Efficacy and Safety of Umbilical Cord Milking
Original Investigation Research
Table 1. Characteristics of the Included Studies
UCM Characteristic
Source
Population
No. of
Times
Speed
Control
Condition
Exclusion Criteria
Alan et al,16 2014
44 Preterm infants
(GA, ≤32 wk) with birth
weight ≤1500 g
3
5 cm/s
ICC
Suspected TTTS or discordant twin, major congenital
and chromosomal anomalies, vaginal bleeding, HDNB or
hydrops fetalis, IUGR, maternal DM treated with insulin
Erickson-Owens et al,17
2012
24 Full-term infants
delivered by elective CD
5
Not reported
ICC
Maternal medical or obstetric complications, maternal
severe anemia or clotting disorders, suspected IUGR,
smoking, congenital anomalies
2-3
20 cm within 2 s ICC
Hosono et al,11 2008
40 Preterm infants
(GA, 24-28 wk)
Katheria et al,15 2014
60 Preterm infants
(GA, 23 wk to 31 wk 6 d)
3
March et al,12 2013
75 Preterm infants
(GA, 24-28 wk)
3
Rabe et al,14 2011
58 Preterm infants
(GA, 24 wk to 32 wk 6 d)
4
20 cm in 2 s
200 Infants
(GA, >34 wk 6 d)
3
10 cm/s
13
Upadhyay et al,
2013
20 cm in 2 s
Not reported
Multiple births, major congenital anomalies, hydrops
fetalis
ICC
Pregnant women with imminent delivery,
monochorionic multiple births, incarcerated mothers,
placenta previa, concerns of abruption, refusal by
obstetrician
ICC
Major congenital anomalies, Rh sensitization, hydrops
fetalis, maternal recent exposure to parvovirus,
elevated peak systolic velocity of the fetal middle
cerebral artery, suspected placental abruption
DCC at 30 s
Multiple pregnancy, hydrops fetalis, Rh sensitization,
major congenital anomalies
DCC within 30 s
Umbilical cord length <25 cm, nonvigorous, meconiumstained amniotic fluid, CD for fetal compromise,
multiple births, Rh-negative mothers, major congenital
anomalies, cord prolapse, placenta previa or abruption,
cord abnormalities (true knots), hydrops fetalis
Abbreviations: CD, cesarean delivery; DCC, delayed cord clamping;
DM, diabetes mellitus; GA, gestational age; HDNB, hemolytic disease of
newborn; ICC, immediate cord clamping; IUGR, intrauterine growth restriction;
TTTS, twin-to-twin transfusion syndrome; UCM, umbilical cord milking.
Assessment of Risk for Bias
Both reviewers independently assessed the risk for bias in each
study using the criteria outlined in the Cochrane Handbook for
Systematic Reviews of Interventions.22 Any disagreements were
resolved by discussion and consensus. The risk for bias was assessed using the following key criteria: sequence generation,
allocation concealment, blinding of assessors, and attrition bias.
We assessed each criterion as having a low, high, or unclear risk
for bias. An overall risk for bias was determined for each study
according to the criteria suggested by Higgins and Green.22
lowing formula: 75th percentile − 25th percentile = 1.35 SD.
When ranges were given, we used the formula suggested by
Hozo et al.25
Data Synthesis and Statistical Analysis
Similar to other meta-analyses, we made no adjustment for
multiple analyses. For categorical outcomes, we calculated
the pooled risk ratio (RR) and the related 95% CI. For continuous outcomes, we calculated the pooled mean difference (MD) and the corresponding 95% CI. When heterogeneity was not observed, the fixed-effect modeling approach
was applied using the Mantel-Haenszel method23 for categorical variables and the inverse variance method for continuous variables. When heterogeneity was detected, a
random-effects model was applied using the DerSimonian
and Laird method. 23 Heterogeneity across studies was
assessed by calculating the I 2 values and Q statistics. 24
P < .10 was defined to note statistical significance in the
analysis of heterogeneity.
When results in studies were presented in different forms,
we attempted to contact authors to obtain original data so that
means and SDs could be computed. When that attempt was
unsuccessful, we assumed normal distribution of the data, substituted the median for the mean, and calculated the SD from
the upper and lower interquartile range (IQR) using the foljamapediatrics.com
Subgroup Analyses
We planned to perform a subgroup analysis by control intervention (none or ICC and DCC); however, the limited number
of studies precluded such subgroup analyses. We assessed
whether the control intervention explained some of the heterogeneous results. Clinical heterogeneity in the included studies was assessed and is reported in Table 1 by describing the
populations included, treatment groups compared, and variations in procedures of UCM. We planned to assess publication bias using a funnel plot if 10 studies were included in a
meta-analysis.
Results
Study Identification
The first screen of titles and abstracts identified 18 potential
citations for secondary review. A detailed evaluation of the retrieved citations identified 7 studies11-17 for inclusion in this review (Figure 1). In addition, we found 3 ongoing studies in the
Current Controlled Trials meta-register.26-28
Description of Studies
The number of infants in each study, description of the UCM
method, how the cord was handled in the control group,
and inclusion and exclusion criteria are depicted in Table 1.
The description of the UCM technique varied among studies, including the number of times the cord was stripped
JAMA Pediatrics Published online November 3, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ on 11/07/2014
E3
Research Original Investigation
Efficacy and Safety of Umbilical Cord Milking
Comparison of Preterm Infants
Figure 1. Summary of Study Selection Process
256 Records identified
through database
search
2 Additional records
identified through
other sources
166 Records after
duplicates removed
148 Records excluded
141 Not relevant
3 Abstract
1 Doctoral dissertation
3 Commentaries
166 Records screened
18 Full-text articles
assessed for eligibility
11 Full-text articles excluded
3 Experimental groups
had UCM and DCC
8 Nonrandomized trials
7 Studies included in
qualitative synthesis
7 Studies included in
quantitative synthesis
(meta-analysis)
DCC indicates delayed cord clamping; UCM, umbilical cord milking.
toward the infant, the milking speed, and whether the cord
was cut before or after milking. Upadhyay et al13 cut the
cord at 25 cm of length from the umbilical stump and then
milked the cord. None of the other included trials milked
the cord this way, which might have affected the volume of
blood transfused. Milking of the umbilical cord was compared with ICC in 5 trials11,12,15-17 and with DCC in 2 trials.13,14
Delayed cord clamping in the control group was defined as
DCC at 30 seconds14 or within 30 seconds.13 The GA groups
of the samples were heterogeneous across trials. EricksonOwens et al17 included infants delivered by cesarean delivery only. All trials but Katheria et al15 excluded infants with
major congenital anomalies, and all but Katheria et al15 and
Erickson-Owens et al17 noted the exclusion of infants with
hydrops fetalis. Multiple pregnancies were excluded by
Hosono et al,11 Upadhyay et al,13 and Rabe et al,14 and Alan
et al16 and Erickson-Owens et al17 excluded fetuses with suspected intrauterine growth restriction. Reporting of outcomes varied across studies. Only Upadhyay et al13 collected
and reported data on maternal hemoglobin levels and antenatal iron supplementation. Hosono et al,11 Rabe et al,14 and
Alan et al16 reported no significant difference in iatrogenic
blood loss between the 2 groups, and only Alan et al16 presented the guideline they followed for transfusion of packed
red blood cells in preterm infants.
Risk for Bias Among Included Studies
Most studies had a low risk for bias. All studies reported that
blinding of the clinicians involved was not possible because
of the nature of the intervention (eTable in the Supplement).
E4
Primary Outcome
Five studies11,12,14-16 of 277 infants reported outcomes of UCM
vs a control condition in preterm infants (GA, <33 weeks). We
found no statistically significant difference in the risk for death
between infants assigned to UCM and the control group (RR,
0.75 [95% CI, 0.35-1.64]; P = .48) (Figure 2A).
Secondary Outcomes
The results of reported secondary outcomes in the trials are
detailed in Table 2. We found a reduced risk for oxygen requirement at 36 weeks (RR, 0.42 [95% CI, 0.21-0.83]) (Figure 2B)
and IVH of all grades (RR, 0.62 [95% CI, 0.41-0.93]) (Figure 2C)
in infants assigned to UCM. Hemoglobin and ferritin levels at
3 to 6 months of age and neurodevelopmental outcomes were
not reported in any of the included studies. Some of the outcomes, which were reported in nonparametric measures, are
described below.
Cord Arterial pH | Rabe et al14 reported no difference in cord arterial pH between groups (UCM group median, 7.3 [range, 6.87.4]; control group median, 7.3 [range, 6.9-7.3]; P = .31). March
et al12 reported no difference in cord pH but did not specify
whether the blood was venous or arterial (UCM group median, 7.3 [IQR, 7.3-7.3]; control group median, 7.3 [IQR, 7.37.4]; P = .44).
Apgar Scores | Hosono et al11 reported higher 1-minute Apgar
scores in infants who underwent UCM (MD, 1.2 [95% CI,
0.02-2.4]), whereas no difference in Apgar scores at 1 minute
was reported by March et al12 (UCM group median, 4 [IQR,
1-5.5]; control group median, 4 [IQR, 2-5]; P = .75), Alan et
al 16 (UCM group median, 7 [range, 3-8]; control group
median, 7 [range, 2-8]; P = .77), and Katheria et al (data
unavailable).15 We found no statistically significant difference in Apgar scores at 5 minutes in the 5 studies that
reported this outcome.11,12,14-16
Duration of Hospital Stay | No difference in median days of hospital stay between comparison groups was reported by Rabe
et al14 (UCM group median, 46 [IQR, 13-315] days; control
group median, 50 [IQR, 1-189] days; P = .58). Alan et al16
reported similar results (UCM group median, 47 [range,
17-72] days; control group median, 53 [range, 17-83] days;
P = .33).
Comparison of Full-term Infants
Upadhyay et al13 and Erickson-Owens et al17 reported outcomes of UCM vs a control condition in 224 infants with a
GA of at least 33 weeks. Neither study reported outcomes on
mortality, cord arterial pH or cord blood gas levels, hypotension requiring intervention, the need for transfusion of
packed red blood cells, or neurodevelopmental outcomes.
Erickson-Owens et al17 reported no statistically significant
difference in Apgar scores at 1 and 5 minutes between the 2
groups but statistically significantly higher hematocrit values in the UCM group (MD, 7.5% [95% CI, 0.7%-1.5%]).
Umbilical cord milking resulted in significantly higher
JAMA Pediatrics Published online November 3, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ on 11/07/2014
jamapediatrics.com
Efficacy and Safety of Umbilical Cord Milking
Original Investigation Research
Figure 2. Outcomes in Umbilical Cord Milking (UCM) vs Control (Immediate or Delayed Cord Clamping) Groups in Preterm Infants
A Mortality before discharge in preterm infantsa
UCM Group
Control Group
No. of
Events
Total
No. of
Events
Alan et al,16 2014
Hosono et al,11 2008
2
2
22
20
2
3
22
20
Katheria et al,15 2014
2
30
1
March et al,12 2013
2
36
4
Rabe et al,14 2011
2
27
10
135
Source
Total
Total Weight, %
Fixed M-H RR
(95% CI)
14.7
22.1
1.00 (0.15-6.48)
0.67 (0.12-3.57)
30
7.4
2.00 (0.19-20.90)
39
28.3
0.54 (0.11-2.78)
4
31
27.5
0.57 (0.11-2.89)
14
142
100.0
0.75 (0.35-1.64)
Favors UCM
Favors Control
Heterogeneity x42 = 1.04 (P = .90); I 2 = 0%
Test for overall effect: z = .71 (P = .48)
0.1
0.2
0.5
1.0
2.0
5.0
10.0
Fixed M-H RR (95% CI)
B
Oxygen requirement at 36 wk, postmenstrual age in preterm infantsa
UCM Group
Control Group
Fixed M-H RR
(95% CI)
No. of
Events
Total
No. of
Events
Alan et al,16 2014
Hosono et al,11 2008
2
0
19
18
3
4
19
17
12.9
19.8
0.67 (0.13-3.55)
0.11 (0.01-1.82)
Katheria et al,15 2014
4
30
12
30
51.4
0.33 (0.12-0.92)
Rabe et al,14 2011
3
27
4
31
16.0
0.86 (0.21-3.51)
Total
9
94
23
97
100.0
0.42 (0.21-0.83)
Source
Total Weight, %
Favors UCM
Favors Control
Heterogeneity x23 = 2.41 (P = .49); I 2 = 0%
Test for overall effect: z = 2.50 (P = .01)
0.005
0.1
1.0
10
200
Fixed M-H RR (95% CI)
C
IVH of all gradesa
UCM Group
Control Group
Fixed M-H RR
(95% CI)
No. of
Events
Total
No. of
Events
Alan et al,16 2014
Hosono et al,11 2008
4
3
22
20
3
5
22
20
6.7
11.2
1.33 (0.34-5.28)
0.60 (0.17-2.18)
Katheria et al,15 2014
8
30
11
30
24.6
0.73 (0.34-1.55)
March et al,12 2013
9
36
20
39
42.9
0.49 (0.26-0.93)
Rabe et al,14 2011
3
27
7
31
14.6
0.49 (0.14-1.72)
27
135
46
142
100.0
0.62 (0.41-0.93)
Source
Total
Total Weight, %
Favors UCM
Favors Control
Heterogeneity x42 = 2.03 (P = .73); I 2 = 0%
Test for overall effect: z = 2.32 (P = .02)
0.1
0.2
0.5
1.0
2.0
5.0
10.0
Fixed M-H RR (95% CI)
Data were obtained from 5 studies included in our meta-analysis. A, Mortality
before discharge. B, Oxygen requirement at postmenstrual age of 36 weeks. C,
Intraventricular hemorrhage (IVH) of any grade. Total columns indicate number
of infants in each group. Weights have been rounded and might not total 100.
M-H indicates Mantel-Haenszel; RR, risk ratio. Whiskers indicate 95% CIs. Data
hemoglobin values in the first 48 hours after birth (for the
224 participants in both studies, MD, 1.2 [95% CI, 0.8-1.5]
g/dL [to convert to grams per liter, multiply by 10.0];
I2 = 34%). We found no significant difference between the 2
groups in peak bilirubin level in 24 participants in the study
by Erickson-Owens et al 17 (MD, 0.6 [95% CI, −1.9 to 3.0]
mg/dL [to convert to micromoles per liter, multiply by
17.104]) and in the need for phototherapy for the 224 participants in both studies13,17 (RR, 5.0 [95% CI, 0.3-49.3]). Upadhyay et al13 reported no polycythemia in either group, and
only 1 case of polycythemia was diagnosed in the UCM
group in the study by Erickson-Owens et al.17 One hundred
seventy infants in the UCM group described by Upadhyay et
al13 had statistically significantly higher levels of hemoglobin (MD, 1.1 [95% CI, 0.7-1.5] g/dL) and ferritin (MD, 79 [95%
jamapediatrics.com
markers vary in size owing to different weights of individual studies in the
meta-analysis.
a
Indicates gestational age of less than 33 weeks.
CI, 58-101] ng/mL [to convert to micrograms per liter, multiply by 2.247]) at 6 weeks of age compared with controls
(P < .05).
Discussion
In what is, to our knowledge, the first systematic review of
UCM, we identified 7 eligible studies of 501 full-term and preterm infants. Most of the included studies were of high quality and had a low risk for bias. We found heterogeneity in the
method of actual implementation of UCM between studies. In
infants with a GA of less than 33 weeks, UCM was not associated with a difference in the primary outcome of the risk for
mortality before discharge; however, UCM was associated with
JAMA Pediatrics Published online November 3, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ on 11/07/2014
E5
Research Original Investigation
Efficacy and Safety of Umbilical Cord Milking
Table 2. Comparison of Umbilical Cord Milking vs Control Intervention in Preterm Infantsa
No. of
Studies
No. of
Participants
Mortality before discharge
5
277
0.75 (0.35 to 1.64)
−0.02 (−0.09 to 0.04)
NA
0
Hypotension requiring volume expanders
3
138
0.71 (0.41 to 1.25)
−0.09 (−0.22 to 0.05)
NA
64
84
Outcome
RR or MD (95% CI)b
RD (95% CI)
I2 Value, %
NNT
Hypotension requiring inotrope support
3
138
0.77 (0.51 to 1.17)
−0.10 (−0.25 to 0.05)
NA
Initial hemoglobin level, g/dL
3
159
2.0 (1.3 to 2.7)c
NA
NA
0
Initial hematocrit level, %
5
277
4.5 (1.5 to 7.4)c
NA
NA
63
Maximal serum bilirubin level, mg/dL
5
277
0.1 (−0.4 to 0.6)
NA
NA
0
Need for phototherapy
2
135
0.95 (0.88 to 1.03)
−0.05 (−0.12 to 0.03)
NA
0
Need for PRBC transfusion
5
271
0.81 (0.62 to 1.05)
−0.16 (−0.32 to 0.00)
NA
62
Respiratory distress syndrome
3
138
1.03 (0.78 to 1.35)
0.01 (−0.14 to 0.17)
NA
52
Oxygen requirement at postmenstrual age of
36 wk
4
191
0.42 (0.21 to 0.83)c
−0.14 (−0.25 to −0.04)
8 (5-28)
38
Oxygen requirement at 28 d
4
194
0.93 (0.53 to 1.64)
−0.01 (−0.13 to 0.10)
NA
0
All
5
277
0.62 (0.41 to 0.93)c
−0.12 (−0.22 to −0.02)
9 (5-50)
3
>III
4
196
0.83 (0.36 to 1.90)
−0.02 (−0.10 to 0.06)
NA
35
Medical management
5
271
1.85 (0.33 to 10.53)
0.02 (−0.04 to 0.08)
NA
60
Perforation
3
156
0.87 (0.25 to 3.01)
−0.01 (−0.08 to 0.07)
NA
15
3
170
0.67 (0.12 to 3.75)
−0.01 (−0.07 to 0.04)
NA
0
All grades
3
168
0.91 (0.73 to 1.13)
−0.05 (−0.16 to 0.06)
NA
23
Requiring treatment
2
98
1.00 (0.18 to 5.51)
0.00 (−0.09 to 0.09)
NA
0
Late-onset sepsis
4
213
0.84 (0.58 to 1.21)
−0.06 (−0.17 to 0.06)
NA
12
Patent ductus arteriosus
3
138
0.91 (0.56 to 1.48)
−0.03 (−0.18 to 0.12)
NA
0
Intraventricular hemorrhage grade
Necrotizing enterocolitis
Periventricular leukomalacia
Retinopathy of prematurity
E6
Abbreviations: MD, mean difference; NA, not applicable; NNT, number needed
to treat; PRBC, packed red blood cells; RD, risk difference; RR, risk ratio.
a
Indicates gestational age of less than 33 weeks.
b
Continuous outcomes are reported as mean difference.
SI conversion factors: To convert bilirubin to micromoles per liter, multiply by
17.104; hematocrit to a proportion of 1.0, multiply by 0.01; and hemoglobin to
grams per liter, multiply by 10.0.
c
Indicates statistically significant results.
higher initial hemoglobin values, a lower risk for oxygen requirement at a postmenstrual age of 36 weeks, and a lower risk
for IVH of all grades. These improvements did not translate into
a reduction in the need for blood transfusion or in the risk for
severe IVH or periventricular leukomalacia. Although hematocrit levels were significantly higher in the UCM group, no
study reported an increased risk for polycythemia or hyperbilirubinemia requiring treatment. In infants with a GA of at
least 33 weeks, UCM was associated with a higher hemoglobin value in the first 48 hours of life and at 6 weeks of age without an increase in the risk for hyperbilirubinemia. None of the
studies evaluated long-term neurodevelopmental outcomes.
We could not make relevant comparisons because of the
absence of previous reviews of UCM. However, we believe that
this review provides a fair comparison between our analysis
and those of DCC. Both procedures lead to placental transfusion in active or passive forms. DeMarsh et al2 reported that
the blood volume of the newborn infant varies according to
the amount of placental transfusion after birth. Depriving infants of placental blood predisposes them to anemia early in
life,29,30 which has been correlated with poor neurodevelopmental outcomes.31 Our result of a reduction in any type of IVH
is in agreement with that of Rabe et al,4 who reported that DCC
was associated with improvement in blood pressure and re-
ductions in the need for blood transfusion and risks for IVH
(all grades) and necrotizing enterocolitis. Takami et al32 reported that UCM in a cohort of very low-birth-weight infants
stabilized cerebral oxygenation and perfusion, as indicated by
near-infrared spectroscopy variables. Moreover, Katheria et al15
demonstrated a greater superior vena caval flow in preterm infants (GA, <32 weeks) allocated to UCM compared with ICC.
Anemia and low systemic blood flow (as measured by superior vena caval flow) are known risk factors for neurologic insults in the first week after birth.33 Placental transfusion resulting from DCC or UCM might improve systemic perfusion
and mitigate fluctuation in cerebral perfusion pressure, leading to a reduction in neurologic injury. However, superior vena
caval flow can be considered a surrogate marker of perfusion
in neonates, and further studies are needed to confirm or refute this finding.
The finding of a significantly lower risk for oxygen requirement at a postmenstrual age of 36 weeks in the UCM group is
different and of interest compared with meta-analyses of DCC
in preterm infants, which found no clear difference between
groups.4 Recently, interest in the potential for progenitor or
stem cell therapies to prevent or treat bronchopulmonary dysplasia in preterm infants has grown.34 In animal models, treatment with cord blood–derived mesenchymal stromal cells
JAMA Pediatrics Published online November 3, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ on 11/07/2014
jamapediatrics.com
Efficacy and Safety of Umbilical Cord Milking
Original Investigation Research
showed promising results in the prevention of adverse effects of hyperoxia-induced lung injury.35 The possibility of
transfer of some progenitor cells during UCM has been speculated.
We planned to perform subgroup analysis by controlled intervention (DCC or ICC). However, the small number of trials
in the present review precluded this analysis. A fundamental
physiological distinction between these two methods warrants careful examination of such comparison when further
information is available. A large body of evidence suggests that
additional flow of blood from the placenta to the infant through
DCC has several advantages to the newborn and does not result in harm. However, controversies exist over what constitutes the optimal time of DCC and the safety of this method
when active resuscitation of the newborn is anticipated. Moreover, the delay in severing the umbilical cord might interfere
with controlling maternal bleeding and suturing the uterine
incision in cesarean delivery or the episiotomy or perineal tear
in vaginal delivery.6 On the other hand, UCM is believed to be
a simple intervention that can be performed in seconds rather
than minutes. This method may be appealing in cases of anticipated birth asphyxia, given the importance of time in such
situations. We can also speculate that pushing certain progenitor cells in some situations, such as autologous blood transfusion, is a subject of human trials in birth asphyxia.36 Given
the suggestions by professional organizations to use DCC as a
method of choice, conceiving any new studies of UCM without considering DCC as the control intervention will be very
difficult, and further research comparing these two methods
is warranted.
This review is, to our knowledge, the first to assess UCM
and its effect on neonatal outcomes. A comprehensive literature search, standard methodologic approach, and contact of original authors to obtain missing information are
some of the strengths of this review. However, we acknowledge the limitations of this review. First, the inclusion and
ARTICLE INFORMATION
Accepted for Publication: July 31, 2014.
Published Online: November 3, 2014.
doi:10.1001/jamapediatrics.2014.1906.
Author Contributions: Drs Al-Wassia and Shah had
full access to all the data in the study and take
responsibility for the integrity of the data and the
accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: All
authors.
Drafting of the manuscript: Al-Wassia.
Critical revision of the manuscript for important
intellectual content: All authors.
Statistical analysis: Al-Wassia.
Administrative, technical, or material support:
Al-Wassia.
Study supervision: All authors.
Conflict of Interest Disclosures: Dr Shah is
supported by an Applied Research Chair in
Reproductive and Child Health Services Research
funding from the Canadian Institute of Health
Research. No other disclosures were reported.
exclusion criteria and measured outcomes varied widely
across studies. Second, although the method of UCM was
described in detail in all trials, the way it was attained varied widely among these trials. Third, methodologic limitations of earlier evidence and the small number of participants, the heterogeneity of the eligible population, and the
varied study outcomes in the current evidence reduce the
power of the meta-analysis to detect statistically significant
differences in the health outcome of interest and to draw
meaningful conclusions. Fourth, the measures used to
describe the central tendency for many outcomes varied
between studies and included parametric and nonparametric measures for the same outcome. We assumed normal
distribution in some cases; however, concern always
remains when such assumptions are made.
Based on the findings of this review of 7 studies, UCM may
have beneficial effects; however, further studies are needed
before widespread use can be recommended, and its use at
present should be restricted to RCTs. Because DCC is recommended for widespread adoption,37 future studies of UCM
should, in most instances, consider DCC as the control intervention. Use of a more consistent study population, uniform
inclusion and exclusion criteria, and well-defined outcome
measures, including delineation of physiological status after
clamping, can help mitigate heterogeneity across studies to better support data synthesis and understand the effects of UCM
on important neonatal and childhood outcomes.
Conclusions
Umbilical cord milking was associated with some benefits and
no adverse effects in the immediate postnatal period in preterm infants (GA, <33 weeks). However, further studies are warranted to assess the effect of UCM on neonatal and long-term
outcomes.
Additional Contributions: Anup Katheria, MD,
Department of Neonatology, Sharp Mary Birch
Hospital for Women and Newborns, San Diego,
California, Serdar Alan, MD, Department of
Pediatrics, Division of Neonatology, Ankara
University School of Medicine, Ankara , Turkey, and
Shigeharu Hosono, MD, PhD, Department of
Paediatrics, Nihon University School of Medicine,
Tokyo, Japan, collaborated in providing data for this
review. They did not receive any compensation for
their contribution.
REFERENCES
1. McCausland AM, Holmes F, Schumann WR.
Management of cord and placental blood and its
effect upon the newborn. Calif Med. 1949;71(3):
190-196.
2. DeMarsh Q, Alt H, Windle W, Hills D. The effect
of depriving the infant of its placental blood on the
blood picture during the first week of life. JAMA.
1941;116(23):2568-2573.
3. Wilson E, Windle W, Alt H. Deprivation of
placental blood as a cause of iron deficiency in
infants. AJDC. 1941;62(2):320-327.
jamapediatrics.com
4. 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 Syst Rev. 2012;8:CD003248.
5. McDonald SJ, Middleton P, Dowswell T, Morris
PS. Effect of timing of umbilical cord clamping of
term infants on maternal and neonatal outcomes.
Cochrane Database Syst Rev. 2013;7:CD004074.
6. Siddall RS, Richardson RP. Milking or stripping
the umbilical cord: effect on vaginally delivered
babies. Obstet Gynecol. 1953;1(2):230-233.
7. Walsh SZ. Early clamping versus stripping of
cord: comparative study of electrocardiogram in
neonatal period. Br Heart J. 1969;31(1):122-126.
8. Lanzkowsky P. Effects of early and late clamping
of umbilical cord on infant’s haemoglobin level. BMJ.
1960;2(5215):1777-1782.
9. Usher R, Shephard M, Lind J. The blood volume
of the newborn infant and placental transfusion.
Acta Paediatr. 1963;52:497-512.
10. Colozzi AE. Clamping of the umbilical cord: its
effect on the placental transfusion. N Engl J Med.
1954;250(15):629-632.
JAMA Pediatrics Published online November 3, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ on 11/07/2014
E7
Research Original Investigation
Efficacy and Safety of Umbilical Cord Milking
11. Hosono S, Mugishima H, Fujita H, et al. Umbilical
cord milking reduces the need for red cell
transfusions and improves neonatal adaptation in
infants born at less than 29 weeks’ gestation:
a randomised controlled trial. Arch Dis Child Fetal
Neonatal Ed. 2008;93(1):F14-F19.
12. March MI, Hacker MR, Parson AW, Modest AM,
de Veciana M. The effects of umbilical cord milking
in extremely preterm infants: a randomized
controlled trial. J Perinatol. 2013;33(10):763-767.
13. Upadhyay A, Gothwal S, Parihar R, et al. Effect
of umbilical cord milking in term and near term
infants: randomized control trial. Am J Obstet Gynecol.
2013;208(2):120.e1-120.e6. doi:10.1016/j.ajog.2012
.10.884.
14. Rabe H, Jewison A, Alvarez RF, et al; Brighton
Perinatal Study Group. Milking compared with
delayed cord clamping to increase placental
transfusion in preterm neonates: a randomized
controlled trial. Obstet Gynecol. 2011;117(2, pt 1):
205-211.
15. Katheria AC, Leone TA, Woelkers D, Garey DM,
Rich W, Finer NN. The effects of umbilical cord
milking on hemodynamics and neonatal outcomes
in premature neonates. J Pediatr. 2014;164(5):
1045-1050.e1. doi:10.1016/j.jpeds.2014.01.024.
16. Alan S, Arsan S, Okulu E, et al. Effects of
umbilical cord milking on the need for packed red
blood cell transfusions and early neonatal
hemodynamic adaptation in preterm infants born
ⱕ1500 g: a prospective, randomized, controlled
trial [published online March 13, 2014]. J Pediatr
Hematol Oncol.
17. Erickson-Owens DA, Mercer JS, Oh W. Umbilical
cord milking in term infants delivered by cesarean
section: a randomized controlled trial. J Perinatol.
2012;32(8):580-584.
18. Liberati A, Altman DG, Tetzlaff J, et al. The
PRISMA statement for reporting systematic reviews
and meta-analyses of studies that evaluate health
care interventions: explanation and elaboration.
J Clin Epidemiol. 2009;62(10):e1-e34.
E8
19. 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 1500 gm. J Pediatr. 1978;92
(4):529-534.
20. Shennan AT, Dunn MS, Ohlsson A, Lennox K,
Hoskins EM. Abnormal pulmonary outcomes in
premature infants: prediction from oxygen
requirement in the neonatal period. Pediatrics.
1988;82(4):527-532.
21. Bell MJ, Ternberg JL, Feigin RD, et al. Neonatal
necrotizing enterocolitis: therapeutic decisions
based upon clinical staging. Ann Surg. 1978;
187(1):1-7.
22. Higgins J, Green S, eds. Cochrane Handbook for
Systematic Reviews of Interventions. Version 5.1.0
[updated March 2011]. The Cochrane Collaboration;
2011. http://handbook.cochrane.org/. Accessed
January 12, 2014.
23. Egger M, Smith GD, Altman DG. Systematic
Reviews in Health Care: Meta-Analysis in Context. 2nd
ed. London, England: BMJ Publishing; 2001:87-108.
24. Higgins JP, Thompson SG, Deeks JJ, Altman
DG. Measuring inconsistency in meta-analyses. BMJ.
2003;327(7414):557-560.
25. Hozo SP, Djulbegovic B, Hozo I. Estimating the
mean and variance from the median, range, and the
size of a sample. BMC Med Res Methodol. 2005;5:13.
doi:10.1186/1471-2288-5-13.
26. Milking the umbilical cord vs immediate
clamping in pre-term infants <33 weeks. NLM
identifier NCT01819532. http://clinicaltrials.gov/ct2
/show/NCT01819532. Accessed May 5, 2014.
27. Milking the umbilical cord for extreme preterm
infants. NLM identifier NCT01666847. http:
//clinicaltrials.gov/ct2/show/NCT01666847. Accessed
May 5, 2014.
28. The effect of cord milking on hemodynamic
status of preterm infants. NLM identifier:
NCT01487187. http://clinicaltrials.gov/ct2/show
/NCT01487187. Accessed May 5, 2014.
29. Gupta R, Ramji S. Effect of delayed cord
clamping on iron stores in infants born to anemic
mothers: a randomized controlled trial. Indian Pediatr.
2002;39(2):130-135.
30. Chaparro CM, Neufeld LM, Tena Alavez G,
Eguia-Líz Cedillo R, Dewey KG. Effect of timing of
umbilical cord clamping on iron status in Mexican
infants: a randomised controlled trial. Lancet.
2006;367(9527):1997-2004.
31. Lozoff B, Jimenez E, Hagen J, Mollen E, Wolf
AW. Poorer behavioral and developmental outcome
more than 10 years after treatment for iron
deficiency in infancy. Pediatrics. 2000;105(4):e51.
32. Takami T, Suganami Y, Sunohara D, et al.
Umbilical cord milking stabilizes cerebral
oxygenation and perfusion in infants born before
29 weeks of gestation. J Pediatr. 2012;161(4):742-747.
33. Kluckow M, Evans N. Low superior vena cava
flow and intraventricular haemorrhage in preterm
infants. Arch Dis Child Fetal Neonatal Ed. 2000;82
(3):F188-F194.
34. Chang YS, Ahn SY, Yoo HS, et al. Mesenchymal
stem cells for bronchopulmonary dysplasia: phase 1
dose-escalation clinical trial. J Pediatr. 2014;164(5):
966-972.e6. doi:10.1016/j.jpeds.2013.12.011.
35. Pierro M, Ionescu L, Montemurro T, et al.
Short-term, long-term and paracrine effect of
human umbilical cord-derived stem cells in lung
injury prevention and repair in experimental
bronchopulmonary dysplasia. Thorax. 2013;68(5):
475-484.
36. Carroll J. Human cord blood for the
hypoxic-ischemic neonate. Pediatr Res. 2012;71
(4, pt 2):459-463.
37. Committee on Obstetric Practice, American
College of Obstetricians and Gynecologists.
Committee opinion No. 543: timing of umbilical
cord clamping after birth. Obstet Gynecol. 2012;120
(6):1522-1526.
JAMA Pediatrics Published online November 3, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ on 11/07/2014
jamapediatrics.com