1. Objectives

To evaluate the short and long term effects of early as compared to delayed clamping of the umbilical cord on preterm infants born at <37 weeks’ gestation and their mothers. Additional objectives are to assess the effect of milking of the umbilical cord and the effect of positioning the infant above or below the introitus at the time of birth

2. How studies were identified

The Cochrane Pregnancy and Childbirth Group’s Trials Register was searched in June 2012, which included records identified from the following databases:

• CENTRAL
• MEDLINE
• EMBASE

In addition, relevant conference proceedings and journals were handsearched

3. Criteria for including studies in the review

3.1 Study type

Randomized controlled trials including cluster-randomized trials

3.2 Study participants

Preterm infants born prior to 37 completed weeks’ gestation and their mothers

3.3 Interventions

Immediate clamping of the umbilical cord versus delayed clamping after 30 seconds or more. The intervention could include: i) holding the baby’s head above or below the level of the placenta; ii) oxytocin administration to the mother; or iii) milking of the umbilical cord towards the infant

(Studies that assessed cord milking with clamping earlier than 30 seconds were also included)

3.4 Primary outcomes

Infant

• Death: prior to discharge from hospital, post discharge from hospital, and total deaths
• Death or neurosensory disability at age two to three years
• Intraventricular haemorrhage via ultrasound diagnosis grade three and four
• Periventricular leukomalacia

Maternal

• Postpartum haemorrhage (blood loss >500 mL)

Secondary outcomes for the infant at birth included: requirement for resuscitation; Apgar score at one, five, and ten minutes; and hypothermia. Other infant outcomes included respiratory outcomes (respiratory distress syndrome during the first 36 hours, use of exogenous surfactant, oxygen dependency, chronic lung disease), cardiovascular outcomes (fluid administration or inotropic administration for hypotension in the first 24 hours, treatment for patent ductus arteriosus), haematological outcomes (treatment of anaemia with blood transfusion, treatment of hyperbilirubinaemia with blood exchange transfusion or phototherapy, haemoglobin and ferritin at six and 12 months), central nervous system outcomes (all grades of intraventricular haemorrhage), gastrointestinal outcomes (necrotising enterocolitis), and length of hospital stay. Secondary outcomes pertaining to the mother included death, manual removal of the placenta, effects on Rhesus-immunisation, psychological wellbeing, bonding with the infant, anxieties, and views; whilst for the father outcomes included psychological wellbeing, bonding with the infant, anxieties, and views.

4. Main results

4.1 Included studies

Fifteen randomized controlled trials, enrolling 738 children, were included in this review

• All trials enrolled preterm infants born between 24 and 36 weeks’ gestation
• For most studies, infant-mother dyads were randomized, although in some studies the unit of randomization was the baby
• Recruited women expected to give birth to preterm infants
• Two studies included only vaginal births, and in one study infants were delivered by exclusively via caesarean section. All 12 remaining studies included women giving birth by both methods
• Included studies investigated a range of interventions aiming to modify placental transfusion, namely, early cord clamping (up to 20 seconds), delayed cord clamping (with a range of 30 to 120 seconds), cord milking, the position of the baby in relation to the placenta, and the use of uterotonic drugs (syntocinon or ergometrine)

4.2 Study settings

• Australia, Germany, Israel, Japan, the Netherlands, South Africa (2 trials), Switzerland, the United Kingdom of Great Britain and Northern Ireland (3 trials), and the United States of America (4 trials), all in hospital settings

4.3 Study settings

How the data were analysed
Less placental transfusion was compared to more placental transfusion in preterm infants. In 14 of the 15 studies, this comparison was made between immediate cord clamping (less placental transfusion) and delayed cord clamping (more placental transfusion), while in one study no milking of the umbilical cord (less placental transfusion) was compared to milking of the cord (more placental transfusion). Both types of comparison were pooled according to whether there was less or more placental transfusion, and subgroup analyses investigated differences between delayed cord clamping and milking of the cord. Where studies were judged to be estimating the same underlying effect, fixed effects meta-analysis was conducted, whilst for studies where substantial clinical variation was suspected, random effects models were used. Mean differences (MD) and standardized mean differences and 95% confidence intervals were employed for continuous data, and risk ratios (RR) with 95% confidence intervals were used for categorical outcomes. Sensitivity analyses were planned for primary outcomes by risk of allocation concealment bias. Investigation of heterogeneity was planned for primary outcomes using the following subgroup analyses:

• By position of the baby relative to the level of the placenta before cord clamping: below the level of the placenta; at the level of the placenta; above the level of the placenta; or position not known/unclear
• By whether the woman was given oxytocin as a uterotonic drug before cord clamping: with oxytocin; no oxytocin before clamping; unclear/not known whether oxytocin was given
• By milking of the cord: with milking; without milking; unclear/not known whether milking occurred
• By birth route: vaginal; abdominal; mixed or not known
• By gestational age at birth: <32 completed weeks’ gestation; ≥32 completed weeks’ gestation; gestational age <32 and ≥32 weeks or not known

Results
Delayed (more placental transfusion) versus immediate (less placental transfusion) cord clamping (15 trials/738 infants)
Infant primary outcomes
Infant death
Overall, the group receiving greater placental transfusion reported fewer infant deaths, with 3.1% (10 deaths/319 infants) in comparison to 4.9% (17 deaths/349 infants) among the group receiving less placental volume. However, this was not statistically significant (RR 0.63, 95% CI [0.31 to 1.28], p=0.20; 13 studies/668 infants).

Severe intraventricular haemorrhage
A non-statistically significant 32% reduction in the risk of severe intraventricular haemorrhage was observed among infants receiving a larger placental transfusion (RR 0.68, 95% CI [0.23 to 1.96], p=0.47; 6 studies/305 infants; 12 events in total).

Periventricular leukomalacia
In the two studies reporting on periventricular leukomalacia, only three events occurred (RR 1.02, 95% CI [0.19 to 5.56], p=0.98; 71 infants).

Death or neurosensory disability at two to three years of age
No studies reported on death or neurosensory disability at two to three years of age.

Pooled results of primary outcomes did not differ in subgroup analyses by delayed cord clamping versus milking of the cord, or by risk of allocation concealment bias.

Maternal primary outcomes
Postpartum haemorrhage (blood loss >500 mL)
For this outcome, none of the included studies reported any results.

Additional outcomes
Mean arterial blood pressure
Infants receiving a larger placental transfusion were found to have a significantly higher mean arterial blood pressure at birth compared to those receiving less placental transfusion, with a MD of 3.52 mmHg (95% CI [0.60 to 6.45], p=0.018; 2 studies/97 infants). The same was true at four hours post-delivery, where infants receiving a larger placental transfusion were found to have a significantly higher mean arterial blood pressure (MD 2.49 mmHg, 95% CI [0.26 to 4.72], p=0.029; 2 studies/111 infants), although the difference had disappeared by 24 hours of age (MD -0.30 mmHg, 95% CI [-6.44 to 5.84], 1 trial/38 infants).

Need for inotropic support
Need for inotropic support was reduced by 58% among infants receiving greater placental transfusion (RR 0.42, 95% CI [0.23 to 0.77], p=0.0045; 5 studies/158 infants).

Intraventricular haemorrhage, all grades
Overall, the group receiving more placental transfusion was at reduced risk of intraventricular haemorrhage of any grade (RR 0.59, 95% CI [0.41 to 0.85], p=0.0048; 10 studies/539 infants).

Necrotising enterocolitis
Infants receiving more placental transfusion were at reduced risk of necrotising enterocolitis (RR 0.62, 95% CI [0.43 to 0.90], p=0.011; 5 studies/241 infants).

Blood transfusions for the treatment of anaemia
Overall, of those infants receiving more placental transfusion, 24% required a blood transfusion for anaemia, compared to 36% of those receiving less placental transfusion (RR 0.61, 95% CI [0.46 to 0.81], p=0.0005; 7 studies/392 infants). Similarly, those infants receiving more placental transfusion exhibited a reduction in the overall number of blood transfusions required (MD -1.26 transfusions, 95% CI [-1.87 to - 0.64], p=0.00006; 5 studies/210 infants,).

Hyperbilirubinaemia (jaundice)
Peak serum bilirubin concentrations for infants receiving more placental transfusion was found on average to be 15.01 mmol/L higher (95% CI [5.62 to 24.20], p=0.0017; 7 studies/320 infants) compared to those receiving less placental transfusion. The risk of treated hyperbilirubinaemia was not significantly different between groups (RR 1.21, 95% CI [0.94 to 1.55], 3 trials/180 infants).

Haematocrit, blood volume, and red cell mass
Infants receiving more placental transfusion had greater haematocrit at birth or at one hour post-delivery (MD 3.26%, 95% CI [1.79 to 4.74], p<0.0001; 7 studies/318 infants), and at four hours (MD 5.40%, 95% CI [3.62 to 7.17], p<0.00001; 5 studies/173 infants), and 24 hours of age (MD 3.28%, 95% CI [1.34 to 5.22], p=0.0009; 3 studies/199 infants). Blood volume after birth was statistically significantly greater after birth among infants receiving more placental transfusion (MD 8.41, 95% CI [2.14 to 14.67], p=0.0086; 2 studies/81 infants), as was red cell mass (MD 5.30, 95% CI [0.05 to 10.55], p=0.048; 1 study/35 infants).

Mean regional tissue oxygenation of the brain
At both four hours and 24 hours of age, mean regional tissue oxygenation of the brain was improved in those infants receiving greater placental transfusion (MD 6.44, 95% CI [5.47 to 7.41], p<0.00001; 1 study/38 infants, and MD 4.29, 95% CI [3.44 to 5.14], p<0.00001; 1 study/38 infants, respectively).

For all other outcomes, no clear differences were identified between groups, or no studies reported data.

5. Additional author observations*

Methodological quality of the included studies was poor, with the majority of studies having largely unclear risk of bias, and only two trials at low risk of allocation concealment bias. Selective outcome reporting was also cited as a concern to the review authors. Clinical practice has changed over time, substantially affecting the risk of infant death, and in some of the trials carried out in low-income settings ventilator support for preterm infants was limited in comparison to modern practices in high-income settings. Similarly, studies carried out in high-income settings reported a low incidence of intraventricular haemorrhage, reducing the power to detect any statistically significant effect of greater placental transfusion on this risk.

Overall, the combined trial results should be interpreted with caution as the vast majority of outcomes had wide confidence intervals. However, delayed cord clamping (more placental transfusion) was associated with a reduction in the risk of all grades of intraventricular haemorrhage, a reduction in the need for transfusions for anaemia or low blood pressure, and a reduced risk of necrotising enterocolitis.

While the findings of this review provide some support for allowing more placental transfusion, the studies to date are small, are at risk of bias, suffer from incomplete reporting of data, report on few outcomes and have no long-term follow up data on neurodevelopmental outcomes. Therefore, there is a need for further large, high quality, multicentre studies that include data on respiratory outcomes, long-term neurodevelopmental outcomes, and maternal outcomes.