Premature infants born at GA 280/7- 33 6/7 weeks at Thammasat University Hospital, Pathumthani, Thailand, between July 1st, 2016, and December 31st, 2018 were enrolled in a randomized controlled trial by the delivery nurse team. We excluded infants who were multiples, diagnosed with severe disabilities, having chromosomal abnormalities, hydrops fetalis, intrauterine growth retardation, or from mothers with placenta previa with hemorrhage, abruption placenta, prolapsed cord, or having fetal distress before birth as well as those mothers who were giving birth before or on arrival or unable to give consent. This study was approved by the Human Research Ethics Committee No.1, Faculty of Medicine,Thammasat University. The informed consent was obtained before delivery by our research staff.
Randomization
Participants were divided into three groups (1:1:1) by a block-of-three randomization using sealed envelopes, performed by a research team. Envelopes were opened by the delivery nurse team, when mothers commenced preterm labor and entered the active phase, meaning regular uterine contractions at least four times in 20 minutes with cervix dilatation > 4 cm or when mothers had premature birth scheduled due to severe preeclampsia for either cesarean opearation or vaginal birth.
Placental transfusion techniques
DCC means delaying cutting the umbilical cord after birth for 45 seconds. DCM-B is DCC for 45 seconds with an obstetrician then performing UCM on about 25 cm length of the cord three times, next cutting the umbilical cord. DCM-A is DCC for 45 seconds with the umbilical cord cut to around 25 cm length; after this, the pediatrician milks the umbilical cord toward the infant three times, and the cord is next cut to standard. Digital clocks were used to time the procedure.
During these interventions, infants born by cesarean section were placed on their maternal abdomen, and those born by vaginal delivery, the obstetrician held the baby at the level of the maternal vaginal level.
After cutting the cord, the preterm infants were wrapped in a plastic bag and placed on a warm mattress in the radiant warmer. Neonatal resuscitation was decided by the attending physician at birth.
All infants received our standard care, which includes a starter iron supplement of 4 mg/kg/day after they can tolerate 120 ml/kg/day of milk for two days, an ultrasound brain screening within 14 and 28 days of life by a pediatric radiologist who was blinded to the intervention, jaundice assessment using serum microbilirubin, verification of hematocrit levels within one hour of birth and checked weekly until discharge. If they are discharged home before 28 days of life, there was an appointment to follow up and check hematocrit levels at the 4th week of age.
Outcomes
All outcomes were analyzed by blinded team researchers who did not know the code of the intervention. The primary outcomes were that premature infants received RBC transfusions within 28 days of age and had hematocrit levels checked on admission, at 1st and 4th week of age. Our pediatric team maintained the authority to determine if transfusions were required based on our guideline. Neonatal outcomes such as IVH (using grading system proposed by Papile et al.16), NEC, respiratory distress syndrome (RDS), patent ductus arteriosus (PDA), bronchopulmonary dysplasia (BPD) defined as oxygen requirement at 28 days of age, retinopathy of prematurity (ROP), and length of hospital stay (LOS) were determined. The morbidities including neonatal polycythemia, hyperbilirubinemia, neonatal death, maternal postpartum hemorrhage and death were collected.
Sample sizes
Katheria AC, et al.6 had 52% of the preterm infants in their DCC group receive blood transfusions. We, therefore, estimated 20% of infants in the DCM-A and DCM-B groups would require blood transfusions. Using 80% power and a confidence value of 0.05 with two-way calculation, 40 preterm infants were required for each group.
Statistical analysis
Demographic data was reported using percentage (%), mean, and SD median. ANOVA test was used to compare continuous variables that were normally distributed; Kruskall-Wallis was used for continuous variables not normally distributed. Pearson chi-square and Fisher’s exact test compared categorical outcome variables. p <0.05 is considered significant. Data analyses were performed using an intent-to-treat basis.