Page 144 - 16Neonatal Jaundice_compressed

Treatment

randomisation used as a computer-generated sequence and the other study 152 used sealed
envelopes for allocation concealment.
The mean gestational ages of the study samples were 27.9 ± 1.4 weeks and 30.4 ± 2.7 weeks, the
mean birthweights were 1019 ± 283 g and 1518 ± 419 g, the mean ages at entry to the study were
38.3 ± 7.1 hours and 58 ± 25.8 hours, and the mean baseline serum bilirubin levels were
109 ± 5 micromol/litre and 168 ± 49 micromol/litre. In all, 107 (51.9%) of participants were male.
There were no statistically significant differences between the groups in terms of the number of
exchange transfusions, number of treatment failures or frequency of rebound jaundice.
Review findings

As the two studies used different populations (preterm and very preterm) and different time-
points for measuring the change in serum bilirubin, it was not possible to pool the results. One
study 151 measured serum bilirubin at 18 hours after initiation of phototherapy and this showed
no statistically significant difference between conventional phototherapy and multiple
phototherapy. The second study, 152 which measured change in serum bilirubin over 72 hours,
found a statistically significant difference in favour of multiple phototherapy
(MD = 11.00 micromol/litre, 95% CI 9.01 to 12.99 micromol/litre).
See below for the overall evidence summary and GDG translation from evidence for
phototherapy in preterm/low-birthweight babies.

Conventional phototherapy versus fibreoptic phototherapy

Description of included studies
Six studies 152-157 with a total of 398 participants were included in this comparison. Four studies were
154
156
carried out in Italy 152;153;155;157 and one apiece in Australia and the Netherlands. The evidence
156
level of all included studies was EL 1+. One study specified the method of randomisation used
as the lottery method while the remaining five studies 152-155;157 used sealed envelopes.
Where reported, the mean gestational age ranged from 27.9 ± 1.4 weeks to 34.4 ± 1.2 weeks,
mean birthweight ranged from 1019 ± 283 g to 2600 ± 382 g, mean age at entry to study ranged
from 26.5 ± 15.0 hours to 63.2 ± 17.8 hours, and mean baseline serum bilirubin levels ranged
from 94 ± 36 micromol/litre to 241 ± 9 micromol/litre. In the studies that reported gender, 162
(54.1%) participants were male.

Review findings
In a meta-analysis of these six studies, there was no statistically significant difference in the
number of exchange transfusions carried out: only five babies who received conventional
phototherapy and only seven babies who received fibreoptic phototherapy required exchange
transfusions. There were no statistically significant differences for treatment failure (defined as
requiring double phototherapy or reaching a predefined serum bilirubin level) between
conventional and fibreoptic groups. No study reported cases of rebound jaundice.

Three studies contributed data on the mean decrease in serum bilirubin: there was no
statistically significant difference between the groups (MD = −1.17 micromol/litre, 95% CI
−3.87 to 1.53 micromol/litre). There was no heterogeneity (I² = 0%). Four studies contributed
data on the mean duration of phototherapy and there was a statistically significant difference
between the groups in favour of fibreoptic phototherapy (MD = 2.63 hours, 95% CI 0.69 to
4.58 hours). There was no heterogeneity (I² = 0%).
See below for the overall evidence summary and GDG translation from evidence for
phototherapy in preterm/low-birthweight babies.
Conventional phototherapy versus LED phototherapy

Description of included studies

Two studies 158;159 with 119 participants were included in this comparison. One study was
carried out in Brazil 158 and the other in Italy. 159 The evidence level in one study 158 was EL 1−

113