Page 113 - 16Neonatal Jaundice_compressed
P. 113
Neonatal jaundice
measurements were taken from the forehead simultaneously with blood sampling. The median
transcutaneous measurement was 209 micromol/litre. The BiliChek overestimated diazo serum
bilirubin by a mean of 34 micromol/litre and VITROS serum bilirubin by a mean of
22 micromol/litre. There was a moderately positive correlation between transcutaneous
bilirubin and serum bilirubin values: diazo (r² = 0.65) and VITROS (r² = 0.66) when bias was
accounted for. [EL III]
Evidence summary – BiliChek
Evidence from good-quality studies indicates that transcutaneous bilirubin measurement from the
forehead using BiliChek correlates moderately well with serum bilirubin values in term and near-
term babies. The correlation coefficient ranged from 0.80 to 0.87. In a study in healthy preterm
babies it was 0.79 whereas in another study in preterm and sick term babies it was 0.88.
BiliChek was less accurate at bilirubin levels greater than 250 micromol/litre. Results from two
studies have reported an increase in the mean difference between serum bilirubin and BiliChek
readings with a rise in bilirubin levels. One study found the BiliChek underestimated serum
bilirubin in healthy term and near-term babies while two studies reported overestimation in
healthy term and preterm babies. Altough there were differences in the populations studied, in
threshold cut-off values of transcutaneous bilirubin and in the levels of laboratory serum
bilirubin used as the reference test, the sensitivity of BiliChek to detect bilirubin levels was
generally reported to be high, with variable results for the specificity. In the study on African
babies, BiliChek readings showed a reasonable correlation with serum bilirubin values but the
difference between transcutaneous bilirubin and serum bilirubin was greatest in babies with
darker skin tones.
GDG translation from evidence – BiliChek
High-quality research suggests that forehead or sternum measurement of transcutaneous
bilirubin by BiliChek is more accurate than visual inspection when used to assess the degree of
jaundice in term and near-term babies with a range of skin tones.
Good-quality studies of BiliChek measurement in preterm babies show a significantly positive
correlation with serum bilirubin but there are no studies that report the use of the BiliChek in
babies with gestational age less than 30 weeks. The GDG considered that, given the lack of
evidence regarding babies of less than 30 weeks of gestation, they could not recommend the
use of the BiliChek in very preterm babies.
BiliChek is less accurate at higher levels of bilirubin and in babies with dark skin tones.
Recommendations
See the end of Section 5.2.
Cost-effectiveness evidence for transcutaneous bilirubinometers
Alternative testing strategies for hyperbilirubinaemia was identified by the GDG as a priority for
an economic analysis. The results are summarised below; further details are available in
Appendix C.
The GDG considered that there were two alternative testing strategies to ‘current practice’ in the
NHS. These two strategies were to either perform a serum bilirubin on all visually jaundiced
babies or undertake a transcutaneous bilirubin measurement on all visually jaundiced babies,
with a serum bilirubin measurement on those with transcutaneous bilirubin estimations above a
certain threshold. They judged that, under their recommended thresholds for treatment (a
relatively high threshold) and further monitoring (a relatively low threshold), either alternative
would be equally effective at preventing cases of kernicterus. Therefore, a cost-minimisation
analysis was undertaken to compare these alternatives. There is insufficient clinical evidence to
determine whether more intensive testing for hyperbilirubinaemia using one of these two
strategies would be more cost-effective than ‘current practice’, in which visual examination is
often used to determine the severity of hyperbilirubinaemia, with less than 10% of visually
jaundiced babies having a serum bilirubin. However, there is very good evidence to show that
visual examination is not reliable in assessing the degree of hyperbilirubinaemia in a jaundiced
84
