Page 112 - 16Neonatal Jaundice_compressed

Recognition

bilirubin from the forehead as 70% of the serum bilirubin limit (300 micromol/litre or 10% of
bodyweight in grams for ill babies and 50 micromol/litre higher for healthy babies), the
sensitivity and specificity in Group 1 babies was 99% and 45%, and for Group 2 100% and
81%, respectively. [EL III]
44
The fourth study was conducted in the UK in a regional teaching hospital and included all
babies in the postnatal ward who were having blood taken for serum bilirubin estimation. A
concurrent transcutaneous bilirubin reading (using BiliChek) was taken but the site was not
specified. A total of 300 babies with gestational age ranging from 33 to 42 weeks were included
in this study. Of these, 18.3% of them had serum bilirubin levels > 250 micromol/litre.
Significant correlation was seen between serum bilirubin levels and transcutaneous bilirubin
readings (r = 0.77; P < 0.001). Although the BiliChek underestimated serum bilirubin levels by
a small value (mean difference −10.6 micromol/litre), the confidence intervals of the difference
were wide, ranging from −80 to +60 micromol/litre. This discrepancy was not found to
increase with rises in bilirubin levels. With a threshold value of > 195 micromol/litre,
transcutaneous bilirubin measurements using BiliChek could detect serum bilirubin levels
> 250 micromol/litre with a sensitivity of 91% and a specificity of 66%. [EL II]
79
The fifth study was conducted in Italy to evaluate BiliChek in preterm babies. The study
population comprised 340 preterm babies with gestational age between 30 to 36 weeks
admitted to the neonatal unit of a tertiary hospital. The mean birthweight of the sample was
2145 ± 518 g. The unit followed a policy of daily bilirubin monitoring for all preterm babies in
the first 120 hours of life. After randomly selecting one of these observations, transcutaneous
bilirubin was measured from the forehead about 10 minutes before drawing blood for serum
bilirubin estimation. All transcutaneous bilirubin measurements were made by the same
investigator, who was blinded to the serum bilirubin results. The correlation coefficient between
the two measurements was 0.79 (P < 0.01). The BiliChek reading overestimated serum
bilirubin level by more than 8.5 micromol/litre in 61% of the sample (209/340), with a mean
difference of 18.8 micromol/litre. This difference was found to increase at higher levels of
bilirubin. The most effective transcutaneous bilirubin threshold values were 111 micromol/litre
to detect serum bilirubin levels > 171 micromol/litre (sensitivity 100% and specificity 40%)
and 171 micromol/litre to detect serum bilirubin levels > 205 micromol/litre (sensitivity 100%
and specificity 72%). [EL Ib]
80
In the sixth study, from Nigeria, transcutaneous bilirubin measurements with BiliChek were
correlated with serum bilirubin values in a group of African babies with varying degrees of skin
pigmentation. The study was conducted at two hospitals; one in a rural setting and the other a
tertiary teaching hospital. The study population comprised 127 term and preterm babies with
jaundice. Transcutaneous bilirubin measurements were taken from the forehead simultaneously
with blood sampling before phototherapy was started. Skin pigmentation was determined by
visual observation and classified as light (54% of babies), medium (36%) and dark (10%).
Transcutaneous bilirubin measurements at the forehead correlated well with the serum bilirubin
values (r = 0.92; P < 0.001) when the data were combined from the two hospitals, and the
mean difference was 8.5 ± 129.2 micromol/litre. When the data were segregated according to
serum bilirubin, correlation for serum bilirubin ≥ 205 micromol/litre was better compared with
serum bilirubin levels < 205 micromol/litre (r = 0.84 versus 0.67). At serum bilirubin levels
≥ 205 micromol/litre, transcutaneous bilirubin measurements underestimated serum bilirubin
with a mean difference of 21.4 micromol/litre, but overestimated it when serum bilirubin levels
were < 205 micromol/litre (mean difference of 35.7 micromol/litre). When the data were
analysed on the basis of skin pigmentation, transcutaneous bilirubin measurements correlated
strongly with all three degrees of pigmentation. Although the mean difference between
transcutaneous bilirubin and serum bilirubin readings was small (8.5 micromol/litre), the
imprecision (standard deviation) increased with increasing degree of pigmentation:
92 micromol/litre for light, 133 micromol/litre for medium, and 197 micromol/litre for dark
pigmentation. [EL II]

81
In the last study, from the USA, transcutaneous bilirubin measurements with BiliChek were
compared with serum bilirubin levels obtained by the diazo and the VITROS® methods. The
study was conducted in a well-baby nursery at a general hospital. The study population
comprised 177 term and preterm babies with suspected jaundice. Transcutaneous bilirubin

83