Page 111 - 16Neonatal Jaundice_compressed

Neonatal jaundice

transcutaneous bilirubin measurements for predicting serum bilirubin determined by the HPLC-
B method. At the threshold value of 187 micromol/litre, transcutaneous bilirubin had a
sensitivity of 93% and specificity of 73% for detecting serum bilirubin > 222 micromol/litre. At
a threshold of 240 micromol/litre, transcutaneous bilirubin had a sensitivity of 90% and
specificity of 87% in detecting serum bilirubin levels > 290 micromol/litre. Transcutaneous
bilirubin measurements showed similar diagnostic accuracy results (sensitivity and specificity)
for detecting hyperbilirubinaemia (serum bilirubin values from HPCL-B method
> 290 micromol/litre). [EL Ib]
In the second observational study, from Malaysia, 345 healthy term babies from different
78
ethnic backgrounds (Malays 63.8% Chinese 30.7% and Indians 5.5%) were studied to assess
whether transcutaneous bilirubin measurement using BiliChek could accurately detect severe
hyperbilirubinaemia. All babies requiring investigation for jaundice had forehead and sternum
transcutaneous bilirubin levels measured within 30 minutes of venous blood being collected for
serum bilirubin estimation. The laboratory technicians were blinded to the transcutaneous
bilirubin readings. The prevalence of severe hyperbilirubinaemia (serum bilirubin
> 300 micromol/litre) in the sample population was 27.5% (95 of 345). The correlation
between the laboratory serum bilirubin levels and transcutaneous bilirubin readings was strong
and statistically significant for forehead and sternum (r = 0.80 and 0.86, respectively; P < 0.001
for both). Minor variation was observed in correlation coefficients for the three ethnic groups,
with the values ranging between 0.79 and 0.84 at the forehead and 0.86 and 0.94 at the
sternum. When these data were segregated according to the timing of serum bilirubin and
transcutaneous bilirubin, the correlation at less than 80 hours of age (r = 0.85) was better than
that seen after 80 hours (r = 0.71) but 79% of the babies with severe hyperbilirubinaemia had
their serum bilirubin estimation done after 80 hours of age.
Forehead transcutaneous bilirubin readings (threshold 250 micromol/litre) had a sensitivity of
100% and specificity of 39% for detecting serum bilirubin levels > 300 micromol/litre, while
the values were 76% and 85% at a transcutaneous bilirubin threshold of 260 micromol/litre. For
sternum transcutaneous bilirubin, the sensitivity and specificity at a threshold of
200 micromol/litre were 100% and 34%, respectively while at a threshold of 280 micromol/litre
the values were 93% and 84%, respectively. When the difference between serum bilirubin and
transcutaneous bilirubin was plotted against the mean serum bilirubin and transcutaneous
bilirubin measurements, the difference widened markedly from the line of agreement at the
mean level of serum bilirubin and transcutaneous bilirubin above 250 micromol/litre, especially
when transcutaneous bilirubin was measured from the forehead. Moreover, the areas under the
curves for different serum bilirubin levels (≥ 250 micromol/litre, ≥ 280 micromol/litre and
≥ 300 micromol/litre) were slightly but consistently larger for the sternum readings compared
with the forehead readings. [EL Ib]
45
In a Danish study, the diagnostic accuracy of BiliChek was evaluated in both sick and healthy
newborn babies. A total of 488 babies comprised the sample population – both preterm babies
with gestational age< 35 weeks and sick term and near-term babies in the NICU formed
Group 1 (n = 261, with mean birthweight 2521 g) while Group 2 was made up of healthy term
and near-term babies with gestational age ≥ 35 weeks in the maternity ward (n = 227, with
mean birthweight 3362 g). Exclusion criteria were well defined but blinding was not specified.
Transcutaneous bilirubin was measured with BiliChek on the forehead, sternum, knee and foot,
following which capillary blood was drawn for laboratory serum bilirubin estimation. In
Group 1 babies, the correlation coefficients for serum bilirubin levels and transcutaneous
bilirubin from the forehead and sternum were high (0.88 and 0.82), while they were 0.77 for
the knee and only 0.51 for the foot. In Group 2, readings from the sternum showed the strongest
correlation (0.90), while it was 0.87 for the forehead, 0.83 for the knee and 0.67 for the foot.
Based on these results, the forehead was recommended as the preferred site for transcutaneous
bilirubin measurement. Although exact data were not given for Bland–Altman analysis, figures
from both groups showed that transcutaneous bilirubin from the forehead underestimated serum
bilirubin levels and this underestimation increased as the serum bilirubin level increased. The
diagnostic accuracy of transcutaneous bilirubin for detecting serum bilirubin levels, where
phototherapy was indicated according to the Danish Pediatric Society guidelines
(www.paediatri.dk/), was also determined. Using a screening threshold for transcutaneous

82