Statistics from Altmetric.com
We read with interest De Luca et al’s comparison of the Medick BiliMed and Respironics BiliChek transcutaneous bilirubinometers.1 We agree that BiliChek provides a reliable measure of skin bilirubin and it performed well in our ethnically diverse population. However, it is comparatively expensive. Therefore it is important to ask whether an alternative transcutaneous device is of sufficient clinical equivalence to BiliChek. The Konica Minolta JM-103 is such an alternative. The basic device has a similar price tag but does not require disposable tips, which considerably reduces the cost over time.
We have compared these two devices and assessed their usefulness in screening for infants likely to develop significant jaundice. We screened 235 well newborns (>35 weeks’ gestation) on the postnatal ward with a JM-103 and 45 of these newborns with a BiliChek at a median age of 21 h. The population screened was representative of the local population (60% Asian from the Indian subcontinent, 21% white, 14% black and 5% mixed race and other).
Babies with a transcutaneous reading (TcB) more than 220 μmol/l or above our local phototherapy chart treatment line had a serum bilirubin (TSB) measured. The correlation coefficients for JM-103 versus serum bilirubin in Asian, white and black babies, respectively, were 0.88, 0.75 and 0.90. Figures 1 and 2 show the relationship of JM-103 to BiliChek.
Overall the transcutaneous devices underestimated serum bilirubin: mean difference between JM-103 and TSB was −11 μmol/l (95% CI −16.9 to −5.2) and mean difference between BiliChek and TSB was −10 μmol/l (95%CI −19.0 to −1.6). Analysis of variance showed an effect of ethnicity on the difference between JM-103 and TSB measurements (F = 15.8, p<0.001) but not on the difference between BiliChek and TSB measurements (F = 0.06, p = 0.94). JM-103 underestimated serum bilirubin in white babies (mean difference −38 μmol/l (95% CI −46.3 to −29.7) and overestimated serum bilirubin in black babies (mean difference 28µmol/l (95% CI 17.2 to 38.8)). In the Asian population, JM-103 gave a closer approximation to the serum bilirubin than BiliChek (mean difference −1 μmol/l (95% CI −7.9 to 5.9) for JM-103 and TSB, and −10 μmol/l (95% CI −23.5 to 3.5) for BiliChek and TSB).
The American Academy of Pediatrics recommends plotting a baby’s predischarge bilirubin measurement (serum or transcutaneous) on an hour-specific nomogram to assess their risk of significant jaundice.2 We plotted initial transcutaneous readings on Bhutani’s hour-specific nomogram3 4 and assessed the ability of the two devices to predict the need for phototherapy (positive predictive value (PPV)) and to identify those babies safe to send home (negative predictive value (NPV)). The PPV for a BiliChek reading above Bhutani’s 75th centile was 35%, for JM-103 it was 29%. The NPV for BiliChek was 96%, with JM-103 it was 98%. There was therefore no difference in the ability of the two devices to identify low-risk or high-risk babies.
We feel that, although BiliChek provides a more accurate measurement of bilirubin across all ethnic groups, in our population the JM-103 is clinically equivalent. We suggest clinicians take ethnicity into account when interpreting JM-103 results.
Competing interests: None.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.