Objectives: To assess the correlation between total serum bilirubin (TSB) and transcutaneous bilirubin
(TcB) values in Malawian newborn infants, and to investigate whether TcB can be used safely to guide
phototherapy treatment in the absence of TSB results.
Methods: 128 newborn jaundiced infants were studied in the neonatal nursery at Queen Elizabeth Central
Hospital, Blantyre. Paired TSB and TcB measurements (from forehead and sternum) were compared using
the linear regression and Bland–Altman methods. Clinical decisions based on TcB results were compared
with those based on ‘gold standard’ TSB results.
Results: For infants not under phototherapy, the lowest TcB reading (from forehead or sternum) gave the
strongest correlation with TSB: r50.83 for term infants and r50.71 for premature infants. For infants
undergoing phototherapy, the highest TcB reading gave the strongest correlation with TSB: r50.66 for term
infants and r50.71 for premature infants. TcB values overestimated the degree of jaundice. For infants not
under phototherapy, the mean (SD) bias and imprecision of TcB was 25 (72) mmol/L for term infants and 37
(73) mmol/L for premature infants. For infants under phototherapy, the mean bias and imprecision was 30
(79) mmol/L for term infants and 44 (77) mmol/L for premature infants. For infants not under phototherapy,
using the lowest TcB reading to guide treatment decisions (n5167) had good sensitivity (91%) and
specificity (90%). For infants undergoing phototherapy, using the highest TcB reading to guide treatment
decisions (n5129) had good sensitivity (94%), but lower specificity (36%).
Conclusions: TcB can be used to safely guide phototherapy treatment in a resource-poor setting.