Comparing a low-cost, hand-held icterometer to transcutaneous and serum measurement of neonatal bilirubin among dark-skinned infants: Can jaundice screening be improved in low-resource settings?

Abstract Neonatal jaundice remains a challenge globally, contributing to significant newborn morbidity and mortality in low- and middle-income countries. This proposal, “Comparing a low-cost, hand-held icterometer to transcutaneous and serum measurement of neonatal bilirubin among dark-skinned infants: Can jaundice screening be improved in low-resource settings?” builds on previous validation research of the Bili-RulerTM, a hand-held icterometer developed at Harvard University. The Bili-Ruler has 6 digitally standardized strips of increasingly yellow hue that allow for comparison against blanched newborn skin. Previous research demonstrated acceptable diagnostic testing metrics in measuring elevated bilirubin, yet data are limited by a small sample size in patients of African descent, raising questions as to how effective the Bili-Ruler is among dark-skinned infants. We propose to conduct a robust validation study among newborns in Kumasi, Ghana, in West Africa (where 99% of the population is from one of 10 African ethnic groups) to address this gap in understanding. We aim to test the Bili-Ruler against both transcutaneous (TCB) and serum (TSB) bilirubin measurements among Ghanaian newborns to determine if this low-cost, hand-held device is a viable option to improve jaundice screening and monitoring in sub-Saharan Africa. Our study includes the following aims: Aim 1: To determine screening accuracy of the Bili-Ruler, we will compare healthcare providers’ Bili- Ruler assessments with measures of TCB on 2000 newborns and TSB on 400 newborns. We will use gestational age at birth, hours since birth, and risk factors for jaundice to determine if TCB and TSB levels are above the recommended threshold for treatment, and we will compare findings against newborns judged to have Bili-Ruler scores of 3 or higher. We will calculate sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under the curve (AUC) for the Bili-Ruler against TCB and TSB. Aim 2: To determine differences in the Bili-Ruler’s accuracy based on both underlying skin tone and severity of jaundice, we will use scores on the Monk Skin Tone Scale (range: 1-10) and increasing thresholds on the Bili-Ruler (e.g. 3+, 4+, 5+) to explore differences in sensitivity, specificity, PPV, NPV and AUC for the Bili-Ruler against TCB and TSB, using healthcare providers’ assessments. Aim 3: To determine inter-rater reliability, we will compare healthcare providers’, researchers’ and mothers’ scores on the Bili-Ruler using correlation, Kappa, and McNemar’s statistical tests. This research is innovative in that it seeks to determine whether a simple, low-tech, hand-held “ruler” could be used as a valid screening tool for neonatal jaundice. If so, the implications for jaundice screening in low- resource settings is significant, given that screening could then be conducted at lower-level facilities or in remote communities without access to TCB and TSB assessments.