Abstract
BACKGROUND: Neonatal hyperbilirubinemia (NHB) is a significant cause of morbidity and mortality, particularly in low and middle-income countries (LMICs). Transcutaneous bilirubinometers offer a non-invasive method for assessing NHB but have limited availability due to cost and maintenance requirements. Visual assessment of jaundice is shown to be inaccurate. Smartphone-based technologies have the potential to provide innovative and accessible healthcare solutions. This study aimed to evaluate the Picterus system, a smartphone-based tool for screening of NHB, in three non-Caucasian populations in LMICs. METHODS: Between 2018 and 2022, cross-sectional studies were conducted in three countries: Mexico, Nepal and the Philippines. Newborns meeting the inclusion criteria were recruited, and data on demographic characteristics, skin type and visual assessment of jaundice were collected. Bilirubin levels were measured using both the Picterus system and total serum bilirubin (TSB) analysis. Correlation analyses, Bland-Altman plots and receiver operating characteristic (ROC) curves were used to evaluate the Picterus system. RESULTS: A total of 416 infants were included in the analysis. The Picterus smartphone system demonstrated a significant positive correlation with TSB levels across all sites (r=0.76). The correlation coefficient was significantly higher in Mexico compared with Nepal and the Philippines. Bland-Altman plots showed limits of agreement ±89.2 µmol/L. Picterus values were underestimated in Mexico, whereas they were overestimated in Nepal and the Philippines. ROC analysis for detection of infants with TSB >225 µmol/L indicated that the Picterus system had higher sensitivity and specificity compared with visual assessment using the Kramer scale. DISCUSSION: This study shows that the Picterus system can potentially be used in screening for neonatal jaundice in populations with moderate dark skin types. Further studies are needed before the system can be used in clinical practice.