Accuracy of non-invasive smartphone bilirubin measurements on neonates before and during phototherapy

Abstract Purpose Previous publications have reported the accuracy of the Picterus Jaundice Pro (Picterus JP), a smartphone-based medical device app for screening neonatal jaundice prior to phototherapy. The purpose of this study was to evaluate the accuracy of the Picterus JP app before and during phototherapy. Methods Between April 2025 and April 2026, newborns whose total serum bilirubin concentration (TSB) reached or exceeded the American Academy of Pediatrics phototherapy thresholds were enrolled in the study after written informed consent was obtained from their parent or legal guardian. A Picterus JP scan was obtained from the chest of the newborn (within 60 minutes of the TSB blood sample collection), a light-occlusive patch was placed on the newborn’s chest, and phototherapy was initiated. Then, up to five Picterus JP scans per newborn were obtained from the newborn’s chest, from the region covered by the phototherapy patch, within one hour of the blood collection for subsequent TSBs. We calculated diagnostic accuracy statistics to summarize the accuracy of Picterus JP. Results A total of 137 paired Picterus JP–TSB measurements were analyzed. Overall, Picterus JP correlated well with TSB (r = 0.720; bias − 38.5 µmol/L; RMSE 73.7 µmol/L). Restricted to a phototherapy course of ≤ 3 days, correlation was strong before phototherapy (n = 45 measurements; r = 0.849; bias − 46.5 µmol/L; RMSE 65.3 µmol/L) and weaker during phototherapy (n = 83 measurements; r = 0.520; bias − 40.7 µmol/L; RMSE 79.6 µmol/L), although the magnitude of bias and RMSE was similar in both periods. The relative (fractional) decline in Picterus JP during phototherapy correlated well with the relative decline in TSB (n = 97 paired change measurements; r = 0.752), and the direction of the bilirubin trend agreed between methods in 83% of assessed trajectories. A fixed offset correction reduced the bias during phototherapy from − 40.7 to + 5.8 µmol/L. Conclusion Although the correlation between Picterus JP results and TSB measurements decreased during phototherapy, the direction of change in Picterus JP results almost always agreed with the direction of change in associated TSB measurements. This suggests that Picterus JP may be useful for monitoring bilirubin trends and treatment response during phototherapy, while also providing non-invasive bilirubin assessment before treatment initiation.