Predictive Models for Neonatal Follow-Up Serum Bilirubin: Model Development and Validation

Bilirubin Measurement and Phototherapy Use After the AAP 2022 Newborn Hyperbilirubinemia Guideline

BACKGROUND AND OBJECTIVES

Guidelines for the management of neonatal hyperbilirubinemia have helped to reduce rates of significant hyperbilirubinemia. However, recent evidence suggesting overtreatment and potential harms of phototherapy have informed the American Academy of Pediatrics clinical practice guideline revision and the accompanying increase in phototherapy thresholds. These changes are predicted to safely reduce overuse; however, to date, the exact effect of these guidelines has not been established.

METHODS

We conducted a retrospective study of newborns born at ≥35 weeks' gestation across a network of 8 hospitals between January 2022 and June 2023. Outcomes included rates of phototherapy and total serum bilirubin (TSB) measurements before and after guideline publication, as well as clinical outcomes, including length of stay, readmissions, and duration of phototherapy.

RESULTS

In our cohort of >22 000 newborns, we observed a 47% decrease in phototherapy utilization, from 3.9% to 2.1% (P < .001). TSB measurements were reduced by 23%, from 712 to 551 measurements per 1000 newborns (P < .001), without an increase in outpatient TSB measurements. We did not observe an increase in readmissions receiving phototherapy, and length of stay increased by only 1 hour (P < .001).

CONCLUSIONS

Our study reveals that the publication of the updated American Academy of Pediatrics 2022 hyperbilirubinemia guidelines has likely yielded a significant reduction in phototherapy use and serum bilirubin measurement. Dedicated quality improvement initiatives may help determine which implementation strategies are most effective. Further population-level studies are needed to confirm safety with ongoing guideline uptake.

Coagulation profile of neonates with hyperbilirubinaemia in full-term newborns

OBJECTIVE

To determine the relationship between coagulation system markers and hyperbilirubinaemia in full-term neonates.

METHODS

This retrospective case-control study enrolled full-term neonates with hyperbilirubinaemia and healthy control full-term neonates. Prothrombin time (PT), thrombin time (TT), fibrinogen (Fbg), activated partial thromboplastin time (APTT) and international normalized ratio (INR) were recorded. The correlation between INR and total bilirubin and between INR and indirect bilirubin was analysed by linear regression analysis. Receiver operating characteristic curve analysis was used to assess the efficacy of INR to identify neonates with hyperbilirubinaemia.

RESULTS

This study enrolled 40 full-term neonates with hyperbilirubinaemia and 30 healthy controls. PT, INR and APTT were significantly higher in the neonates with hyperbilirubinaemia compared with the healthy controls. There was a significant positive correlation between INR and the level of total bilirubin in neonates with hyperbilirubinaemia (R = 0.3327). There was also a significant positive correlation between INR and the level of indirect bilirubin in neonates with hyperbilirubinaemia (R = 0.3406). INR in neonates with hyperbilirubinaemia significantly achieved an area under the curve of 0.800 (95% confidence interval, 0.6288, 0.9712; cut-off value, 1.060; specificity, 71.43%; sensitivity, 80.00%).

CONCLUSION

These findings suggest that INR is a novel biomarker for the diagnosis of neonatal hyperbilirubinaemia in full-term neonates.

Neonatal Jaundice Diagnosis Using a Smartphone Camera Based on Eye, Skin, and Fused Features with Transfer Learning

Neonatal jaundice is a common condition worldwide. Failure of timely diagnosis and treatment can lead to death or brain injury. Current diagnostic approaches include a painful and time-consuming invasive blood test and non-invasive tests using costly transcutaneous bilirubinometers. Since periodic monitoring is crucial, multiple efforts have been made to develop non-invasive diagnostic tools using a smartphone camera. However, existing works rely either on skin or eye images using statistical or traditional machine learning methods. In this paper, we adopt a deep transfer learning approach based on eye, skin, and fused images. We also trained well-known traditional machine learning models, including multi-layer perceptron (MLP), support vector machine (SVM), decision tree (DT), and random forest (RF), and compared their performance with that of the transfer learning model. We collected our dataset using a smartphone camera. Moreover, unlike most of the existing contributions, we report accuracy, precision, recall, f-score, and area under the curve (AUC) for all the experiments and analyzed their significance statistically. Our results indicate that the transfer learning model performed the best with skin images, while traditional models achieved the best performance with eyes and fused features. Further, we found that the transfer learning model with skin features performed comparably to the MLP model with eye features.