A LASSO-Based Nomogram for Predicting Acute Bilirubin Encephalopathy in Newborns

Shenzhen Children's Hospital collected data from 287 newborns with severe hyperbilirubinemia between January 2015 and December 2022. The dataset was used to develop and validate a predictive nomogram for acute bilirubin encephalopathy (ABE) using LASSO regression and logistic analysis. Author Qifa Hu published the supporting file on figshare under a CC-BY-4.0 license.

Use Cases

• Developing a clinical risk prediction tool based on identified predictors like delivery method and total bilirubin.
• Validating logistic regression models for neonatal encephalopathy using a training and validation split.
• Benchmarking LASSO regression feature selection techniques for medical datasets.
• Conducting decision curve analysis (DCA) to evaluate the clinical utility of a predictive model.

Strengths

• Model validation was performed on distinct training (200 cases) and validation (87 cases) groups.
• The nomogram model achieved an area under the curve (AUC) of 0.822 on the validation set, indicating good predictive performance.
• The Hosmer-Lemeshow goodness-of-fit test results (P-values of 0.558 and 0.805) suggest good model calibration.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect geographic bias inherent to a single hospital in Shenzhen.

Provenance

Source

figshare

Collection Method

A retrospective analysis was conducted on patient records.

Time Range

January 2015 to December 2022

Freshness

Last updated 2026-05-15 05:40:57; freshness should be verified.

Geography

Shenzhen Children's Hospital, China