Multicenter Validation of an XGBoost Model for Predicting Intrapartum High Fever

Bo Liu's research dataset includes 747 parturients from three Chinese hospitals who received labor analgesia and developed intrapartum fever. It contains six predictors selected via LASSO regression, including BMI and blood cell ratios, used to train and validate an XGBoost model for predicting high fever (≥38.5 °C). The dataset was last updated on 2026-05-07.

Use Cases

• Validating a machine learning model for predicting intrapartum high fever based on clinical variables like BMI and blood cell ratios.
• Analyzing feature importance for maternal fever prediction using SHAP values derived from the XGBoost model.
• Conducting external validation of a clinical prediction model across multiple hospital sites.

Strengths

• Dataset includes 747 patient records from a multicenter study, providing a basis for external validation.
• Model performance metrics are reported, including an AUC of 0.716 in the test set and 0.674 in external validation.
• Six key predictive features, such as BMI and monocyte/lymphocyte ratio, were identified and retained.

Limitations

• Row count for the underlying data tables is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is very small at 21.0 KB, indicating limited raw data scope.

Provenance

Source

figshare, author Bo Liu.

Collection Method

Multicenter retrospective study extracting candidate variables from electronic health records (EHR).

Freshness

Last updated 2026-05-07 05:24:41; freshness should be verified.

Geography

Three hospitals in China.