Clinical Data for Predicting COPD Exacerbation Risk in 2,102 Patients

2,102 patient records from a retrospective study of Chronic Obstructive Pulmonary Disease (COPD) patients admitted between 1 January 2019 and 31 December 2024. The dataset was used to develop and compare machine learning models for predicting the risk of acute exacerbations, with the XGBoost model achieving an AUC of 0.960 in training and 0.824 in testing. It was authored by Dapeng Kuang and shared under a CC-BY-4.0 license.

Use Cases

• Training machine learning models to predict COPD exacerbation risk based on clinical and laboratory variables.
• Identifying key risk factors for acute COPD exacerbations using SHAP analysis.
• Benchmarking predictive performance of models like XGBoost against other algorithms for clinical outcomes.

Strengths

• Contains data from 2,102 patient records, providing a substantial sample size.
• Model performance is documented with specific metrics, including an AUC of 0.960 in the training set.
• Key predictive features like creatinine, neutrophil percentage, and D-dimer are identified in the description.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small (1.2 KB), suggesting it may contain only summary statistics or a limited feature set.

Provenance

Source

figshare

Collection Method

Retrospective analysis of patient admissions.

Time Range

2019-01-01 to 2024-12-31

Freshness

Last updated 2026-05-14 05:43:31; freshness should be verified.