Breast Cancer-Related Lymphedema Risk Prediction Model with Internal Validation

A clinical prediction model developed from a retrospective cohort of 234 breast cancer surgery patients, of whom 27 developed lymphedema. The model incorporates four surgical and pathological variables to predict lymphedema risk and was internally validated using bootstrap methods. The dataset, authored by Yufei Fan and last updated in June 2026, is shared under a CC-BY-4.0 license.

Use Cases

• Validating lymphedema risk prediction models based on the four identified clinical variables.
• Benchmarking new predictive algorithms against the reported AUC of 0.964.
• Studying the relationship between surgical factors and lymphedema outcomes based on the cohort data.

Strengths

• Model performance is quantified with specific metrics, including an optimism-corrected AUC of 0.954.
• The dataset is derived from a clearly defined cohort of 234 patients with 27 outcome events.
• The model was developed using Firth penalized logistic regression to address sparse-data bias.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The model has only undergone internal validation, with external validation pending as noted in the description.

Provenance

Source

figshare

Collection Method

Retrospective cohort study of patients undergoing breast cancer surgery.

Freshness

Last updated 2026-06-03 04:47:20; freshness should be verified.