Data Sheet 1_Development of an online prediction tool for immunotherapy-related adverse ev

A dataset of 550 patients with advanced non-small cell lung cancer (NSCLC) treated with immune checkpoint inhibitors (ICIs) was used to develop a machine learning model for predicting immune-related adverse events (irAEs). The data includes six clinical features such as neutrophil count, lymphocyte count, and ECOG performance status. The dataset was created by Ling-Chun Cao and published on figshare in May 2026.

Use Cases

• Predicting the risk of immune-related adverse events based on clinical features like neutrophil and lymphocyte counts.
• Training interpretable machine learning models for clinical decision support using logistic regression.
• Validating predictive models for immunotherapy safety using a temporally distinct patient cohort.
• Developing web-based clinical tools for early identification of high-risk patients during hospitalization.

Strengths

• Dataset includes 550 patient records with a clear split into development (n=420) and temporal validation (n=130) cohorts.
• The optimal logistic regression model achieved an AUC of 0.855 in the test set and 0.801 in the validation set.
• Six essential clinical features are identified, including neutrophil count, lymphocyte count, and ECOG performance status.

Limitations

• Row count and column-level documentation are unknown, requiring manual inspection after download.
• The dataset is small in scale at 501.3 KB, which may limit the complexity of models it can support.
• Geographic and institutional provenance of the patient data is not specified, which may indicate source bias.

Provenance

Source

Ling-Chun Cao

Collection Method

Clinical data collected from patients treated with immune checkpoint inhibitors.

Time Range

Patient treatment dates from January 2019 to November 2024.

Freshness

Last updated 2026-05-01 04:19:03; freshness should be verified.