CT-ADE: Adverse Drug Event Prediction from Clinical Trial Results

CT-ADE is an evaluation benchmark for predicting adverse drug events (ADEs) from clinical trial results, released by ds4dh and updated in March 2025. It provides a standardized framework for assessing models that identify adverse drug reactions (ADRs) within clinical study outcomes.

Use Cases

• Predicting adverse drug events (ADEs) from clinical trial result text
• Benchmarking NLP models on clinical trial outcome extraction
• Evaluating the correlation between trial results and reported adverse drug reactions (ADRs)

Strengths

• MIT licensed for open redistribution
• Specifically designed as an evaluation benchmark for ADE prediction
• Updated in 2025

Limitations

• Unknown sample size and record count
• Lack of detailed column documentation in the source metadata
• Potential for class imbalance between common and rare adverse events

Provenance

Source

ds4dh (GitHub)

Collection Method

Curated from clinical trial results

Freshness

Last updated March 27, 2025.