ClinBench: 1,487 Primary Outcome Comparisons for Clinical Trial Prediction

ClinBench is a curated benchmark dataset for predicting the success or failure of clinical trial primary outcomes. It contains 1,487 primary outcome comparisons from 780 unique trials registered on ClinicalTrials.gov. The dataset was authored by Vladimir Aladinskiy and last updated on May 6, 2026.

Use Cases

• Benchmarking machine learning models for predicting trial success based on pre-results information.
• Analyzing factors correlated with statistical significance (p < 0.05) of primary endpoints.
• Training models to assess the likelihood of a clinical trial's primary outcome.

Strengths

• Contains 1,487 primary outcome comparisons from 780 unique clinical trials.
• Specifically designed as a benchmark for evaluating predictive models.
• Data is sourced from the authoritative ClinicalTrials.gov registry.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

ClinicalTrials.gov

Collection Method

Curated from registered clinical trials.

Time Range

null

Freshness

Last updated 2026-05-06 21:10:06; freshness should be verified.

Geography

null