Trial Outcome Prediction (TOP)

Benchmarks for predicting Phase 1-3 trial outcomes from structured + text features.

Composite
76.5
Experimental validation
Clinical
Stages
Phase IIIClinical Development
Modalities
cross-modality
Task types
trial-outcome-prediction
Size
trials: 17,000
phases: 3
License
MIT
First release
2022
Last updated
2024
Official site
→ project page
Leaderboard
→ leaderboard
Dataset
→ dataset
Code / GitHub
→ repository
HuggingFace
→ HF
Paper
Artificial intelligence for clinical trial design · Harrer S, Shah P, Antony B, Hu J · 2019 · paper · doi:10.1016/j.tips.2019.05.005 · 520 citations
Flags
none
Experts
Tianfan Fu
Groups
Fu Lab (RPI), Sun Lab (Georgia Tech)
Hosted by
TrialBench / HINT / TOP
Related benchmarks
HINT / TrialBench

Rubric (7-criterion)

rigor
4
coverage
3
maintenance
4
adoption
3
quality
4
accessibility
5
industry_relevance
4

Notes

Often reported alongside HINT.

← Back to all benchmarks

Compare:
Open comparison →