# Score every preclinical model by how often it predicted the clinical result

Source: https://onco.cc/ideas/idea-tr2-model-report-cards/  
OnCo record `idea-tr2-model-report-cards` (Idea). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

For each type of laboratory model, keep a public record of how often its predictions came true in patients, so that researchers know which models to trust for which question.

## Summary

Model predictivity is asserted, not measured. Linking preclinical efficacy claims (from publications and investigational new drug packages) to subsequent clinical outcomes would yield per-model, per-indication predictive values: for instance how often cell-line xenograft regression preceded objective responses in the same indication. Failures are essential to this calculation, which is why they must be recorded.

## Fields

- Kind: Idea
- Last checked: 2026-09-08
- Hypothesis: Report cards will show at least two-fold differences in positive predictive value between model classes within the same indication, and this information will change model choice in subsequent grant applications.
- Rationale: Systematic reviews in stroke and neuroscience showed that animal model results predicted clinical results poorly; oncology has never computed the equivalent at scale despite having the most trials.
- Proposed test: Link 300 drug-indication pairs with published preclinical data to trial outcomes; compute predictive values by model class; publish and update annually.
- Maturity: speculative
- Actor: data

## Sources

- Bottleneck evidence (Failures are hidden): Anderson et al., Compliance with results reporting at ClinicalTrials.gov (NEJM 2015): https://doi.org/10.1056/NEJMsa1409364

## Connected records

- collections: [Cancer Models (PDCM Finder) & HCMI](https://onco.cc/collections/cancer-models/)
- ideas: [A machine-readable taxonomy of why cancer drugs fail](https://onco.cc/ideas/idea-tr2-failure-taxonomy/), [Living systematic reviews of animal and organoid evidence before every new trial](https://onco.cc/ideas/idea-tr2-preclinical-living-reviews/)
- technologies: [Patient-derived organoids](https://onco.cc/technologies/organoids/), [Patient-derived xenografts](https://onco.cc/technologies/pdx-models/)
- bottlenecks: [Failures are hidden](https://onco.cc/bottlenecks/b-negative-results/), [Lab models that fail to predict what happens in patients](https://onco.cc/bottlenecks/b-preclinical-models/)
- key papers: [Compliance with results reporting at ClinicalTrials.gov](https://onco.cc/key-papers/paper-anderson-n-engl-j-med/)

---
JSON: https://onco.cc/api/v1/entities/idea-tr2-model-report-cards.json