# Paid independent statistical review for preclinical papers that inform trials

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

## TL;DR

Clinical trials get expert statistical review; the laboratory studies that justify them usually do not. Paying statisticians to review these papers before they influence a trial would catch errors early.

## Summary

Statistical errors in preclinical papers (pseudo-replication, inappropriate tests, no correction for multiple comparisons) are common and are rarely caught by biologist reviewers. Journals could not afford it; a funder-supported service that provides paid statistical review for preclinical papers cited in trial protocols or investigational new drug packages, with a published statistical report, would target the small fraction of papers that matter most for translation.

## Fields

- Kind: Idea
- Last checked: 2026-09-08
- Hypothesis: Statistical review will identify errors changing the interpretation in at least 20% of translationally important preclinical papers, and trials informed by reviewed papers will have lower early futility rates.
- Rationale: Statistical review is standard in clinical journals and reduces error; its absence in preclinical publishing is an economic rather than a scientific choice.
- Proposed test: Review 200 preclinical papers cited in registered trial protocols; publish the error rate and types; follow the trials.
- Maturity: speculative
- Actor: philanthropy

## Sources

- Bottleneck evidence (Preclinical results do not reproduce): Errington et al., Investigating the replicability of preclinical cancer biology (eLife 2021): https://doi.org/10.7554/eLife.71601

## Connected records

- ideas: [Independent replication of the key experiment before first-in-human academic trials](https://onco.cc/ideas/idea-tr2-replication-before-ind/), [Run automated statistics and image checks on every cancer manuscript before review](https://onco.cc/ideas/idea-tr2-automated-stats-check/)
- bottlenecks: [Preclinical results do not reproduce](https://onco.cc/bottlenecks/b-reproducibility/), [The valley of death between lab and product](https://onco.cc/bottlenecks/b-translational-valley/)

---
JSON: https://onco.cc/api/v1/entities/idea-tr2-paid-statistical-review.json