Home  /  Insights  /  Regenerative science & regulation
Insights

Sample size and control groups

Two concepts that explain most of the gap between what small studies find and what larger ones later confirm.

All insights

Why small studies mislead

Small samples produce unstable estimates. A trial of fifteen people can show a striking result through chance alone, and small studies that happen to produce large effects are more likely to be published than small studies that find nothing.

The consequence is a well-documented pattern: early small studies report large effects, later larger studies report smaller ones. This is not evidence of fraud. It is what happens when the visible literature is filtered.

What controls control for

A control group accounts for everything other than the treatment — natural history, regression to the mean, placebo response, the effect of attention and expectation, and any co-interventions.

Without one, improvement cannot be attributed to the treatment. This is the single most common weakness in regenerative medicine case series, and it is usually acknowledged in the papers themselves.

Blinding

Blinding prevents expectation from influencing reported outcomes. Patient-reported measures such as pain scores are especially susceptible, and those are the primary endpoints in most of this literature.

Sham-controlled procedural trials are difficult and sometimes ethically contested, but where they exist they are the most informative studies available.

This article covers general science and published research. Whether any approach is appropriate for you is a clinical question, answered by a licensed provider through a good-faith exam.

Talk with a provider

Educational content. This page provides general information in line with the disclosures below; it is not medical advice or a claim of results. Eligibility for any service is determined by a licensed provider through a good-faith exam.