Sample Case Study #1

Case Study: Direct-to-Consumer Microbiome Testing

Question

Did the public-facing narrative match the paper?

What I found

The article captured the broad cautionary message of the study, but some consumer-facing advice and interpretive language were stronger than the paper itself directly supported.

Why it matters

That kind of drift changes how readers interpret reliability, risk, and what they think they should do next.

Signals

  • Evidence fidelity: Moderate

  • Claim traceability: Moderate

  • Communication risk: High

  • Source-study stability: Moderate to high

See the report

Sample Case Study #2

When a strong paper becomes a stronger story

I compared the Segata 2025 source paper against both its press release and a media article to see how the evidence changed as it moved into public-facing language.

What the paper did

The source paper developed a microbial ranking framework intended to distinguish species associated with healthier versus poorer health states across large metagenomic datasets. In this case, the paper itself did not present as a major measurement-instability problem. The main issue was not that the science collapsed under scrutiny. The issue was what happened next.

What I found

Both downstream communication layers moved beyond the paper.

The press release and the media article each pushed the narrative into stronger predictive, benchmark, and implementation language than the paper did. In plain English, the public-facing story became more confident and more actionable than the source evidence justified.

Why it matters

This was not a case of fake science. It was a case of translation drift.

That distinction is important. Many evidence problems do not start with a bad paper. They start when association-level findings are rewritten as if they were validated prediction tools, clinical benchmarks, or ready-to-use implementation frameworks.

Audit result

My review found:

  • substantial narrative drift in both the press release and the media article

  • weak claim traceability from public claims back to the paper

  • high communication risk, because the messaging encouraged a stronger interpretation than the study comfortably supported

  • relatively stable source-study signal, meaning the biggest problem was the translation layer, not the paper itself

Bottom line

The Segata 2025 case is a good example of why evidence translation audits matter. The source study was not the main problem. The problem was how quickly the downstream narrative became more certain than the evidence underneath it.

That is exactly the gap I help clients identify before a message goes public.

See sample report #2