Topic guide

Screening and diagnostics

Evidence checks about screening, diagnostic tests, AI scans, blood tests, biomarkers, early detection and clinical utility claims.

What to watch for

Common ways headlines can go too far

  • Technical accuracy presented as patient benefit
  • Earlier detection framed as improved survival without outcome data
  • AI or blood-test performance reported without clinical utility

Screening and diagnostic headlines often confuse earlier detection, technical accuracy or AI performance with improved outcomes for patients.

Common hype patterns

  • Sensitivity and specificity framed as proof of patient benefit.
  • Earlier detection presented as improved survival without outcome data.
  • AI or blood-test performance reported without real-world clinical utility.
  • Overdiagnosis, false positives and follow-up harms omitted.

Questions to ask before trusting the headline

  • Was the claim tested in people, or only in cells, animals or models?
  • Is the outcome something patients notice, or only a surrogate marker?
  • Does the headline distinguish association from causation?
  • Are absolute risks, comparators and uncertainty shown clearly?
  • Would the finding change real-world behaviour, or is it mainly hypothesis-generating?

How deHype.io reads this topic

deHype.io separates the public claim from the underlying evidence. The aim is not to dismiss early science, but to show whether the headline has kept the finding in proportion.

Seen a headline in this topic? Paste the link and get a deHype evidence check.
Check a diagnostic headline