A validated score to triage the lytic bone lesion

Design

Retrospective external validation, four reviewers (Diagnostic Level III)

Population

423 long-bone lytic lesions on extremity radiographs

Primary outcome

Performance for malignancy risk stratification

Effect

Odds ratio 9.7 per point (95% CI 6.2-14.9); sensitivity 97.1%, specificity 67.4%, NPV 92.4%

An incidental lucent bone lesion on an extremity radiograph forces a judgement: reassure, follow up, or refer for possible malignancy. Bone-RADS, an American College of Radiology scoring system using five radiographic features plus malignancy history, aims to standardise that call, and this study externally validated it.

Across 423 long-bone lytic lesions scored by four reviewers, a higher Bone-RADS score strongly tracked malignancy, with an odds ratio of 9.7 (95% CI 6.2 to 14.9) per one-point increase. As a triage test it was highly sensitive at 97.1% with a negative predictive value of 92.4%, while specificity was more modest at 67.4%. Agreement on the full four-tier score was only moderate, but the clinically important benign-versus-malignant split was more reproducible.

For practice this supports Bone-RADS as a structured triage aid: a low score is reassuring, while a higher score flags a lesion that warrants specialist referral and further imaging. Its high sensitivity makes it good at not missing malignancy; the modest specificity means some benign lesions will still be referred.

• External validation of Bone-RADS on 423 long-bone lytic lesions scored by four reviewers.
• Each one-point score rise raised malignancy odds nearly tenfold (odds ratio 9.7, 95% CI 6.2 to 14.9).
• Sensitivity was 97.1% and negative predictive value 92.4%; specificity was 67.4%.
• Benign-versus-malignant classification was more reproducible than the full four-tier score.
• Use a low score to reassure and a higher score to trigger specialist referral and further imaging.

Why it matters

It gives a non-specialist a structured, validated way to decide which lucent lesion is safe to watch and which needs urgent referral.

Don't overread it

This was a retrospective validation and the score is a triage aid, not a diagnosis; a worrying lesion still needs specialist assessment and, where indicated, biopsy regardless of score.

The statistics, in plain English

High sensitivity (97%) with a 92% negative predictive value means a low score rarely misses cancer; the modest specificity (67%) means a positive result is less certain, so some benign lesions will be referred unnecessarily.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.