Age- and probability-adjusted D-dimer cut-offs held up in prospective use

Design

Systematic review and meta-analysis of prospective diagnostic management studies

Population

6 studies, 12,194 patients with suspected PE

Primary outcome

3-month VTE or PE-related death after rule-out without imaging

Effect

Overall failure 0.12% (0.06–0.26%); adaptive window 0.50% (0.20–1.11%)

Adaptive D-dimer strategies raise the threshold for excluding pulmonary embolism above the fixed 500 ng/mL — by age (age × 10 in those over 50) or by low clinical probability. Earlier safety data came mostly from cohorts managed with the fixed cut-off. This meta-analysis pooled only prospective management studies where the adaptive threshold actually decided whether patients had imaging.

Six studies with 12,194 patients were included. When PE was excluded without CT pulmonary angiography, the three-month failure rate was 0.12% (95% CI 0.06 to 0.26%), with no heterogeneity. In the 'adaptive window' — D-dimer above 500 but below the adjusted threshold — it was 0.50% (0.20 to 1.11%). About 61% of patients avoided imaging under the algorithms.

A failure rate under 2% is the accepted standard for safe exclusion, and the upper limit here sits well below it. This supports using adaptive thresholds routinely in non-pregnant adults, reducing CT use and contrast exposure, especially in older patients whose D-dimer rises with age. Pregnant women and children were excluded.

• Use an age-adjusted D-dimer cut-off (age × 10 ng/mL over 50) in patients with low or intermediate probability
• Consider a higher threshold in patients with low clinical probability under a validated algorithm such as YEARS
• Check whether your laboratory reports D-dimer in FEU or DDU units before applying a cut-off
• Do not use these thresholds in high-probability patients, who need imaging directly
• Apply pregnancy-specific algorithms, not these, in pregnant women

Why it matters

It confirms that patients in the grey zone above 500 can safely avoid CT, which matters most in older adults.

Don't overread it

Pregnant women and children were excluded, and the studies used specific validated algorithms rather than ad hoc thresholds.

The statistics, in plain English

A failure rate of 0.50% means about one missed clot in 200 patients in the adaptive window over three months. The upper confidence limit of 1.11% is still below the 2% generally accepted for safe rule-out.

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.