Nine studies (13,299 participants) were included: five assessed diagnosis of DVT and four assessed diagnosis of pulmonary embolism. The main limitations of the included studies were potential for differential verification bias (78%), not blinding test interpreters (78%) and not reporting intermediate results (69%).
Diagnosis of DVT (five studies, 6,560 participants): Prevalence of DVT ranged from 15% to 39%. There was no clear association with age. Sensitivity ranged from 96% to 100% (four studies). Specificity ranged from 29% to 70% (four studies) and decreased with increasing age. Safety ranged from 0% to 4% (five studies) and efficiency ranged from 21% to 40% (five studies).
Diagnosis of pulmonary embolism (four studies, 6,739 participants): Prevalence of pulmonary embolism ranged from 20% to 24%, and increased with increasing age. Sensitivity was 100% (one study) and specificity ranged from 50% to 22% (one study), decreasing with increasing age (one study). Safety was 1% (two studies). Efficiency ranged from 30% to 32% (two studies).