Pulmonary embolism (PE) is a potentially fatal disease if left untreated. The prevalence of PE increases markedly with age, and its diagnosis in elderly people is difficult since many cardiopulmonary conditions may mimic the clinical presentation of PE and age may unfavourably influence the characteristics of diagnostic tests for PE. The modern approach to PE is based on sequential diagnostic strategies based on clinical probability, D-dimer measurement, lower limb compression ultrasonography, ventilation/perfusion lung scans and helical computed tomography. Pulmonary angiography is rarely necessary as the noninvasive diagnostic work-up is usually conclusive.

Age reduces the clinical usefulness of D-dimer measurement and the ventilation/perfusion lung scan. D-dimer measurement permits PE to be ruled out in only 5% of patients aged >80 yrs, compared with 60% aged <40 yrs. Similarly, the rate of inconclusive ventilation/perfusion lung scans is almost twice as high (58%) in patients aged >70 yrs compared with patients aged <40 yrs (32%). In contrast, the diagnostic accuracy of clinical probability assessment, either empirical or as determined by prediction rules, is unchanged by ageing, and the diagnostic characteristics of helical computed tomography do not to appear to be influenced by advancing age.

Compression ultrasonography was mandatory when using single-slice computed tomography. However, recent studies suggest that compression ultrasonography is no longer mandatory when using multi-slice computed tomography (MSCT). In outpatients, a strategy based on assessment of clinical probability, D-dimer measurement and MSCT has been shown to be cost-effective, even in elderly patients. In in-patients D-dimer measurement is probably useless. Compression ultrasonography remains useful in patients with contraindications to MSCT, as an attempt to avoid further imaging.