Prevalence and clinical history of incidental, asymptomatic pulmonary embolism: a meta-analysis

Context: Recently, there has been an increasing number of reports of incidental pulmonary embolism (PE) in patients undergoing chest computer tomography (CT) for reasons other than the research of suspected PE. Natural history of incidental PE remains unclear.

Objectives: To estimate the prevalence of incidental PE, to assess potential factors associated with incidental PE, and to evaluated its clinical history.

Data sources: MEDLINE, EMBASE databases (up to January 2009).

Study selection: Studies were included if the prevalence of incidental PE was assessed using CT scanning.

Data extraction: The prevalence of incidental PE in these patients was documented. Separate data for inpatients and outpatients and according to the reason for CT scanning were collected. Weighted mean proportion of the prevalence of incidental PE was calculated. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated to test the association with potential predictors of incidental PE.

Results: Twelve studies for a total of more than 10 thousand patients were included. The weighted mean prevalence of incidental PE was 2.6% (95% CI 1.9, 3.4). Hospitalization at the time of CT scanning and the presence of cancer were associated with a significantly increased risk of incidental PE (OR 4.27 and OR 1.80 respectively).

Conclusions: The prevalence of incidental PE is clinically relevant. Future studies are necessary to properly evaluate the clinical history of these patients.