Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy.

PURPOSE

The purpose of this investigation is to estimate the prevalence of acute pulmonary embolism (PE) in a general hospital, its frequency among patients who died, and the ability of physicians to diagnose PE antemortem.

METHODS

The prevalence of acute PE among 51,645 patients hospitalized over a 21-month period was assessed in 1 of the 6 clinical centers (Henry Ford Hospital) that participated in the collaborative study, prospective investigation of pulmonary embolism diagnosis (PIOPED). The diagnosis of PE was made by pulmonary angiography, or in those who did not undergo pulmonary angiography because they declined or were ineligible for randomization to angiography in PIOPED, the diagnosis was based on the ventilation/perfusion (V/Q) lung scan. Based on data in PIOPED, PE was considered to be present in 87% of patients with high probability V/Q scam interpretations, 30% with intermediate probability interpretations, 14% with low probability interpretations, and 4% with nearly normal V/Q scans.

RESULTS

The estimated prevalence of acute PE in hospitalized patients was 526 of 51,645 (1.0%; 95% confidence interval [CI], 0.9 to 1.1%). Based on extrapolated data from autopsy, PE was estimated to have caused or contributed to death in 122 of 51,645 (0.2%; 95% CI, 0.19 to 0.29%). Pulmonary embolism was observed at autopsy in 59 of 404 (14.6%; 95% CI, 11.3 to 18.4%). Among patients with PE at autopsy, the PE caused or contributed to death in 22 of 59 (37.3%; 95% CI, 25.0 to 50.9%) and PE was incidental in 37 of 59 (62.7%; 95% CI, 49.1 to 75.0%). Among patients at autopsy who died from PE, the diagnosis was unsuspected in 14 of 20 (70.0%; 95% CI, 45.7 to 88.1%). Most of these patients had advanced associated disease. In these patients, death from PE occurred within 2.5 h in 13 of 14 (92.9%; 95% CI, 66.1 to 99.8%).

CONCLUSION

Pulmonary embolism is common in a general hospital. The prevalence of PE at autopsy has not changed over 3 decades. The frequency of unsuspected PE in patients at autopsy has not diminished. Even among patients who die from PE, the PE is usually unsuspected. Such patients, however, typically have advanced disease. Among moribund patients, incidental PE is rarely diagnosed. Patients who suffer sudden unexplained catastrophic events in the hospital are a group in whom the diagnosis might be suspected more frequently if physicians maintain a high index of suspicion.