Correlation of diagnostic imaging and subsequent autopsy findings in patients with pulmonary embolism

We determined the frequency of correct antemortem diagnosis in patients who underwent lung scintigraphy for suspected pulmonary embolism and for whom the pulmonary embolism was proven at autopsy.

We performed a longitudinal follow-up study of 1210 consecutive patients who underwent scintigraphic imaging for suspected pulmonary embolism. Patients for whom pulmonary embolism was proven at autopsy within 30 days of lung scintigraphy were studied. Diagnostic testing and risk factors were compared in patients with and without an antemortem diagnosis of pulmonary embolism.

The antemortem diagnosis was made in four of eight patients with autopsy-proven pulmonary embolism, yielding a sensitivity of 0.5 (95% confidence interval, 0.16-0.84). The diagnosis was made by high-probability lung scintigraphy in two patients, by pulmonary arteriography in one patient, and by lower extremity venous sonography in another patient. The imaging investigation for suspected pulmonary embolism included only lung scintigraphy in four patients in whom the antemortem diagnosis was incorrect: two with low-probability scintigraphy and two with intermediate-probability scintigraphy. Risk factors were similar for patients with and without pulmonary embolism.

In this cohort study of patients suspected of having pulmonary embolism, the antemortem diagnosis was not made in 50% of patients in whom pulmonary embolism was later proven at autopsy. Underuse of diagnostic testing may have been to blame. We therefore believe that further study is needed to help identify patients with suspected pulmonary embolism who may benefit from additional diagnostic testing after low- or intermediate-probability lung scintigraphy.