PIOPED II (2006)

The accuracy of CT pulmonary angiography (CTPA) for diagnosing pulmonary embolism (PE) was uncertain, particularly in a broad patient population. This study aimed to evaluate the diagnostic performance of CTPA, alone and in combination with clinical probability, for detecting PE.

This prospective, multicenter cohort study enrolled 1967 patients with suspected PE. Patients underwent clinical probability assessment, CTPA, and a reference standard of ventilation-perfusion (V/Q) scan, pulmonary angiography, or clinical follow-up for 3 months.

The sensitivity of CTPA for PE was 83% and specificity was 96%. When combined with a high clinical probability, a positive CTPA had a positive predictive value of 96%. With a low clinical probability, a negative CTPA had a negative predictive value of 96%. Overall, PE was confirmed in 23% of patients, and CTPA had an indeterminate rate of 10%.

CTPA is a useful diagnostic tool for PE, especially when interpreted in conjunction with clinical probability. A negative CTPA in a patient with low clinical probability effectively rules out PE.

The study's reference standard was complex and varied, and a significant proportion of patients had indeterminate CTPA results. The generalizability to all patient populations, especially those with comorbidities, may be limited.