More physicians are using spiral computed tomographic pulmonary angiography (CTPA) in the evaluation of patients with suspected pulmonary embolism (PE) because of its increasing accuracy and availability. However, CTPA’s inadequate sensitivity for excluding PE has prompted some to question the safety of CTPA as a definitive diagnostic test for PE. Moores and colleagues performed a systematic review of the literature and a meta-analysis to examine the rate of subsequent fatal and nonfatal PE in patients who did not receive anticoagulants because of a negative CTPA evaluation for PE.

Twenty-three of the 640 potentially relevant abstracts were included in the analysis. Inclusion criteria were: a consecutive sample of patients, a well-defined diagnostic strategy, withholding of anticoagulation in patients with negative CTPA tests, data on subsequent PEs, and a minimum three-month follow-up period. Eight studies were retrospective, and fifteen were prospective. All but one study included additional diagnostic tests in conjunction with CTPA, such as D-dimer measurements, lower extremity ultrasonography, and objective imaging. Computed tomography techniques also varied between studies.

Final analysis included 4,657 patients with negative CTPA results. At three-month follow-up, 1.4 percent suffered a PE—0.51 percent were fatal. Among patients in the prospective studies, the rates of nonfatal and fatal PE were 1.2 and 0.26 percent, respectively. In the nine studies with follow-up periods longer than three months, only one PE occurred after the third month. In a separate analysis of 327 patients whose CTPA results were inconclusive but did not receive empiric anticoagulation, the authors found a 16.2 percent rate of subsequent PE.

The authors conclude that withholding anticoagulation after a negative CTPA result is safe. Although a small number of patients developed nonfatal or fatal PE within three months, the results are similar to those found with conventional pulmonary angiography. The authors suggest that because 14 percent of patients who received anticoagulation despite negative CTPA results (and were excluded from the pooled analysis) had the potential for PE as confirmed by ultrasound, physicians should not use CTPA without concomitant ultrasonography.