Special Feature

The Use of Chest CT for the Diagnosis of PE

By Esther Chen, MD and Stephanie Abbuhl, MD, FACEP Dr Chen is Assistant Professor, University of Pennsylvania, Philadelphia. Dr. Abbuhl is Vice Chair, Department of Emergency Medicine, The Hospital of the University of Pennsylvania; Associate Professor of Emergency Medicine, University of Pennsylvania School of Medicine, Philadelphia. Dr. Chen and Dr. Abbuhl have reported no relationships with companies having ties to the field of study covered by this CME program.

Several recent studies have significantly expanded the evidence for the central role of computed tomography (CT) in the evaluation of the patient with suspected pulmonary embolism (PE). We review these studies below and summarize the current understanding of the advantages and disadvantages of CT.

CT Technique

The current 16-detector row CT scanner allows imaging of the entire thorax with 1-mm collimation within a single breath hold in fewer than 10 seconds,1 compared with a SDCT, which scans with 2-5 mm collimation in a 32-second breath hold or during shallow respirations.2,3 The reduced acquisition time of MDCT diminishes motion artifacts, particularly important in dyspneic patients. Unfortunately, even with the 10-second breath hold, severely dyspneic patients often have large respiratory movements that can cause respiratory motion artifact, contributing to a low, but significant number of nondiagnostic studies (about 9-10% in several recent trials).2-4

Moreover, MDCT scanners have significantly improved the detection of small peripheral emboli because they provide high spatial resolution data that can be reformatted for two- and three-dimensional visualization. MDCT allows visualization of clot in pulmonary vessels down to the sixth order of branching, termed subsegmental vessels1 with higher interobserver correlation than pulmonary angiography (PA).5 Consequently, PA is no longer considered the gold standard for subsegmental PE, or, in the view of some experts, for PE in general.6 While small, previously undetectable peripheral clots now may be detected, the clinical significance of isolated subsegmental PE is still unclear. One study showed that in a small group of patients with isolated subsegmental emboli, 25% had concurrent lower extremity deep venous thrombosis (DVT).4 In the absence of DVT, however, treating patients with isolated subsegmental emboli remains controversial.

Finally, compared with VQ, CT scanning takes less time to perform, may diagnose other cardiopulmonary diseases once PE has been excluded (25-76% of the time),2,7 is readily available at most hospitals, and can be obtained easily after hours. Unfortunately, there continue to be some disadvantages to CT, including radiation exposure and the hazards of giving intravenous contrast to patients with renal insufficiency. Furthermore, with the high volume of CT studies that currently are performed, there are issues with managing and storing the enormous amounts of computerized data.

Outcome Studies

Previous PE studies evaluated the reliability of CT by comparing its sensitivity and specificity to PA, but in the past few years, studies have switched to more practical and realistic designs assessing patient out comes. There is a growing body of literature that evaluates the clinical validity of CT by withholding anticoagulation in patients with a negative CT evaluation and subsequently determining their venous thromboembolic (VTE) rate at three-month follow-up. Each study, however, incorporates CT in a unique diagnostic algorithm, often using a strategy that includes some combination of pretest clinical probability assessment, d-dimer testing, and lower extremity ultrasound (US). Therefore, one must consider that the negative predictive values in these studies are attributed to the CT study as part of an entire diagnostic strategy, and not as used in isolation. The post-test probability of having a VTE event following a negative chest CT evaluation is directly related to the prevalence of PE in the population studied and, therefore, the selection of low, moderate, or high pretest probability patients will play an important role in determining the clinical validity of CT for PE.

One of the recent landmark studies by Musset and colleagues3 was a prospective multicenter trial that enrolled 1041 patients with suspected PE, risk-stratified to low, intermediate, or high clinical probability, who underwent SDCT and lower extremity US. Patients with a low to intermediate clinical probability and negative CT and US studies had a VTE rate of 1.8% (95% CI, 0.8-3.3%) at three-month follow-up. Alternatively, high risk patients with negative CT and US evaluations had a VTE rate of 5.3% (95% CI, 1.5 13.1%), confirmed by VQ, angiography, or both. Patients with isolated subsegmental emboli and negative US (n = 12) were considered to have a nondiagnostic CT study, of which three (25%) patients were subsequently diagnosed with PE by VQ or PA. Of interest, 16% of patients with a negative CT study had positive US findings; more than half of the DVTs were below the popliteal veins. The authors concluded that in the low to intermediate risk patients, both negative CT and US studies are necessary to reliably exclude PE; high-risk patients must undergo additional testing.

The reliability of SDCT was evaluated further in a study2 of 512 patients with suspected PE who underwent CT imaging. In this protocol, there were no pretest probabilities or d-dimers used, and patients with negative or inconclusive CT studies were evaluated with lower extremity US. Almost half (48.6%) of the study group had a normal CT study, of which only two patients had DVT detected by US on day 1, and no further DVTs were identified on days 4 and 7. The overall VTE rate at three-month follow-up was 0.8% (95% CI, 0.2-2.3%). Of note, if US had not been done and the two patients who were identified had returned with DVT on follow-up, the rate of VTE would have been 1.3% (95% CI, 0.4 3.1%). The authors concluded that CT imaging may be used reliably as the primary imaging study to exclude PE, and suggested that US has little additional value.

Prompted by the flurry of trials using CT, Quiroz and colleagues8 performed a systematic review comparing the results of studies using CT in various diagnostic strategies for PE. Fifteen studies, published between 1990 to 2004, met the study criteria, which included 3500 patients using predominantly SDCT (80%). The overall negative likelihood ratio of VTE after a negative CT study for PE was 0.07 (95% CI, 0.05-0.11) with a negative predictive value of 99.1%. The NLR of a VTE after a negative SDCT was 0.08 and after MDCT was 0.15. The authors concluded that anticoagulation might be withheld safely in patients with a negative CT study because of the low incidence of VTE at follow-up.

Finally, in the most recently published prospective trial evaluating the utility of MDCT, researchers combined clinical probability, d-dimer testing, and lower extremity US to evaluate patients with suspected PE.9 This was an international multicenter study of 756 patients who were risk-stratified into low, intermediate, and high clinical risk of PE using the Geneva score. In the low-intermediate probability group, 34% had a negative d-dimer and not a single VTE event was found at three-month follow-up. Of those low-intermediate risk patients with a positive d-dimer and negative CT and US, the VTE rate at follow-up was 1.7% (95% CI, 0.7-3.9) and only two patients with a negative CT scan were found to have DVT. In the 82 patients in the high pretest probability group, 95% had a positive CT scan and only one patient with a negative CT scan had a DVT. In all pretest probability groups, 0.9% (95% CI, 0.3-2.7) of patients had proximal DVT despite a negative CT scan. Because of the small numbers in the high clinical probability group, the authors suggested that US be eliminated in all patients, except those at high risk for PE.

New Questions

Lower Extremity Evaluation. Lower extremity CT venography increasingly is being used in conjunction with chest CT to detect DVT. Although variable by institution, a typical protocol utilizes only one contrast injection for CT pulmonary angiography, followed by scanning from the iliac crest to the popliteal fossa, after a 120-second delay from the chest evaluation. With approximately three minutes added to the scanning time10 and no additional contrast, it is easy to see why this study is so appealing.

The gonadal radiation exposure, however, is increased 500-2000 fold compared with chest CT alone, although this is still within the thresholds provided by the ICRP-60 guidelines.11 Although this risk is small, it is important to carefully weigh the risks of the extra radiation exposure with the benefits of the additional detection of concurrent DVT.

Since most studies have shown only a small number of patients who have DVT in the absence of PE, the additional evaluation may not be necessary in most
patients, particularly patients of childbearing age. US may be a better alternative in those patients with a nondiagnostic CT study, when only isolated subsegmental emboli are detected, or when there is high probability of PE with a negative CT study.

Radiation Considerations. As both CT angiography and CT venography are used increasingly, there is mounting concern about the risks of radiation exposure, particularly in children and younger adults.12-14 It has become clear with results from a longitudinal (> 50 years) study of atomic bomb survivors, that low-dose radiation (similar to the exposure incurred with a single abdominal CT study) is associated with a small, but statistically significant, risk of excessive cancer over a lifetime.14 Of particular concern is that some patients undergo multiple CT imaging, thereby accumulating the radiation dose over their lifetime.15 Some authors have suggested that because of the absorbed breast dose, younger women with a normal chest x-ray (high likelihood of diagnostic VQ) should be evaluated further with a VQ instead of CT.16

Fortunately, health care providers are more cognizant of this potential radiation exposure, such that clinicians avoid ordering unnecessary CT studies, radiologists and technologists make individual adjustments to scan parameters (based on indication and patient size), and the CT industry is encouraged to develop newer technology to reduce radiation dose. New studies have shown promising results in reducing radiation by using low-dose techniques,12 testicular gonadal shields,17 and CT scanners that have automatic exposure control for changing radiation dose to different regions of the body that require fewer photons for good imaging.15

Recommended Algorithm

In patients with a low-intermediate clinical suspicion for PE, a negative ELISA d-dimer assay effectively excludes the diagnosis. A negative CT evaluation in patients with a low-intermediate PE risk and an elevated d-dimer assay also excludes PE, with an acceptable miss rate similar to that of pulmonary angiography. However, a negative CT evaluation in high-risk patients may not reliably exclude PE, and requires lower extremity imaging to exclude concurrent DVT Further outcome studies in high-risk patients would help to clarify this strategy. Moreover, when a lower extremity evaluation is indicated in patients of childbearing age, US should be the test of choice to minimize unnecessary radiation. In older patients, however, the risk/benefit ratio of CT venography may support its use. Finally, patients with inconclusive CT studies or isloated subsegmental embolirequire further diagnostic imaging.


1. Schoepf UJ, et al. CT angiography for diagnosis of pulmonary embolism: State of the art. Radiology 2004; 230:329-337.

2. van Strijen MJ, et al. Single-detector helical computed tomography as the primary diagnostic test in suspected pulmonary embolism: A multicenter clinical management study of 510 patients. Ann Intern Med 2003; 138:307-314.

3. Musset D, et al. Diagnostic strategy for patients with suspected pulmonary embolism: A prospective multicentre outcome study. Lancet 2002;360:1914-1920.

4. Revel MP, et al. Diagnosing pulmonary embolism with four-detector row helical CT: Prospective evaluation of 216 outpatients and inpatients. Radiology 2005; 234: 265-273.

5. Stein PD, et al. Reassessment of pulmonary angiography for the diagnosis of pulmonary embolism: Relation of interpreter agreement to the order of the involved pulmonary arterial branch. Radiology 1999;210: 689-691.

6. Ravenel JG, et al. CT in the diagnosis of pulmonary embolism. AJR Am J Roentgenol 2005;184:1707; author reply 1707-1708.

7. Kavanagh EC, et al. Risk of pulmonary embolism after negative MDCT pulmonary angiography findings. AJR Am J Roentgenol 2004;182:499-504.

8. Quiroz R, et al. Clinical validity of a negative computed tomography scan in patients with suspected pulmonary embolism: A systematic review. JAMA 2005; 293:2012-2017.

9. Perrier A, et al. Multidetector-row computed tomography in suspected pulmonary embolism. N Engl J Med 2005;352:1760-1768.

10. Cham MD, et al. Deep venous thrombosis: Detection by using indirect CT venography. The Pulmonary Angiography-Indirect CT Venography Cooperative Group. Radiology 2000;216:744-751.

11. 1990 Recommendations of the International Commission on Radiological Protection. Ann ICRP 1991;21: 1-201.

12. Tack D, et al. Multi-detector row CT pulmonary angiography: Comparison of standard-dose and simulated low-dose techniques. Radiology 2005;236:318-325.

13. Golding SJ. Multi-slice computed tomography (MSCT): The dose challenge of the new revolution. Radiat Prot Dosimetry 2005;114:303-307.

14. Slovis TL. Children, computed tomography radiation dose, and the As Low As Reasonably Achievable (ALARA) concept. Pediatrics 2003;112:971-972.

15. Frush DP, et al. Computed tomography and radiation risks: What pediatric health care providers should know. Pediatrics 2003;112:951-957.

16. Ravenel JG, et al. CT angiography with multidetector-row CT for detection of acute pulmonary embolus. Semin Roentgenol 2005;40:11-19.

17. Hohl C, et al. Radiation dose reduction to the male gonads during MDCT: The effectiveness of a lead shield. AJR Am J Roentgenol 2005;184:128-130.