Imaging/original research
Incidence and Predictors of Repeated Computed Tomographic Pulmonary Angiography in Emergency Department Patients

https://doi.org/10.1016/j.annemergmed.2008.08.015Get rights and content

Study objective

Use of contrast-enhanced computed tomography (CT) of the pulmonary arteries to evaluate for pulmonary embolism has increased, raising concern about radiation and contrast toxicity. We sought to measure the frequency of repeat CT pulmonary angiography in emergency department (ED) patients.

Methods

This was a prospective, longitudinal follow-up of ED patients who underwent first-time CT pulmonary angiography as part of a research protocol for diagnosis of pulmonary embolism in 2001 to 2002. Two authors (DMB and MCK) searched electronic medical record databases to measure the frequency of repeated CT scans performed within 5 years. Primary outcome was greater than or equal to 1 repeated CT pulmonary angiography examination. Radiologist-written interpretations of CT pulmonary angiography were categorized by 2 observers (DMB and JAK). Cox regression was used to estimate hazard ratios for 24 clinical variables.

Results

A cohort of 675 ED patients was observed for a median of 1,989 days: 226 of 675 (33%) had at least 1 additional CT pulmonary angiography scan, and 60 died with no repeated CT pulmonary angiography, leading to a mortality-adjusted frequency of repeated CT pulmonary angiography scanning of 226 of 615, or 37%. Seventy-three percent of the cohort had 1 or more subsequent CT scans of any body part, and 31 patients (5%) had 5 or more repeated CT pulmonary angiography scans. The pulmonary embolism (positive) prevalence was 57 of 675 (8.4%; 95% confidence interval [CI] 6.5% to 10.8%) on the first CT pulmonary angiography versus 8 of 226 (3.5%; 95% CI 1.5% to 6.9%) on the second CT pulmonary angiography scan. Hazard ratios indicated that respiratory rate, active malignancy, previous coronary artery disease, and previous or new diagnosis of venous thromboembolism were positively associated with repeated CT pulmonary angiography scanning.

Conclusion

At least one third of ED patients who undergo CT pulmonary angiography scanning will have a second CT pulmonary angiography result that will be negative for pulmonary embolism. New methods are needed to exclude pulmonary embolism recurrence without use of ionizing radiation.

Introduction

Pulmonary embolism manifests a wide variety of clinical presentations, ranging from classic (eg, sudden onset of dyspnea in a patient with risk factors and abnormal vital signs), to atypical (eg, altered mental status with normal vital signs).1 If undiagnosed, pulmonary embolism can kill rapidly.2 As a result, emergency physicians have a low threshold for ordering diagnostic testing for pulmonary embolism.3 Most emergency physicians in the United States rely on contrast-enhanced computed tomographic (CT) pulmonary angiography as a definitive diagnostic test to diagnose and exclude pulmonary embolism.3 Clinical practice guidelines generally suggest that emergency clinicians use pretest probability before testing for pulmonary embolism.4, 5 Clinical features that increase risk of pulmonary embolism in the emergency department (ED) setting include older age, unexplained dyspnea, increased pulse rate, low pulse oximetry reading, recent surgery, immobility, history of malignancy, and previous venous thromboembolism.6, 7, 8 These risk factors recur as fixed or intermittent features of the clinical picture each time the patient presents to the ED, leading to recurrent CT pulmonary angiography scanning. To address this issue, we report longitudinal follow-up data from a cohort of patients who underwent first-time CT pulmonary angiography in 2001 to 2002.

Section snippets

Theoretical Model of the Problem

Radiation and contrast material increase risk of malignancy and kidney injury.9, 10 These toxicities, together with the economic cost, provide motivation to avoid repeated CT pulmonary angiography, especially if the repeated CT pulmonary angiography images seldom show emergency radiologic findings. Accordingly, the first objectives of this study were to measure the frequency of repeated CT pulmonary angiography scans during 5 years and categorize the clinical significance of the radiologic

Results

From January 1, 2001, until September 30, 2002, we enrolled 695 unique patients who underwent what we believe was their first CT pulmonary angiography ordered from within our hospital system. (Hereafter, the day of enrollment will be referred to as the index visit.) Follow-up in 2007 was incomplete in 20 of 695 (3%) patients for the following reasons: unexplained absence of patient identification in any of 3 databases (n=12), no evidence of the written order or report of the first CT pulmonary

Limitations

Our estimated frequency of repeated CT pulmonary angiography scanning is based on data from the Carolinas Healthcare System, and this limitation probably caused underestimation of the primary outcome. However, because all patients in this study were prescreened by a clinical protocol that included pretest probability assessment and D-dimer and alveolar dead-space testing, if our population is compared with that of an ED that uses no such protocol, it might be found that our patients were

Discussion

Use of CT imaging in the ED has increased sharply in the past decade. Approximately 10% of all ED patients undergo some type of CT scan, and about half of these include intravenous contrast.17 A recent multicenter study of 12 US EDs and 1 New Zealand ED found that approximately 1.5% of all ED patients underwent CT pulmonary angiography to evaluate for suspected pulmonary embolism.3 Each CT pulmonary angiography scan poses inherent risk, including probable increased risk of cancer and contrast

References (21)

There are more references available in the full text version of this article.

Cited by (74)

  • Multi-Energy CT Applications: Problem-Solving in Emergency Radiology

    2023, Radiologic Clinics of North America
    Citation Excerpt :

    Image quality is a frequent challenge in ER CTA studies, with many causes including motion, patient positioning, and poor contrast enhancement due to bolus timing or IV infiltration. This is especially common with CT pulmonary angiography (CTPA), where adequate enhancement of the pulmonary arterial tree is essential to make an accurate diagnosis, and repeat studies are not uncommonly performed.40 Low-keV VMI, by accentuating iodine, will improve vascular enhancement in CTA studies and thereby improve image quality.

  • Pulmonary embolism rule-out criteria (PERC) rule in European patients with low implicit clinical probability (PERCEPIC): a multicentre, prospective, observational study

    2017, The Lancet Haematology
    Citation Excerpt :

    In overcrowded emergency departments, this decrease in the length of stay is beneficial not only for patients with low clinical probability and negative PERC but also for all emergency department patients.27 Moreover, PERC rule implementation could avoid potential adverse effects (eg, radiation exposure and iodinated contrast injection), cost-effectiveness imbalance, and incidental findings, which not only cause anxiety among patients but also prompt a series of follow-up imaging studies.28 Of note, the diagnostic strategy applied in our study was based, in case of low or moderate clinical probability, on a D-dimer test with age-adjusted cutoff (age × 10 in patients aged 50 years or older) but without any implication for patients with negative PERC (by definition, younger than 50 years) and we did not observe any false negatives of the age-adjusted cutoff during follow-up.29,30

View all citing articles on Scopus

Provide feedback on this article at the journal's Web site, www.annemergmed.com.

Supervising editor: William R. Mower, MD, PhD

Author contributions: JAK conceived the work, obtained funding, supervised the project, performed database queries, contributed to the primary analyses, and wrote the article. DMC performed the statistical and graphic analyses and edited the article. DMB, MCK, and MS performed database queries, data entry, and assisted with article production. JAK takes responsibility for the paper as a whole.

Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article that might create any potential conflict of interest. See the Manuscript Submission Agreement in this issue for examples of specific conflicts covered by this statement. JAK is inventor on US patent numbers 6,575,918; 6,881,193; 7,104,964. JAK owns stock in a company involved in developing a medical device referenced herein.

Publication date: Available online October 5, 2008.

View full text