Non-ECG-gated CT pulmonary angiography and the prediction of right ventricular dysfunction in patients suspected of pulmonary embolism

Research output: Contribution to journalJournal articleResearchpeer-review

PURPOSE: Right ventricular dysfunction (RVD) is an important prognostic factor of 30-day mortality in patients with acute pulmonary embolism (PE). The aim of our study was to evaluate whether non-electrocardiogram (ECG)-gated cardiovascular parameters attained during computed tomography pulmonary angiography (CTPA) could predict RVD in patients suspected of PE using ECG-gated cardiac CT angiography as reference.

METHODS: Consecutive patients suspected of PE were referred to a ventilation/perfusion single-photon emission tomography (V/Q-SPECT) as first-line imaging procedure. Patients had a V/Q-SPECT/CT, a CTPA and an ECG-gated cardiac CT angiography performed the same day.

RESULTS: A total of 71 patients were available for analysis. Seventeen patients (24%) had RVD. The non-ECG-gated dimensions of left and right ventricle and the major vessels were correlated with ECG-gated cardiac dimensions. The size of the pulmonary trunk could identify patients with RVD: AUC (0·67, 95% confidence intervals (CIs) 0·52-0·82) as seen in the ROC curve (P<0·05). With a cut-off value of the pulmonary trunk of 29 mm, the sensitivity and specificity were 70·6% and 55·5%, respectively. The positive predictive and negative predictive values for detection of RVD were 59·1% and 85·7%, respectively.

CONCLUSION: In the present study, we demonstrated correlation between ECG-gated cardiac dimensions and non-ECG-gated cardiovascular parameters, however with only moderate diagnostic accuracies. We demonstrated that the dimension of the pulmonary trunk might be of value in detection of patients with RVD. We suggest further studies on the potential value of non-ECG-gated cardiac dimensions in patients suspected of PE.

Original languageEnglish
JournalClinical Physiology and Functional Imaging
Volume37
Issue number6
Pages (from-to)575–581
Number of pages7
ISSN1475-0961
DOIs
Publication statusPublished - Nov 2017

ID: 164753829