Predicting Paroxysmal Atrial Fibrillation in Cerebrovascular Ischemia Using Tissue Doppler Imaging and Speckle Tracking Echocardiography

Research output: Contribution to journalJournal articleResearchpeer-review

  • Flemming Javier Olsen
  • Peter Godsk Jørgensen
  • Rasmus Møgelvang
  • Jan Skov Jensen
  • Thomas Fritz Hansen
  • Jan Bech
  • Biering-Sørensen, Tor

BACKGROUND: Often the underlying cause of cerebral ischemia (CI) cannot be found during a routine diagnostic investigation, but paroxysmal atrial fibrillation (PAF) could be the culprit.

AIM: The objective of the study is to investigate whether advanced echocardiography improves the diagnostic approach for PAF in CI.

METHODS: The study included 286 CI patients with an echocardiogram in sinus rhythm. Patients were divided by PAF occurrence (PAF: n = 86, non-PAF: n = 200). PAF was defined as 1 or more reported episodes of atrial fibrillation. Echocardiograms consisted of conventional measures, tissue Doppler imaging (TDI), and speckle tracking. TDI was performed to acquire myocardial peak velocities during systole/ventricular contraction (global s'), early diastole/ventricular filling (global e'), and late diastole/atrial contraction (global a'). Speckle tracking was performed for myocardial strain analysis, thereby retrieving global longitudinal strain and global strain rate (s, e, a) values.

RESULTS: Patients with PAF exhibited significantly impaired atrial contractile measures: global a' (-7.0 cm/second versus -5.7 cm/second, P < .001) and global strain rate a (.97 second(-1) versus .81 second(-1), P < .001). Both were univariable markers of PAF, and along with age remained the only independent significant determinants of PAF after multivariable logistic regression. Area under the curve (AUC) for age, global a', and global strain rate a significantly exceeded AUC for age alone (.79 versus .76, P = .032). Cutoff values with the highest sensitivity and specificity for these 3 parameters improved the diagnostic accuracy (sensitivity = 97%, specificity = 32%, negative predictive value = 95%, and positive predictive value = 38%).

CONCLUSIONS: Atrial contractile measures by advanced echocardiography are significant determinants of PAF in CI. However, there is no discriminatory power to make them clinically useful at the current moment.

Original languageEnglish
JournalJournal of Stroke & Cerebrovascular Diseases
Volume25
Issue number2
Pages (from-to)350-9
Number of pages10
ISSN1052-3057
DOIs
Publication statusPublished - Feb 2016

    Research areas

  • Journal Article, Research Support, Non-U.S. Gov't

ID: 164456224