|
Background: T-wave
alternans (TWA) has been proposed as an alternative to electrophysiologic
testing (EPS) for sudden cardiac death risk stratification. The purpose of
this study was to evaluate the extent to which these two different tests
yield concordant vs. complementary information.
Methods: We prospectively evaluated 307 pts (228 men, mean age 62?14 yrs) who were referred for EPS for risk stratification or due to a
history of syncope or documented sustained ventricular arrhythmias. All pts
underwent TWA testing during rapid atrial pacing (550 msec) with the
Cambridge Heart CH2000 system or HearTwave system at the time of EPS. TWA was
interpreted using standard criteria by two physicians blinded to the results
of EPS and pts with indeterminate tests were not included in this cohort. The
EPS protocol used programmed stimulation at two ventricular sites with up to
triple extrastimuli with and without isoproterenol or dobutamine. Positive
EPS was defined as the induction of sustained monomorphic ventricular
tachycardia with triple ventricular extrastimuli or
ventricular fibrillation with up to double ventricular extrastimuli. Concordance
was evaluated using Cohen's kappa (κ)
statistic, which measures the agreement between two different evaluations of
the same object. κ = 1
indicates perfect agreement, whereas κ = 0
indicates that agreement is no better than chance.
Results: For
the overall population, κ =
0.07 (p = NS). There was no significant agreement between TWA and EPS in any
subgroup when the population was stratified with respect to: 1) gender, 2) presence
of coronary artery disease, 3) history of myocardial infarction, 4) presence
of left ventricular aneurysm, 5) presence of valvular heart disease, 5) use
of beta-blockers at the time of testing, 6) presence of left
bundle branch block, 7) history of sustained ventricular tachyarrhythmias, 8)
history of cardiac revascularization, and 9) history of congestive heart
failure (all κ
< 0.13, all p = NS).
Conclusion: There is
no significant concurrence between the results of TWA testing and EPS in
patients being evaluated for ventricular arrhythmias. Thus, these tests yield
complementary rather than concordant information.
|