Risk assessment of R-on-T event based on modeled QT-RR relationship.

Methods: We retrospectively analyzed the Spontaneous Ventricular Tachyarrhythmia Database and the HAWAI Registry, which include a total of 397 RR interval recordings from 116 implantable cardioverter defibrillator patients. For each RR interval time series, QT intervals were estimated from the weighted average of preceding RR intervals using Bazett, Fridericia, and linear formulas. The risk score (RS) of each cycle was calculated to quantify the probability of R-on-T event based on the timing of R-wave relative to the estimated T-end. We identified 52,440 ectopic beats (EBs) episodes, 280 nonsustained VT (NSVT) episodes, and 352 sustained VT/VF episodes. The RS of episode onset and the prematurity index (PMI) of the initiating beat were compared. Results: Using different QT-RR models, R-on-T events were respectively detected in 9% EB, 45% NSVT, 69% VT/VF (Bazett); in 6% EB, 41% NSVT, 65% VT/VF (Fridericia); and in 7% EB, 42% NSVT, 66% VT/VF (linear). No R-on-T event was found in normal beats. Consistent among three QT-RR models, the RS of episode onset rises sharply from EB to NSVT and to VT/VF episodes. In contrast, no trend in PMI is found. Conclusions: The risk of R-on-T can be estimated from RR intervals, based on modeled QT-RR relationship. An episode onset with higher RS has increased risk of developing into NSVT or VT/VF. (PACE 2011; 1-9).