Application of probabilistic analysis for precisely correcting the QT interval for heart rate in telemetered common marmosets

Introduction: QT intervals are strongly influenced by preceeding heart rate history and are also characterized by rate-independent variability, leading to difficulty in precise rate-correction of the raw QT interval. The present study elucidates a novel analytical method that effectively addresses this problematic phenomenon in telemetered common marmosets. Methods: ECGs were collected from telemetered common marmosets (male and female) and analyzed by computerized algorithms. Descriptive statistics were calculated from the mean of QT intervals for 5-ms increments of RR. The QT interval was corrected for the RR interval by applying Bazett's, Fridericia's, and individual probabilistic QT rate-correction formulae. Results: The linear regression of log-transformed QT and RR intervals derived from a probabilistic approach yielded a well-correlated QT-RR fit. Assessed as the slope of the QTc-RR interval, application of individual probabilistic QT rate-corrections resulted in the most effective dissociation of the effects of rate from the raw QT interval, compared to generic rate-correction formulae. Using individual corrections, the QTc was stable while the interquartile range (IQR) of the QTc distribution was stable, spanning 5-10 ms for each subject over all physiological RR intervals. Heart rate variability distributions were centered about unity during both photoperiods and sinus arrhythmia was far less pronounced compared with measurements in dogs. Discussion: Probabilistic QT rate-correction eliminated the confounding effects of heart rate and provided a stable QTc baseline. These results indicate that application of this method of analysis in telemetered common marmosets results in a high degree of sensitivity for the consistent detection of small (5-10 ms) changes in the QTc interval.