We evaluated whether the predictive value of Tpe for SCA improves with heart rate correction and sought to determine an optimal cutoff value for Tpec in the context of SCA risk.
Cases of SCA (n = 628; mean age 66.4 ± 14.5 years; n = 416, 66.2% men) from the Oregon Sudden Unexpected Death Study with an archived electrocardiogram available prior and unrelated to the SCA event were analyzed. Comparisons were made with control subjects (n = 819; mean age 66.7 ± 11.5 years; n = 559, 68.2% men). The Tpe interval was corrected for heart rate using Bazett (TpecBa) and Fridericia (TpecFd) formulas, and the predictive value of Tpec for SCA was evaluated using logistic regression models.
The area under the curve for Tpec predicting SCA improved with both correction formulas. TpecBa and TpecFd were shown to have an area under the curve of 0.695 and 0.672, respectively, as compared with a baseline of 0.601 with an uncorrected Tpe. A TpecBa value of >90 ms was predictive of SCA, independent of age, sex, comorbidities, QRS duration, corrected QT interval, and severely reduced left ventricular ejection fraction (≤35%; odds ratio 2.8; 95% confidence interval 1.92–4.17;
Correcting Tpe for heart rate, using either the Bazett or the Fridericia formula, improved the independent predictive value of this marker for the assessment of SCA risk. Prolongation of TpecBa beyond 90 ms was associated with a nearly 3-fold increased risk of SCA.