A comprehensive formula for computing corrected QT intervals in patients with wide QRS

Background: There is increasing interest in developing appropriate methods for correction of the QT intervals in patients with prolonged depolarization of the ventricles. From an experimental model of controlled heart rate and QRS widening, we aimed to develop a new formula with better performance.

Methods: After exclusions, incremental pacing was performed (AAI mode; 50-120 bpm) in 17 patients admitted for self-expanding aortic valve implantation in two different phases: before and immediately after the release of the prosthesis. Mixed linear models were built to predict variation in QRS and JT intervals, and subsequently calculate the corrected QT (QT_QRS-C). Internal and external validation on a new cohort of 48 patients was performed.

Results: Significant QRS widening (32.5 ms; CI95%:31.5-33.6) and QT prolongation (35.4 ms; CI95%:33.4-37.4) was observed after the release of the prosthesis. The best fit formula was computed as QT_QRS-C = QT_measured - (1.058*QRS_measured - 108.397) - (0.0496*JT_measured - 11.038). Internal validation provided a good correlation between measured and predicted QT (Pearson's coefficient:0.76; CI95%:0.70-0.80). In an external cohort (n = 48), the QRS widening was 28.4 ms(CI95%:21.8-35.1) and the JT widening was 16.58 ms(CI95%:8.73-24.4). Compared to other formulas, our proposed formula tends to display better performance as a result of a combination of appropriate correlation (Pearson's:0.75), being the closer to identity line (slope 0.83) and minimizing the relative standard error of the estimates (RSE 0.11).

Conclusions: We developed a formula with better performance for QT correction in patients with wide QRS.