- Email: [email protected]
Manual and automatic measurements of moxifloxacin-induced QT prolongation
Available online at www.sciencedirect.com
Journal of Electrocardiology 42 (2009) 607 – 613 www.jecgonline.com
Poster Session 1
Conclusions: The results confirm the hypothesis that people with higher BMI, a growing section of the population, have lower ECG amplitudes. Therefore, the Sokolow-Lyon voltage criteria underestimate the presence of LVH for subjects with higher BMI. Our analysis suggests that computerized electrocardiography for the diagnosis of LVH should incorporate the BMI factor to improve sensitivity while keeping the specificity of LVH detection.
commercial algorithms to detect moxifloxacin-induced QTc prolongation should be investigated. Methods: A total of 124 healthy, young subjects (50 females) were randomized to moxifloxacin or placebo treatment in a thorough QT study. 12-Lead electrocardiograms (ECGs) were extracted from Holter in triplicates at 12 timepoints on the days 1 (baseline) and 7 (treatment). Manual measurement of RR, QT, and QTcF intervals was performed for each ECG by a commercial ECG core laboratory. The measurements were done in 3 consecutive beats in lead II, and the average value across the beats was reported. Automatic measurements were performed on a global beat using GE Healthcare's 12SL algorithm. For each ECG, the difference between the automatic and manual QT and QTcF measurements was evaluated. Time-matched change from baseline (Δ) was evaluated for all day 7 recordings and used to compute placeboadjusted change from baseline (ΔΔ) for the moxifloxacin group. Results: The correlation (r) between automatic and manual QT measurements across all recordings was 0.96 for QT and 0.88 for QTcF. The differences between automatic and manual measurements were normally distributed and showed generally longer intervals for the automatic method for both QT: 10.0 ± 7.5 milliseconds (mean ± SD) and QTcF: 10.6 ± 8.3 milliseconds, P b .001. For ΔQTcF-values, the difference between automatic and manual measurements was 1.5 ± 0.1 milliseconds (mean ± SE), P b .001. For placebo-adjusted QTcF change from baseline (ΔΔ), the difference between automatic and manual measurements was 0.7 ± 0.2 milliseconds (mean ± SE), P b .001. There was no difference (P = .40) in mean ΔΔQTcF at the time of maximum change (4 hours postdose) between manual, ΔΔQTcF = 9.4 milliseconds (90% confidence interval, 5.7-13.1 milliseconds) and automatic ΔΔQTcF = 9.5 milliseconds (90% confidence interval, 6.3-12.8 milliseconds). Conclusion: The study showed clear correlation between manual and automatic QT /QTc measurements with a bias toward longer automatic measurements for the individual ECG. This bias is markedly reduced when evaluating changes in QTc—there was no practical difference in ΔQTc and ΔΔQTc values between the 2 methods. Both the automatic and manual methods detected the moxifloxacin-induced QTc prolongation, and there was no significant difference in ΔΔQTc at the time of maximum change between the 2 methods.
doi:10.1016/j.jelectrocard.2009.08.007
doi:10.1016/j.jelectrocard.2009.08.008
Manual and automatic measurements of moxifloxacin-induced QT prolongation M.P. Andersen,a J. Matz, b C. Graff, a J.K. Kanters, c E. Toft, a J.J. Struijk, a ( aAalborg University, Aalborg, Denmark; bH. Lundbeck A/S, Copenhagen, Denmark; cUniversity of Copenhagen, Copenhagen, Denmark)
Philips Microvolt Electrocardiogram Amplitude Periodicity algorithm for detection and quantification of T-wave alternans Saeed Babaeizadeh, Eric D. Helfenbein, Jim M. Lindauer, Sophia H. Zhou, (Advanced Algorithm Research Center, Philips Healthcare, Thousand Oaks, CA, USA)
Background: Accurate measurement of QT intervals is of pivotal importance in cardiac safety trials. Thorough QT studies must be designed to detect an 8- to 12-millisecond ΔΔQTc prolongation in a positive control group treated with moxifloxacin. QT measurements are typically performed manually to maximize measurement accuracy. Automatic QT measurements are more cost-effective, but the ability of
Background: Microvolt Electrocardiogram (ECG) Amplitude Periodicity (MEAP) algorithm, introduced in our other presentation, is able to detect and quantify patterned beat-to-beat variability in multilead ECG recordings. This study examines the performance of this algorithm for detecting and quantifying T-wave alternans (TWA) as one form of patterned behavior in a part of an ECG beat.
Negative correlation relationship between left ventricular hypertrophy Sokolow-Lyon and body mass in 41 806 Swiss conscripts Roger Abächerli, a Richard Kobza, b Bernhard Niggli, b Lingchuan Zhou, a Johann-Jakob Schmid, a Franz Frey, c Paul Erne, b ( aSCHILLER AG, Biomed. Research and Signalprocessing, Baar, Switzerland; bDivision of Cardiology, Cantonal Hospital Luzern, Switzerland; cMedical Service, Swiss Army, Ittigen, Switzerland) Background: Conventional electrocardiographic diagnosis criteria for left ventricular hypertrophy (LVH), such as the Sokolow-Lyon voltage, have been limited by low sensitivity at acceptable levels of specificity. A number of studies have demonstrated that body mass index (BMI) is associated with decreased sensitivity of electrocardiogram (ECG) LVH classification in hypertensive patients. The objective of this study is to investigate the correlation relationship between Sokolow-Lyon voltage criteria and BMI in Swiss conscripts. Methods: A database of 41 806 young Swiss people who underwent compulsory conscription for the Swiss Army was compiled. Along with other medical data, an ECG was taken. Statistical analyses, such as linear regression and calculation of correlation coefficient, were carried out between the Sokolow-Lyon voltages and BMI. Results: The mean (SD) age in the studied population was 19.2 (1.1) years, with a median age of 19 years (range, 17-38 years). We found an overweight prevalence of 25.1%. The results showed that body habitus had significant association with Sokolow-Lyon voltages. A mean decrease of 13%, 5%, 19%, and 14% for the 4 studied indexes were found between reference range subjects (in kilograms per meter squared: 18.5 ≤ BMI b 25) and obese subjects (25 ≤ BMI).
0022-0736/$ – see front matter
Journal of Electrocardiology 42 (2009) 607 – 613 www.jecgonline.com
Poster Session 1
Conclusions: The results confirm the hypothesis that people with higher BMI, a growing section of the population, have lower ECG amplitudes. Therefore, the Sokolow-Lyon voltage criteria underestimate the presence of LVH for subjects with higher BMI. Our analysis suggests that computerized electrocardiography for the diagnosis of LVH should incorporate the BMI factor to improve sensitivity while keeping the specificity of LVH detection.
commercial algorithms to detect moxifloxacin-induced QTc prolongation should be investigated. Methods: A total of 124 healthy, young subjects (50 females) were randomized to moxifloxacin or placebo treatment in a thorough QT study. 12-Lead electrocardiograms (ECGs) were extracted from Holter in triplicates at 12 timepoints on the days 1 (baseline) and 7 (treatment). Manual measurement of RR, QT, and QTcF intervals was performed for each ECG by a commercial ECG core laboratory. The measurements were done in 3 consecutive beats in lead II, and the average value across the beats was reported. Automatic measurements were performed on a global beat using GE Healthcare's 12SL algorithm. For each ECG, the difference between the automatic and manual QT and QTcF measurements was evaluated. Time-matched change from baseline (Δ) was evaluated for all day 7 recordings and used to compute placeboadjusted change from baseline (ΔΔ) for the moxifloxacin group. Results: The correlation (r) between automatic and manual QT measurements across all recordings was 0.96 for QT and 0.88 for QTcF. The differences between automatic and manual measurements were normally distributed and showed generally longer intervals for the automatic method for both QT: 10.0 ± 7.5 milliseconds (mean ± SD) and QTcF: 10.6 ± 8.3 milliseconds, P b .001. For ΔQTcF-values, the difference between automatic and manual measurements was 1.5 ± 0.1 milliseconds (mean ± SE), P b .001. For placebo-adjusted QTcF change from baseline (ΔΔ), the difference between automatic and manual measurements was 0.7 ± 0.2 milliseconds (mean ± SE), P b .001. There was no difference (P = .40) in mean ΔΔQTcF at the time of maximum change (4 hours postdose) between manual, ΔΔQTcF = 9.4 milliseconds (90% confidence interval, 5.7-13.1 milliseconds) and automatic ΔΔQTcF = 9.5 milliseconds (90% confidence interval, 6.3-12.8 milliseconds). Conclusion: The study showed clear correlation between manual and automatic QT /QTc measurements with a bias toward longer automatic measurements for the individual ECG. This bias is markedly reduced when evaluating changes in QTc—there was no practical difference in ΔQTc and ΔΔQTc values between the 2 methods. Both the automatic and manual methods detected the moxifloxacin-induced QTc prolongation, and there was no significant difference in ΔΔQTc at the time of maximum change between the 2 methods.
doi:10.1016/j.jelectrocard.2009.08.007
doi:10.1016/j.jelectrocard.2009.08.008
Manual and automatic measurements of moxifloxacin-induced QT prolongation M.P. Andersen,a J. Matz, b C. Graff, a J.K. Kanters, c E. Toft, a J.J. Struijk, a ( aAalborg University, Aalborg, Denmark; bH. Lundbeck A/S, Copenhagen, Denmark; cUniversity of Copenhagen, Copenhagen, Denmark)
Philips Microvolt Electrocardiogram Amplitude Periodicity algorithm for detection and quantification of T-wave alternans Saeed Babaeizadeh, Eric D. Helfenbein, Jim M. Lindauer, Sophia H. Zhou, (Advanced Algorithm Research Center, Philips Healthcare, Thousand Oaks, CA, USA)
Background: Accurate measurement of QT intervals is of pivotal importance in cardiac safety trials. Thorough QT studies must be designed to detect an 8- to 12-millisecond ΔΔQTc prolongation in a positive control group treated with moxifloxacin. QT measurements are typically performed manually to maximize measurement accuracy. Automatic QT measurements are more cost-effective, but the ability of
Background: Microvolt Electrocardiogram (ECG) Amplitude Periodicity (MEAP) algorithm, introduced in our other presentation, is able to detect and quantify patterned beat-to-beat variability in multilead ECG recordings. This study examines the performance of this algorithm for detecting and quantifying T-wave alternans (TWA) as one form of patterned behavior in a part of an ECG beat.
Negative correlation relationship between left ventricular hypertrophy Sokolow-Lyon and body mass in 41 806 Swiss conscripts Roger Abächerli, a Richard Kobza, b Bernhard Niggli, b Lingchuan Zhou, a Johann-Jakob Schmid, a Franz Frey, c Paul Erne, b ( aSCHILLER AG, Biomed. Research and Signalprocessing, Baar, Switzerland; bDivision of Cardiology, Cantonal Hospital Luzern, Switzerland; cMedical Service, Swiss Army, Ittigen, Switzerland) Background: Conventional electrocardiographic diagnosis criteria for left ventricular hypertrophy (LVH), such as the Sokolow-Lyon voltage, have been limited by low sensitivity at acceptable levels of specificity. A number of studies have demonstrated that body mass index (BMI) is associated with decreased sensitivity of electrocardiogram (ECG) LVH classification in hypertensive patients. The objective of this study is to investigate the correlation relationship between Sokolow-Lyon voltage criteria and BMI in Swiss conscripts. Methods: A database of 41 806 young Swiss people who underwent compulsory conscription for the Swiss Army was compiled. Along with other medical data, an ECG was taken. Statistical analyses, such as linear regression and calculation of correlation coefficient, were carried out between the Sokolow-Lyon voltages and BMI. Results: The mean (SD) age in the studied population was 19.2 (1.1) years, with a median age of 19 years (range, 17-38 years). We found an overweight prevalence of 25.1%. The results showed that body habitus had significant association with Sokolow-Lyon voltages. A mean decrease of 13%, 5%, 19%, and 14% for the 4 studied indexes were found between reference range subjects (in kilograms per meter squared: 18.5 ≤ BMI b 25) and obese subjects (25 ≤ BMI).
0022-0736/$ – see front matter