Evaluation of differences in automated QT/QTc measurements between Fukuda Denshi and Nihon Koden systems

Motoaki Sano, Yoshiyasu Aizawa, Yoshinori Katsumata, Nobuhiro Nishiyama, Seiji Takatsuki, Shigeo Kamitsuji, Naoyuki Kamatani, Keiichi Fukuda

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Background: Automatic measurement becomes a preference, and indeed a necessity, when analyzing 1000 s of ECGs in the setting of either drug-inducing QT prolongation screening or genome-wide association studies of QT interval. The problem is that individual manufacturers apply different computerized algorithms to measure QT interval. We conducted a comparative study to assess the outcomes with different automated measurements of QT interval between ECG machine manufacturers and validated the related heart rate correction methods.

Methods and Results: Herein, we directly compared these different commercial systems using 10,529 Fukuda Denshi ECGs and 72,754 Nihon Kohden ECGs taken in healthy Japanese volunteers. Log-transformed data revealed an equal optimal heart rate correction formula of QT interval for Fukuda Denshi and Nihon Kohden, in the form of QTc = QT/RR-0.347. However, with the raw data, the optimal heart rate correction formula of QT interval was in the form of QTc = QT+0.1566(1- RR) for Fukuda Denshi and QTc = QT+0.1526(1-RR) for Nihon Kohden. After optimization of heart rate correction of QT interval by the linear regression model using either log-transformed data or raw data, QTc interval was ,10 ms longer in Nihon Kohden ECGs than in those recorded on Fukuda Denshi machines. Indeed, regression analysis revealed that differences in the ECG machine used had up to a two-fold larger impact on QT variation than gender difference. Such an impact is likely to be of considerable importance when ECGs for a given individual are recorded on different machines in the setting of multi-institutional joint research.

Conclusions: We recommend that ECG machines of the same manufacturer should be used to measure QT and RR intervals in the setting of multi-institutional joint research. It is desirable to unify the computer algorithm for automatic QT and RR measurements from an ECG.

Original languageEnglish
Article numberA1326
JournalPLoS One
Volume9
Issue number9
DOIs
Publication statusPublished - 2014 Sep 17

Fingerprint

Electrocardiography
heart rate
Heart Rate
gender differences
Linear Models
volunteers
regression analysis
screening
drugs
Genome-Wide Association Study
Research
Linear regression
Regression analysis
methodology
Healthy Volunteers
Screening
Genes
Regression Analysis
Pharmaceutical Preparations

ASJC Scopus subject areas

  • Agricultural and Biological Sciences(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Medicine(all)

Cite this

Evaluation of differences in automated QT/QTc measurements between Fukuda Denshi and Nihon Koden systems. / Sano, Motoaki; Aizawa, Yoshiyasu; Katsumata, Yoshinori; Nishiyama, Nobuhiro; Takatsuki, Seiji; Kamitsuji, Shigeo; Kamatani, Naoyuki; Fukuda, Keiichi.

In: PLoS One, Vol. 9, No. 9, A1326, 17.09.2014.

Research output: Contribution to journalArticle

Sano, Motoaki ; Aizawa, Yoshiyasu ; Katsumata, Yoshinori ; Nishiyama, Nobuhiro ; Takatsuki, Seiji ; Kamitsuji, Shigeo ; Kamatani, Naoyuki ; Fukuda, Keiichi. / Evaluation of differences in automated QT/QTc measurements between Fukuda Denshi and Nihon Koden systems. In: PLoS One. 2014 ; Vol. 9, No. 9.
@article{9e8f27f11d2b4a0fb78ba2a32d9f3f91,
title = "Evaluation of differences in automated QT/QTc measurements between Fukuda Denshi and Nihon Koden systems",
abstract = "Background: Automatic measurement becomes a preference, and indeed a necessity, when analyzing 1000 s of ECGs in the setting of either drug-inducing QT prolongation screening or genome-wide association studies of QT interval. The problem is that individual manufacturers apply different computerized algorithms to measure QT interval. We conducted a comparative study to assess the outcomes with different automated measurements of QT interval between ECG machine manufacturers and validated the related heart rate correction methods.Methods and Results: Herein, we directly compared these different commercial systems using 10,529 Fukuda Denshi ECGs and 72,754 Nihon Kohden ECGs taken in healthy Japanese volunteers. Log-transformed data revealed an equal optimal heart rate correction formula of QT interval for Fukuda Denshi and Nihon Kohden, in the form of QTc = QT/RR-0.347. However, with the raw data, the optimal heart rate correction formula of QT interval was in the form of QTc = QT+0.1566(1- RR) for Fukuda Denshi and QTc = QT+0.1526(1-RR) for Nihon Kohden. After optimization of heart rate correction of QT interval by the linear regression model using either log-transformed data or raw data, QTc interval was ,10 ms longer in Nihon Kohden ECGs than in those recorded on Fukuda Denshi machines. Indeed, regression analysis revealed that differences in the ECG machine used had up to a two-fold larger impact on QT variation than gender difference. Such an impact is likely to be of considerable importance when ECGs for a given individual are recorded on different machines in the setting of multi-institutional joint research.Conclusions: We recommend that ECG machines of the same manufacturer should be used to measure QT and RR intervals in the setting of multi-institutional joint research. It is desirable to unify the computer algorithm for automatic QT and RR measurements from an ECG.",
author = "Motoaki Sano and Yoshiyasu Aizawa and Yoshinori Katsumata and Nobuhiro Nishiyama and Seiji Takatsuki and Shigeo Kamitsuji and Naoyuki Kamatani and Keiichi Fukuda",
year = "2014",
month = "9",
day = "17",
doi = "10.1371/journal.pone.0106947",
language = "English",
volume = "9",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "9",

}

TY - JOUR

T1 - Evaluation of differences in automated QT/QTc measurements between Fukuda Denshi and Nihon Koden systems

AU - Sano, Motoaki

AU - Aizawa, Yoshiyasu

AU - Katsumata, Yoshinori

AU - Nishiyama, Nobuhiro

AU - Takatsuki, Seiji

AU - Kamitsuji, Shigeo

AU - Kamatani, Naoyuki

AU - Fukuda, Keiichi

PY - 2014/9/17

Y1 - 2014/9/17

N2 - Background: Automatic measurement becomes a preference, and indeed a necessity, when analyzing 1000 s of ECGs in the setting of either drug-inducing QT prolongation screening or genome-wide association studies of QT interval. The problem is that individual manufacturers apply different computerized algorithms to measure QT interval. We conducted a comparative study to assess the outcomes with different automated measurements of QT interval between ECG machine manufacturers and validated the related heart rate correction methods.Methods and Results: Herein, we directly compared these different commercial systems using 10,529 Fukuda Denshi ECGs and 72,754 Nihon Kohden ECGs taken in healthy Japanese volunteers. Log-transformed data revealed an equal optimal heart rate correction formula of QT interval for Fukuda Denshi and Nihon Kohden, in the form of QTc = QT/RR-0.347. However, with the raw data, the optimal heart rate correction formula of QT interval was in the form of QTc = QT+0.1566(1- RR) for Fukuda Denshi and QTc = QT+0.1526(1-RR) for Nihon Kohden. After optimization of heart rate correction of QT interval by the linear regression model using either log-transformed data or raw data, QTc interval was ,10 ms longer in Nihon Kohden ECGs than in those recorded on Fukuda Denshi machines. Indeed, regression analysis revealed that differences in the ECG machine used had up to a two-fold larger impact on QT variation than gender difference. Such an impact is likely to be of considerable importance when ECGs for a given individual are recorded on different machines in the setting of multi-institutional joint research.Conclusions: We recommend that ECG machines of the same manufacturer should be used to measure QT and RR intervals in the setting of multi-institutional joint research. It is desirable to unify the computer algorithm for automatic QT and RR measurements from an ECG.

AB - Background: Automatic measurement becomes a preference, and indeed a necessity, when analyzing 1000 s of ECGs in the setting of either drug-inducing QT prolongation screening or genome-wide association studies of QT interval. The problem is that individual manufacturers apply different computerized algorithms to measure QT interval. We conducted a comparative study to assess the outcomes with different automated measurements of QT interval between ECG machine manufacturers and validated the related heart rate correction methods.Methods and Results: Herein, we directly compared these different commercial systems using 10,529 Fukuda Denshi ECGs and 72,754 Nihon Kohden ECGs taken in healthy Japanese volunteers. Log-transformed data revealed an equal optimal heart rate correction formula of QT interval for Fukuda Denshi and Nihon Kohden, in the form of QTc = QT/RR-0.347. However, with the raw data, the optimal heart rate correction formula of QT interval was in the form of QTc = QT+0.1566(1- RR) for Fukuda Denshi and QTc = QT+0.1526(1-RR) for Nihon Kohden. After optimization of heart rate correction of QT interval by the linear regression model using either log-transformed data or raw data, QTc interval was ,10 ms longer in Nihon Kohden ECGs than in those recorded on Fukuda Denshi machines. Indeed, regression analysis revealed that differences in the ECG machine used had up to a two-fold larger impact on QT variation than gender difference. Such an impact is likely to be of considerable importance when ECGs for a given individual are recorded on different machines in the setting of multi-institutional joint research.Conclusions: We recommend that ECG machines of the same manufacturer should be used to measure QT and RR intervals in the setting of multi-institutional joint research. It is desirable to unify the computer algorithm for automatic QT and RR measurements from an ECG.

UR - http://www.scopus.com/inward/record.url?scp=84907217705&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84907217705&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0106947

DO - 10.1371/journal.pone.0106947

M3 - Article

VL - 9

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 9

M1 - A1326

ER -