American Statistical Association
New York City
Metropolitan Area Chapter

Mailman School of Public Health
Columbia University
Department of Biostatistics Colloquium



Yi Tsong
Food and Drug Administration


QT interval is defined as the time from beginning of ventricular depolarization through repolarization as seen on the ECG. QT interval lengthens as heart rate decreases. Medically, drug induced prolongation of QT interval is considered as a surrogate for arrhythmic or “torsade de pointes.” There were numerous drugs withdrawn from market or restricted its prescription due to torsadogenic event since 1988, regulatory agent required QT trials before marketing approval. International Community of Harmonization defined a QT prolongation free drug as “A negative thorough QT/QTc study is one in which the upper bound of the 95% one-sided confidence interval for the largest time-matched mean effect of the drug on the QTc interval is <= 10ms. This definition is chosen to provide reasonable assurance that the mean effect of the study drug on the QT/QTc interval is not greater than around 5ms, which is the threshold level of regulatory concern” in E14. Accordingly, the intersection-union test of mean QTc difference between test treatment and placebo adjusted for baseline measurement at each time point against a non-inferiority margin 10 ms is considered as the standard approach. Although it was encouraged to study QTc differences only at and around maximum concentration, IU test was often applied to more than 5 time points and sometimes suffered with power lost with designated sample size. A few alternative approaches were proposed to compare the mean QTc difference between test treatment and placebo at maximum concentration of test drug with the non-inferiority margin. This approach requires a precise prediction of QTc difference at the concentration not measured through concentration-QTc difference modeling. Theoretically, it is expected to be more conservative than the E14 method when the test drug prolongs QTc interval following the assumed model. The conventional linear or linear-log model used to assess QT prolongation is evaluated with five data sets. The results indicate that the conventional models lead to biased results and more complicated models are needed.

Date: Thursday, October 11, 2007
Time: 4:00 - 5:00 P.M.
Location: Mailman School of Public Health
Department of Biostatistics
722 West 168th Street
Judith Jansen Conference Room
4th Floor - Room 425
New York, New York


Refreshments will be served at 3:30 P.M. in the
Biostatistics Conference Room (R627).

Home Page | Chapter News | Chapter Officers | Chapter Events
Other Metro Area Events | ASA National Home Page | Links To Other Websites

Copyright © 2007 by New York City Metropolitan Area Chapter of the ASA
Designed and maintained by Cynthia Scherer
Send questions or comments to