The new diabetes guidance has reinforced the fact that cardiac safety has become the No. 1 reason for drug withdrawals and labeling changes during the past several years. Many clinical trial sponsors do not consider the consequences of decentralized ECG collection and review when planning their pivotal studies. This can result in data problems that would generally be addressed throughout the research process by a central ECG laboratory.

These processes include critical activities such as reconciling data, correcting demography issues, consistently measuring interval duration measurements, interpreting the ECGs consistently, and managing the ECG database and database lock. The decentralized approach, where the ECG is interpreted at the site and the findings documented in the case report form, can result in transcription errors and other unexpected results due to an investigator’s inability to adequately interpret the ECG data. Most importantly, the site reading of these ECGS will likely contain readings that are difficult to analyze such as probable and possible myocardial infarcts, poor R wave progression, etc. When a meta-analysis is performed to assess the number of cardiovascular events on diabetes compounds, these findings will make the interpretation of the data more difficult.  A database that contains centralized readings is likely to have assessed the entire subject and should provide readings that are consistent, for each subject, throughout the trial.

Proper safety assessments require analytical and operational rigor at every level and, for cardiac safety, the use of an ECG core lab is essential for optimal data integrity, consistency, and accuracy.