2021 FDA Science Forum
Validation Framework for Datasets from QT Studies: Leveraging Data Standards to Enable Automation
- Authors:
- Center:
-
Contributing OfficeCenter for Drug Evaluation and Research
Abstract
Since 2005, all new drugs are required to undergo a clinical study to assess their potential to prolong the QT interval in the electrocardiogram and cause a potentially fatal arrhythmia. ICH E14 guideline provides recommendations concerning the design, conduct, analysis, and interpretation of this QT assessment. To reduce the data management time required to produce analysis ready datasets from data collected in these clinical studies, the Interdisciplinary Review Team for Cardiac Safety Studies (IRT) recently published a Technical Specifications Document for QT Studies (QT-TSD), which is based on existing data standards. This research investigated whether there are automation opportunities that could reduce the time needed to validate datasets submitted following the QT-TSD. A systematic review of the validation process of QT-TSD datasets documented the steps that could be automated in human- and machine-friendly spreadsheets and python scripts (plugins). The spreadsheets list QT-TSD data elements and their associated validation rules. Plugins implement the algorithm of each validation rule. Using these spreadsheets and plugins, an automatic validation framework in python reproduces the manual validation of any QT-TSD dataset as follows: the validator takes the QT-TSD dataset as input, reads and iterates through the spreadsheets rows executing the plugin rules, and produces a validation report. To facilitate fixing dataset errors, the report includes description for errors found in the dataset. Lastly, if the validation is passed without errors, the framework also derives an analysis ready dataset that can be used out-of-the-box by IRT’s analysis tools. Validation of 6 QT-TSD datasets showed the potential of the framework to speed up the review time of QT-TSD datasets, although further testing is ongoing. In addition, the validation framework can be applied to other therapeutic areas (e.g., ambulatory blood pressure monitoring) by replacing the QT-TSD spreadsheets and plugin rules with those of the therapeutic area of interest. Technical Specifications Documents like the QT-TSD facilitate efficient interchange of clinical datasets, usually with a focus on a specific area. This framework illustrates that TSDs also present an opportunity to reduce the time needed to validate, analyze and review clinical datasets when analysis needs and processes are coupled and well-defined.
