To maintain high-quality data in clinical trials, it is crucial to be able to detect systematic anomalies in time series data at the site and subject level, which often stem from protocol misinterpretation or device miscalibration. Traditional approaches can be slow and resource intensive. We therefore aim to develop an analytics approach that can be run with minimal effort and with a high degree of reliability.

Goal

  • To develop a statistical analytics package that detects inconsistencies in subject time series data at both site and subject levels, facilitating timely intervention.
  • To validate the package by piloting it at member companies and explain its utility

Expected Outcomes

  • Open-source r package that is ready for use by the industry
  • White paper sharing the results of piloting the package by member companies and its utility
  • Webinar demonstrating the use of the package

Lead(s)

Resources

Open-source Clinical Trial Anomaly Spotter (CTAS) R Package

The Clinical Trial Anomaly Spotter (CTAS) is a powerful open source tool for Central Statistical Monitoring that identifies outliers and anomalies efficiently and accurately in clinical trial time series. Its main focus is on flagging sites with one or more study parameters whose profiles differ from those of the other sites. In addition, the results can be used to identify anomalies for individual subjects.

Development of the package was spearheaded by Pekka Tiikkainen, Principal Clinical Data Scientist at Bayer, and tested and adapted for user flexibility across organizations by members of IMPALA’s Anomaly Detection in Clinical Data Work Product Team. The underlying algorithm of the CTAS operates by defining one or more time series for each parameter. The algorithm summarizes time series into a set of features such as Mean, Standard deviation, Unique value count, and Autocorrelation. These features help in identifying individual subjects with suspicious data.

CTAS represents a significant advancement in clinical trial data analysis; however, IMPALA’s vision for CTAS extends beyond its current capabilities and usage. CTAS has the potential to be an industry-standard tool that can significantly enhance the integrity and reliability of clinical trial data, leading to more accurate research outcomes and ultimately, better patient care.

IMPALA proudly invites all interested partners to test, utilize, and provide feedback for this innovative package.