Insights from a Data Scientist: eIQ Review Use Case Spotlight | Part 1

The way trials are being conducted has changed significantly with decentralized clinical trials (DCT) becoming mainstream and the collection of clinical data from wearables and other remote-monitoring devices becoming common practice. As a result, the volume and variety of clinical data being captured per trial has grown exponentially – and at a faster pace than ever before.

Managing this deluge of clinical trial data has become challenging and is driving the need to evolve traditional data management approaches. The integration of AI/ML can address the limitations of traditional approaches and provide opportunities for data management teams to conduct clinical data review more efficiently, reduce manual efforts, and ensure data quality.

elluminate^® IQ Review (eIQ Review) provides AI-enabled data review capabilities for several use cases to automate and create efficiencies across clinical data review processes. Two examples of the use cases deployed within eIQ Review are highlighted below.

Concomitant Medication Abnormal Duration (CMAD)

The manual review of data, particularly through exception listings, is a time-consuming and labor-intensive process. One crucial task involves identifying concomitant medications (CM) with unusually short or long durations. While essential for patient safety, this task is prone to errors and inefficiencies. The CMAD model aims to automate parts of this process using machine learning techniques and insights from historical data.

The primary goal of CMAD is to alleviate the burden on data monitors and enhance data review efficiency in clinical trials. These capabilities within eIQ Review can be leveraged to identify approximately 86% of anomalous data, reducing the amount of data that requires review by data managers. As a result, data monitors can avoid reviewing a significant portion of data while still capturing the anomalous data points, greatly reducing their workload.

Automated Adverse Event Duration (AAED)

In the realm of adverse events (AEs), occurrences that exhibit unusually short or prolonged durations often warrant a closer look. Such instances may either hold significant clinical implications or stem from data entry errors. An automated solution for detecting these anomalies can substantially enhance the efficiency of data review processes. The AAED (Automated Adverse Event Duration) model harnesses an unsupervised machine learning algorithm to automatically pinpoint these exceptional adverse events and surface them to users within eIQ Review.

The primary objective of this use case is to shine a spotlight on data points that might otherwise slip under the radar of data reviewers. This crucial task empowers data monitors to uncover data inconsistencies that might easily elude detection in a conventional study scenario. In essence, the AAED module acts as a guardian, safeguarding the integrity of your data by bringing into focus those critical data nuances that would typically go unnoticed, ensuring the highest standards of data quality.

In clinical development, the potential impact of AI/ML is vast, and for data management teams, the integration of AI/ML can address the limitations of traditional approaches that are simply not scalable enough to support that pace at which clinical trials are evolving. In the weeks ahead, we will be highlighting additional eIQ Review use cases in our blog, but in the meantime, if you’d like to learn more about eIQ Review, you can view the Introduction to eIQ Review webinar on demand.

Authors

Yekta Zahed

Senior Data Scientist, eClinical Solutions

Alison Cronin

Director of Product Marketing, eClinical Solutions