A proposal for a new approach to intergroup cancer trials

Area of Science:

• Clinical Trials
• Data Management
• Biostatistics

Background:

• North American intergroup trials traditionally rely on a single lead group for data management.
• This centralized approach, while successful in patient accrual, creates significant data flow challenges for statistical centers.
• Previous successes with pooled analyses from independently conducted trials informed a new data management strategy.

Purpose of the Study:

• To propose and evaluate an alternative data management model for North American intergroup trials.
• To address the data flow bottlenecks experienced with the current centralized data management system.
• To maintain the integrity and features of intergroup studies while enhancing data management efficiency.

Main Methods:

• Implementing a common protocol and data capture elements across participating groups.
• Assigning responsibility for data collection and quality control to each individual group.
• Establishing a common, periodically updated data set for interim and final analyses.
• Utilizing a pooled analysis approach based on independently managed data.

Main Results:

• Demonstrated success in combining data from three independently conducted trials for a planned pooled analysis.
• Experience gained from pooled analysis informed the design of a forthcoming intergroup trial using a decentralized data management approach.
• The proposed model aims to mitigate data flow peaks by distributing data management responsibilities.

Conclusions:

• The proposed decentralized data management model offers advantages over the current lead-group responsibility system for intergroup trials.
• This approach maintains the distinct characteristics of intergroup studies, differentiating them from meta-analyses.
• Shared data collection and quality control can improve the efficiency and feasibility of large-scale intergroup clinical trials.