Good Practices and Research Reproducibility in the Department of Veterans Affairs (VA) Cooperative Studies Program

• Activity Number: 289
• Type: Topic Contributed
• Date/Time: Tuesday, August 2, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistical Consulting
• Author(s): Rebecca B. McNeil*+ and Domenic J. Reda
• Companies: Durham VA Medical Center and Hines VA Hospital
• Keywords: reproducible research ; consulting statistics ; good practice

Abstract

The VA Cooperative Studies Program (CSP) provides coordinating centers for the development and implementation of qualified investigator-initiated clinical trials. Global SOPs for CSP address issues of reproducibility and good practice using four approaches. First, the coordinating center maintains independence from the principal investigator with respect to quality management, despite acting as scientific collaborators. Second, the statistician's role is expanded to include that of team lead, with responsibility for protocol development and management of the internal merit review submission, study start-up, and extensive monitoring. Reflecting this increased level of responsibility, the statistician is included in the Planning and Executive Committees. Third, commercial co-sponsor involvement in data monitoring committee organization and manuscript preparation is prohibited. Finally, a rigorous validation process is required during the development of data management systems and statistical analysis code. We provide a detailed overview of these reproducibility and good practice approaches, and discuss their perceived effects on the statistician, team dynamics, and research quality.