Abstract: Data Management and Montitoring in the CombiRx Randomized Trial (The 27th Annual Meeting of the CMSC and the 5th Cooperative Meeting of the CMSC-ACTRIMS)

Thursday, May 30, 2013

Tarah Gustafson, BSN , Corinne Goldsmith Dickinson Center for Multiple Sclerosis, New York, NY

C. Steve Powell, PhD , Biostatistics, University of Alabama at Birmingham, Birmingham, AL

Amber Salter, MPH , Biostatistics, University of Alabama at Birmingham, Birmingham, AL

Jing Wang, MS , Biostatistics, University of Alabama at Birmingham, Birmingham, AL

Fred D. Lublin, MD , Corinne Goldsmith Dickinson Center for Multiple Sclerosis, New York, NY

Jerry Wolinsky, MD , University of Texas Health Science Center, Houston, TX

Robin Conwit, MD , NINDS, NIH, Bethesda, MD

Gary Cutter, PhD , Biostatistics, University of Alabama at Birmingham, Birmingham, AL

Stacey S Cofield, PhD , Biostatistics, University of Alabama at Birmingham, Birmingham, AL

PDF

Background: Randomized clinical trials are large undertakings, financially and administratively. Electronic data capture has reduced the need for on-site data monitoring by allowing for built in rules prohibiting clearly incorrect or out of range data entry. Data queries and on-site monitoring are necessary for in-range transcription errors and validation of key study outcomes. The CombiRx trial allows for assessment of an electronic data system with both on-site and remote data monitoring.

Objectives: Assess the CombiRx data entry, management and monitoring system through the number of data changes made by on-site monitors (SM), remote monitoring (RM), and by the clinical site (CS).

Methods: CombiRx randomized 1008 participants, across 69 sites, with 2 screening visits, a baseline visit, quarterly visits (up to 90 months), and unscheduled visits; including 39 case report forms. Data changes were tracked using a web-based administrative site that recorded the individual making the change, the form, field, original value, new value and reason for change. Changes made at the site to a completed form required an authorization code and recorded similar information.

Results: Over 23,465,000 data fields were entered on more than 267,000 forms between January 2005-March 2012. A total of 22,937 changes were made by monitoring: SM 14,983; RM 17,954; and 2,914 forms were unlocked by 66 sites to make changes. The SM had 8 personnel make changes, RM 12, with no difference in the number of changes made between RM and SM personnel (p=0.82), but requiring SM personnel to execute more than 300 on-site visits. The most common forms with data changes were concomitant medications (6,098 field edits), MSFC (4,220), and Functional Systems and EDSS (4,220); with the least number of changes made to single use forms Terminating Study Therapy (5) and Death Notification (2). Additional analyses of field specific changes and fiscal costs of monitoring will also be presented.

Conclusions: Electronic data capture has allowed for increased monitoring at the site level at the time of data entry. In addition, remote monitoring with statistical analysis allows for monitoring of fields overtime to detect data values out of expected ranges without the travel required for external site monitoring visits. On-site monitoring of data could be reduced with increased remote monitoring personnel and would reduce administrative costs of large clinical trials.

DX32 Data Management and Montitoring in the CombiRx Randomized Trial