W 26: Impact of Internal Data Review and Source Data Verification on Overall Data Quality

Poster Presenter

Ron Taylor

Director, Clinical Trial Management
Seattle Genetics, Inc.
United States

Objectives

The primary objective is to measure the impact data review has on the overall integrity of Seattle Genetics’ clinical data on 14 oncology studies and determine if it supports recent evidence suggesting that SDV has minimal effect on data integrity (Transcelerate Biopharma Inc. May 2013)

Method

An analysis of data changed as a result of the review of data at the clinical sites and internally at Seattle Genetics is presented for each of 14 studies and is further compared in two global Phase 3 trials with the key difference being the approach to reviewing the data at the sites.

Results

The overall data change rate averages 5.3% across the 14 studies evaluated. This includes data that may change from original entry as the result of system edit checks detecting general entry errors and as the result of data review processes that correct discrepancies. The data change rate resulting from the resource-intensive data review processes (both from source data verification and internal centralized monitoring review) averages 2.5% across the 14 studies evaluated. A detailed comparison between the two global Phase 3 trials was performed, both of approximately the same size and complexity, with the key difference being the approach to reviewing the data at the sites. One used the industry standard of 100% SDV; the other used a risk-based monitoring approach that focused more effort on critical variables (variables identified as directly impacting study endpoints) and less on non-critical variables. Results show minimal differences in the data change rates (2.1% vs 1.7%). This is primarily due to differences in data change rate of critical variables (3.4% vs 2.8%), with minimal impact to the data change rate for non-critical variables.

Conclusion

Of the 14 oncology trials evaluated, only 2.5% of the data changed from original entry as a result of SDV and internal data review, supporting earlier research (Transcelerate Biopharma Inc. May 2013; Medidata Solutions. April 2012). Analysis suggests that the reduction in SDV has not impacted data quality on the global Phase 3 trial using Risk Based Monitoring. Analysis supports further expanding a risk-based approach to source data verification and to internal data review.