An In-Process Scaling Model: A Potential Framework for Data Monitoring Committees and Clinical Trial Quality Improvement
- PDF / 195,192 Bytes
- 5 Pages / 602.986 x 782.986 pts Page_size
- 40 Downloads / 196 Views
An In-Process Scaling Model: A Potential Framework for Data Monitoring Committees and Clinical Trial Quality Improvement
Drug Information Journal 46(1) 8-12 ª The Author(s) 2012 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0092861511427864 http://dij.sagepub.com
Alan Smith, MS1, and Jonathan Seltzer, MD, MBA, FACC2
Abstract This article discusses the application of a statistical process control methodology to improve the quality of real-time clinical trial monitoring, an especially valuable tool for data monitoring committees. This article outlines a method that may have value in bringing a data-based approach using the measures of site performance to define patterns of similarity or dissimilarity between the sites. These can then be used to provide an efficient and objective mechanism to identify those sites for closer scrutiny, monitoring, or further training, either at the end of the study or on an ongoing basis. Keywords clinical monitoring, adverse event, statistical process controls, metric, data monitoring committee, data safety monitoring board, multidimensional scaling, unblinding
Introduction Currently, clinical trials lack a widely accepted effective quality management system—that is, a framework that embodies continuous feedback and improvement, first described by Deming.1 Although this does not exist for clinical trials, the quality by design efforts for pharmaceutical medicine manufacture embodied in International Conference on Harmonization (ICH) guidances Q8 and Q9 establish a quality systems framework and are thought to both improve quality and reduce regulatory oversight activities. Application of a similar system for clinical trials has been urged at the highest levels of the Food and Drug Administration (FDA). Currently, however, quality in clinical trials is guided by Good Clinical Practices and ICH guidelines. The most widely accepted method of quality control seems to be clinical monitoring. Recent work by the Clinical Trial Transformation Institute (CTTI) finds that approximately 80% of industry studies employ clinical monitors as a quality management system, but that number drops to less than a third when looking at academia, government, and nongovernmental organization sponsored studies. Despite the widespread use of monitoring, CTTI reports that monitoring strategies vary widely, and there is little evidence to support any particular monitoring approach.
An exception may be the ADAMON project in Germany, for which specific risk-based monitoring strategies have been implemented prospectively in an effort to improve quality and reduce overall monitoring burden.2 However, even in this latter case, clinical monitoring, because it is by nature a post hoc compliance activity, is an ‘‘inspection-based’’ activity, not a ‘‘quality-based’’ system. We believe that the emergence of real-time data, in concert with clinical data standards, will allow for the development and deployment of clinical trial statistical process controls (SPCs). Statistical process controls a
Data Loading...
