Ensuring Uniform Protocol Revision Usage
in Multi-Site
Randomized Prospective Clinical Trials
Ensuring that the correct version of any given protocol is in use by all the centers in a large clinical trial can present challenges to any sponsor. Often, it is a time consuming and lengthy process that may require the trial be halted and subsequently restarted to ensure site-level compliance. Additionally, there may be regulatory compliance issues associated if the protocol change required relates to patient safety. Beyond the burden imposed on the sponsor or CRO in managing the process, stop-start delays can be costly in terms of lost strategic opportunity as it relates to product approval or claim generation, and thus any protocol-version change implementation process must concomitantly address burden, timeliness of resolution, and regulatory compliance.
This paper briefly describes the methodology employed by a major medical device company (Guidant Corporation, Indianapolis, IN) to affect a uniform protocol-version change throughout a 112 site, 1571 patient (at enrollment closure), randomized, prospective, international clinical trial during late 2003.
Situation
In the INTRINSIC RV study, an electronic research organization (eRO) had been retained by Guidant to provide data capture, management and reporting services. The eRO, MedNet Solutions Inc. (Minneapolis, MN), provided advanced technology solutions to affect these services, including web-based data capture and real-time data analytics.
Review of early, non-endpoint related data made available "at will" by the eRO via their implementation of the web-based system caused the Principal Scientist for the project to raise concerns with management over the protocol required settings for the implanted device and their effects on patient randomization. Further analysis indicated that a protocol amendment was required urgently in order to adequately address the issue and ensure a sufficient number of patients were available for endpoint analysis within the original total "n" projections and assumptions of the statistical analysis plan.
© Timothy Pratt 2005
