Computable Protocol Models for Interactive Trial Design

Abstract

Clinical trial protocols must clearly describe complex relationships between study events, treatments, patient safety, and data collection/monitoring tasks. Protocol writers must be constantly aware of the trade-offs between scientific, regulatory, and operational requirements and constraints. Many features of a protocol's design could impact on the likelihood of success when that trial is placed into the field. Yet, protocol authors often do not have access to information on potential operational issues as they write the protocol. Without this information, study designs that are scientifically sound may be exceedingly difficult or expensive to operationalize successfully.

We present a new methodology for evaluating a protocol for operational clarity and risk assessment during the design of the protocol. We introduce a new technology based on formal computable models for representing clinical trial protocols and studies. In creating computable models, many operational ambiguities surface. From these computable models, numerous quantitative analyses can be performed to understand the key operational assumptions and risk elements inherent in the proposed design. Because the end product is a structured protocol model, execution tools such as site management and study monitoring tools can be automatically configured to execute the model.