By Michelle Marlborough and Joshua Pines   

December 19, 2011 | Guest Commentary | Protocol complexity continues to increase as studies target ever more narrow patient populations and organizations embrace adaptive designs, in order to do more with less. Additionally, many organizations view detailed protocol feasibility as a luxury that relies on access to key investigators or external data sources, and for which there is often not enough time to analyze and adopt the feedback. What is lacking in most protocol design processes is an understanding of how the past can be an indicator of a study’s feasibility and likelihood of success. By thinking of themselves as data archaeologists, today’s protocol authors can leverage the information from previous studies, as well as their operational outcomes, to ensure better protocol design and guarantee operational feasibility.   

Hunt for Past Protocol or Find Yourself in The Temple of Doom    

Every sponsor conducting clinical trials has at their fingertips a huge amount of information about the likelihood of a trial’s success—from enrollment rates for studies with similar eligibility criteria, to dropout rates in studies utilizing specific procedures, to the most common reasons for needing to amend protocols in a specific therapeutic area. The challenge in leveraging this mountain of information is that all of the critical items needed to guide protocol design are buried deep in a series of unsearchable, disconnected documents.   

Much as an archaeologist might use a flat-edge shovel to scrape away the upper layers of earth, then a hand pick followed by small brushes to clear away the final bits of dirt to reach the artifact being tracked, a number of tools must be in place to enable authors to drive smarter protocol design with data from both their own organization and industry benchmarks.    

The protocol author’s shovel is the ability to search for studies with a similar set of design characteristics. Adoption of a structured approach to protocol authoring can enable organizations to establish a repository of design information as part of standard authoring activities.   

Structured protocol simply means taking information that is typically written in a document and specifying it in a way that allows a study’s metadata to be captured. For example, a protocol’s title contains many important pieces of metadata about a trial. Therefore, instead of having a paragraph in a document that reads, “A Phase 3, Multicenter, Double-blind, Randomized, Placebo-controlled, Parallel-group Study to Evaluate the Safety and Efficacy of Drug X in Patients with disease Y,” you can alternately capture this in a structured format.    

Thus, it becomes not only possible to include the right textual information in a document automatically, but also simple to search for studies of the same design, and to reuse the information for submission to disclosure registries. This concept can and should be expanded to all study design components, including objectives, endpoints, eligibility criteria, and schedules.   

Our hand pick is the addition of defined relationships between the structured core clinical components. For example, the linking of related objectives, endpoints, procedures, and CRFs develops a line of sight that allows benchmark data and other advanced analytics tools to provide the author feedback on the feasibility of the study they have designed.   

The final tool, the fine grain brush, is the ability to capture key operational information and indicators of success alongside the design criteria: enrollment rate, number of sites, dropout rate, and reasons for protocol amendments.   

Grab Your Fedora and Whip Your Protocol Design Process into Shape   

Once a database of structured protocol designs has been established and links are created among the protocol procedures and clinical and operational data, it is possible to analyze the study design in light of both your organization’s history and also the industry’s experience.   

For instance, it is possible to compare the study design with the standard of care for the disease under consideration. If any procedure, particularly a complex or invasive one, is being conducted more times than typical in the disease area, patients are unlikely to consent. Having this data (see figure 1 for an example) at the author’s fingertips during the design process, as well as knowing the reason for the procedure—via the linked structured data—allows teams to optimize the design and thus improve the feasibility of the study.   

Employing advanced analytical methods to compare the study design to other studies of a similar design type can help provide key performance indicators in multiple areas critical to study design and planning. Protocol authors can use it to better understand typical performance information such as enrollment rates, dropout rates, and number of sites needed. However, in order to capitalize on these capabilities, sponsors must think about their protocol development in a fundamentally different way.    

Just as modern archaeology is built on the fundamental belief that artifacts of a previous society should be studied to understand their culture, so too must past performance in protocol design be explored to properly comprehend data from the studies associated with that protocol. Few sponsors currently do this in a replicable manner. In order to avoid losing this critical institutional knowledge, they must adopt a study design process that leverages structured protocol. Otherwise, they might as well be storing prior study information in an unmarked wooden crate.   

Michelle Marlborough is director of product management and Joshua Pines is senior product marketing manager for Medidata Solutions.