Cytel Introduces Advanced Design Framework: Part 2- The Need for A Quantitative Evaluation Approach for Deciding Together
Cytel has recently revealed its Advanced Design Framework, a method developed by Cytel’s thought leaders that draws on decades of experience increasing clinical development productivity. The Framework illustrates how advances in design processes and technology can help development teams deliver greater business results, unifying statistics and strategy in the era of cloud computing, and making strategic use of well-resourced statisticians.
The framework consists of three parts: Thoroughly Explore, Decide Together, and Communicate Trade-Offs. This week we take a deeper look into the second part of this Framework, revealing how to effectively incorporate varied perspectives to efficiently design innovative clinical trials. Opportunities for quantitative evaluation criteria and design without bias help R&D teams sift through the thousands of trial designs options to optimize for speed, success, and savings.
Has your organization ever completed the execution of a long, expensive Phase 3 clinical trial only to learn that the organization is unable to commercialize the therapy? Has a faster competitor ever locked your organization out of a market? One contributing factor may have been the historical process for selecting statistical trial designs. The emergence of cloud-powered trial optimization capabilities warrants an examination of the end-to-end approach to statistical trial design.
The second component of Cytel’s new Advanced Design Framework, “Decide Together”, helps trial sponsors select statistical designs that can successfully deliver greater value to patients and shareholders. Deciding together requires three key improvements:
Unify team perspectives into a consistent, quantitative evaluation approach
Use statisticians strategically as well as tactically
Employ computational power to efficiently evaluate the expansive design space
Barriers to an Effective Selection Process
Optimized clinical trial designs have the potential to accelerate trial completion, increase the probability of success, and reduce trial costs. Recent research demonstrates that the use of innovative trial designs can increase clinical development productivity 10-20%. However, despite their emergence over a decade ago, adoption of these advanced statistical approaches has lagged.
Historically, selecting the right statistical design for clinical trials from among promising options has been a tedious process, and several barriers have inhibited development teams from deciding upon the optimal design. Time and technical constraints made it impractical to explicitly incorporate all cross-functional perspectives, leaving the organization vulnerable to an evaluation approach that may overlook important considerations. Processes at some organizations have evolved to only consult the “core” team members whose perspectives can be modeled. Even when extended team members are consulted, often their perspectives are incorporated qualitatively, opening the team to biases and decisions influenced by rank or politics. This situation becomes further exacerbated when team members find themselves facing competing goals. Often they try to establish a decision process and criteria only after determining the operating characteristics of the options before them, leaving personal bias in the choice of optimal clinical study design. The emergence of cloud-computing provides an opportunity to establish best practices in clinical trial design by expanding identified opportunities and quantifying decision criteria.
1: Unify team perspectives into a consistent, quantitative evaluation approach
It is important to examine the managerial influences on decision-making. Is there an established process for the approach to decisions? Does one key role have decision-making authority or is it decision by consensus? And if consensus, who is on the team? Are all relevant stakeholders included to ensure their perspectives are considered? Will the team establish consistent, quantifiable decision criteria prior to evaluating the various options, and does corporate culture promote data-driven decision-making? Will the team unify the criteria into a quantitative metric of success that supports corporate goals, and are team members’ managerial incentives aligned with that definition?
A quantitative evaluation approach is one that mitigates bias caused by one or more stakeholders having undue influence over discussions about trial design. Team members agree that certain elements of the clinical development situation (e.g. trial costs, trial speed, probability of success) will be given particular weights in decision-making, and these weights are then used to come to a collective decision.
2: Strategically involve statisticians early in the process
Especially as organizations seek the benefits of innovative trial designs to help improve trial speed, success, and savings, it is important to ensure biostatisticians are involved in the strategic decision-making very early in the process. Biostatisticians have expertise in converting risk into opportunity and can facilitate data-driven, quantitative decision-making to better support business objectives if they understand the context. In addition, biostatisticians can introduce additional promising options for consideration if included in the discussion sufficiently early in the process. For example, a biostatistician who knows that a competitor might be edging a product out of the market might have insight into whether to increase sample size to obtain even more rigorous results or give the whole trial an opportunity to stop earlier for futility. A biostatistician who is unaware of the strategic stakes will not know to identify these opportunities for the R&D team.
3: Employ computational power to efficiently evaluate the expansive design space
Traditionally, trial designs were compared by manual scoring, a process by which each trial design under consideration was compared to every other in a process of pairwise comparison. For example, five potential designs would lead to 24 comparisons (4x3x2x1=24).
In the age of cloud-computing, which can surface an exponentially larger set of options for consideration, technology is necessary to help evaluate the options since manual scoring is no longer feasible. Arming the biostatistician with the latest software technology is important to being able to identify the optimal design. For example, Solara enables organizations to evaluate hundreds of thousands of design options within minutes by applying a quantitative decision framework that consistently incorporates critical non-statistical considerations. This approach helps development teams to efficiently identify the optimal design that best support business goals within operational constraints.
Join the first webinar in our C-Suite Webinar Series to learn how preliminary testing of Cytel’s Advanced Design Framework has consistently identified opportunities to increase clinical development productivity 10-20%.
Read part 1 of this blog series here and part 3 here.
Cytel recently released Solara™, a collaborative decision-support platform that unifies statistics and strategy to optimize clinical trial design. Contact us to learn more.