Supercharging Quantitative Decision-Making with Simulation-Guided Trial Design
Those familiar with simulation-guided design (SGD) know that it can be used for a wealth of clinical trial options: endpoint selection, number and timing of interim analyses, hitting recruitment targets, managing risk, and building in trial efficiencies. Recently, decision-makers have commented on SGD’s ability to “super-charge” quantitative decision-making for “productive evidence-based conversations.” In other words, SGD has gone from serving as a vital tactical tool, to one with the power to frame and permeate essential strategic conversations. A subtle but significant paradigm shift, a supercharged quantitative decision-making framework affords trial sponsors insights that they have never had before.
Building a map of the strategic terrain
According to Dr. Yannis Jemiai and Dr. Albert Kim, decision-makers might be aware of the number of uncertainties facing a clinical trial, from delayed treatment effects to fluctuating enrollment rates. The way in which these individually affect a clinical trial are hard enough to quantify — but such quantification is also insufficient. Having to account for their collective impact on the speed and operational costs of a clinical trial is crucial but raises even more challenges. Unfortunately, even this set of insights is not enough. A full view of the strategic terrain needs to offer decision-makers visibility to variability in estimates of uncertainty. This requires those powerful tools that are reliably provided by certain forms of SGD.
The Solara® clinical strategy platform, for example, uses massive cloud compute to simulate tens of millions of modelled events in minutes. This allows product development teams ample opportunity to pressure-test against a number of scenarios that might arise in the course of a clinical trial, carefully exploring every tradeoff and opportunity with quantitative insight.
Unlocking funding opportunities
Simulation-guided design has sometimes been used by smaller pharma and biotechs to attract venture capital. A recent case study highlights how a Cytel client used simulation-guided design to alter the financial prospects of investing in a clinical trial, while also de-risking the trial from a medical and clinical perspective.
The client chose a staged-investment strategy, which aligns R&D decisions as well as financial planning with the interim looks of a clinical trial (Case Study page 3). In less than 30 minutes, Solara simulated 59 million trials to ensure that this strategy also minimized the number of patients and recruitment sites, while buoying up statistical power and technical probability of success.
While each of these tactics might have occurred to various strategic consultants, their cumulative benefit is far more intuitively captured using SGD. According to the client’s CMO: “Stage-gating investments require innovative designs. Finding innovative designs that are simple to operationalize requires the type of rapid design generation facilitated by Solara.”
The new optimization problem
Supercharged quantitative decision-making sometimes arises with what we have previously called the New Optimization Problem. The amount of data that tens of millions of simulated models now available, transforms not only the answers received but also the questions we might ask of data: are we leaving opportunities on the table (discoverable through Pareto Analysis)? Are we prioritizing the right business parameters (insights gleaned by scoring functions)?
As questions become more evidence-driven, new methods necessarily arise. While statisticians can help explicate on these questions, it is worthwhile for decision-makers to familiarize themselves with what information these new methods can provide. This will ensure that these simulations guide business specific strategic goals.
Read more from Perspectives on Enquiry and Evidence:
Sorry no results please clear the filters and try again
Simulation-Guided Design Is Reshaping Clinical Trial Strategy
Bayesian Methods across the Clinical Development Journey
Topics in Bayesian Statistical Methods
If you'd like updates on our blog posts, sign up for email updates below.