Applying Systems Thinking to the Future of Clinical R&D

Conversations about the future of clinical R&D usually center on what's coming next and how to capitalize on it. Less often discussed is whether the industry is positioned to use what's coming next any differently than it used what came before. A decade of investment in modernization has produced refinement at the margins, but the fundamentals of how we approach clinical development remain unchanged.

The reasons are largely structural. Drug development carries enormous risk, causing leaders to value stability and reliability over more risk-taking. Clinical research is a coordination challenge involving a web of stakeholders and handoffs inherent in the process. Each new study is its own special snowflake with a custom design and system architecture. Those dynamics shape how lasting change happens (or doesn’t) across the industry. It may explain why the cost and timeline for development has not improved over decades. We keep reverting to the mean despite successive eras of innovation.

Meaningful, scalable change in clinical R&D will be shaped less by what new tools become available—including “AI-powered” ones—and more by the resolve of industry leaders to re-wire what’s possible at a system level. Systems thinking is vital to the future of this industry. And it starts with personal curiosity.

Adopting a posture of curiosity

It’s easy to look around the drug development landscape and wonder, “Who designed this?” The simple answer is no one. Every participant in this relay race is driven by specific goals, incentives and constraints (real or perceived) that drive rational behavior from a siloed perspective. What we see as a fragmented and complex ecosystem is simply the inevitable result of these factors playing off one another.

We are unlikely to change a system we don’t comprehend on a structural level. One of the most important roles a leader can play for their organization is to be an avid observer of the workflows unfolding within their teams and to seek to understand more deeply the system at work. Taking a bird’s eye view of the complete system within your horizon, including inflows and outflows, can shine a light on what levers may have meaningful rather than marginal impact. It can show you the outcome the system is set up to produce. It may surprise you that it isn’t the outcome you or others have intended.

This posture can feel several steps removed from the presenting problem or the conventional wisdom around how to solve it. But take the example of slow enrollment: Interventions that take place downstream, after the problem is first noticed, may involve a mixture of increased direct-to-patient recruitment or assigning new third parties to rescue campaigns. Whether these actions are effective in the short term is one question. Another is whether treating potential root causes that exist further upstream—often before or during protocol design—might have prevented the problem.

We know intuitively that anticipating and addressing issues upstream can be far less costly. What we may understand less is how the system disincentivizes this behavior and what levers we might start pulling to reinforce more intentional design.

Visualizing the system around you

As any writer or architect knows, the mental model you have in your head is worth little until you commit it to a medium that exists outside yourself.

This is why having visual maps of the systems at work in clinical research can help. System maps turn invisible structure and hidden patterns into a concrete schema. They are not linear—instead, they illustrate the dynamic relationships and interdependencies between actions, the inflows and outflows, and the pressure points. They reduce the cognitive load necessary to fully understand a problem simply by wrestling it down on paper.

The beauty of system maps is that they can be zoomed in or out based on the topic of concern. You can map the entire healthcare system in the US or you can map the waiting area in a specialist’s office. You can map study startup or you can map Visit 2. The point is to shed light at a scale reasonably sufficient to include your presenting problem (more than likely a symptom) and the influences and stakeholders surrounding it. If your map doesn’t invite new questions or highlight a potential leverage point or constraint, you probably need to widen its boundaries.

Drawing a map may sound trite, particularly in the age of AI. But the most important reasons to visualize are to become clear on your own assumptions and to invite debate and feedback from others. If you are willing to be told why your map is wrong and humble enough to change it, you are practicing your own version of the scientific method. With regular practice and use of visual system maps, we make abstract understanding more concrete. And we have a basis from which to identify the right problems to solve.

Optimizing for the whole

A common system trap is to seek the wrong goal. And in clinical research, we are inundated with solutions that have been optimized for a goal that is too narrowly defined. When these solutions interact with the broader system a user must operate within, the result is uneven (across sites and across studies). At their worst, they lead to unintended consequences that result in poorer system performance overall.

Modernizing in the right way means looking at solutions in context. A platform that streamlines one workflow can still add complexity to a site's day if it doesn't account for the eight other systems that workflow now sits next to.

The goal of interoperability in clinical research illustrates a similar dynamic. The technical capability to exchange study specification data across systems is no longer the primary constraint. Standards exist. Frameworks exist. The path forward is reasonably well understood by the people doing the work.

What's harder is the decision to begin adopting those standards and for this decision to be repeated by multiple stakeholders at scale. Organizations have invested significantly in their existing technology environments. Sites have done the same. Moving toward more interoperable models requires absorbing additional cost on top of investments already made, and it requires prioritizing long-term efficiency over short-term convenience. It is rooted in a higher-order thinking that benefits the system as a whole.

Building a future

The work of structural change in clinical R&D is often described as something the “industry” needs to do. But it is more accurate to say it is something individual leaders, in individual organizations, in individual conversations, choose to do or not do. Those choices accumulate across an ecosystem that spans sponsors, sites, regulators, service providers and patient groups. The choice may be as simple as calling attention to a hidden cause of a vicious feedback loop and proclaiming it loudly to those who must hear it. Lasting change won’t come from within one discipline. It will come from seeing the whole system and integrating all relevant points of view.

Kelsey Jakee, senior director, portfolio strategy & innovation, TransCelerate BioPharma