Why Site Engagement Isn't Just A Role Anymore
By Dan Schell, Chief Editor, Clinical Leader
I went into my conversation with GSK’s Laura O’Donnell thinking we’d talk about site engagement as a function. A team. Maybe even a job description. About 10 minutes in, that idea was gone.
What she described instead was something much harder to define, and honestly, much harder to execute. Site engagement, in her view, isn’t owned by a person or even a department. It’s a capability that has to exist across an entire organization, whether that organization is ready for it or not.
I really had to pause and think about every part of that statement.
So, it’s not a role someone has, and it’s more intrinsic or foundational to the entire company. Got it. I really thought that statement was pretty interesting … no, that sounds weak. To me, it was … illuminating (Ha! See, I can write without a thesaurus!). Basically, it made me think; it gave me pause.
Now, sure, you could say, “Duh Dan, anyone with the title Sr. Director, Head, Global Site Engagement is going to say something like this.” Yeah, maybe. But, I also heard something similar a few months ago at SCOPE from Wouter Daniels, clinical trial patient and site engagement lead at Bohringer Ingelheim (Yes, I know, another site-centric title). He told me sponsors can’t really claim to be patient-centric if they’re not site-centric first. Sites, as he put it, are the bridge. If that bridge is shaky, everything downstream suffers.
A bridge. I like that metaphor. And what O’Donnell is focused on now is what it actually takes to reinforce that bridge in a system that has become far more complicated than it used to be.
The CRA Model Isn’t Coming Back
There was a time when site relationships were relatively simple. A CRA was the face of the sponsor, often for years at a time. They knew the site, the staff, and sometimes even the investigator’s vacation schedule better than the investigator did. Those days are over, though, thanks to tons of turnover in the CRA position, more complex trials (which may be part of the cause for that turnover), and more roles getting involved with the site relationships. The institutional knowledge that once lived in a single person’s head now gets scattered across teams, systems, and sometimes lost entirely.
O’Donnell doesn’t see a path back to that older model. Instead, she’s focused on replacing what was lost with something more durable, even if it looks less personal on the surface.
“We do exist as a network of individuals,” she said. “It takes a village to do clinical trials.
The challenge is making that village feel like a single, coordinated partner instead of a rotating cast of characters.”
More Data Isn’t The Solution
According to O’Donnell, if sponsors can’t reduce trial complexity, they need to figure out ways to absorb it. “More specifically, you have to keep that burden from falling squarely on the site. That’s where much of the current innovation is focused, not on eliminating steps, but on making those steps easier to navigate. Better tooling, clearer processes, faster answers. It’s less about reinvention and more about making complexity survivable.”
A big part of that effort revolves around data. Sponsors are investing heavily in systems that centralize site performance, engagement history, and operational metrics. The goal is to create a shared view of each site so that anyone interacting with them can pick up the thread without missing a beat. In practice, though, it’s not quite that clean.
“Even with all of the AI advancements, there’s still a limit to what data can tell you,” she explained. “Sure, that data can tell you what a site did in the past in a very specific environment. But, it will not necessarily tell you what that same site will do on a different study today.”
Staffing changes, competing studies, shifting priorities — all of it matters. Which means that data alone isn’t enough. You still need context, and more importantly, you still need judgment.
Just think about the feasibility process. Sites are asked how many patients they can enroll. They provide a number, often based on incomplete information. That number gets adjusted, reinterpreted, and sometimes outright changed as it moves through different systems. By the time the study starts, there may be several “versions” of that commitment floating around. Then everyone wonders why projections don’t hold up.
“What would it look like if we were all super transparent?” O’Donnell asked. “If a site said, ‘We can do eight,’ and they did eight every time, they’d be the best site in the universe. But they won’t put eight, because they’re afraid they won’t get selected.” It’s a system that rewards optimism over accuracy and then penalizes sites when reality catches up.
Service Is Becoming The Differentiator
One of the more interesting shifts in the conversation was how often “service” came up alongside “engagement.” Not everything needs to be a relationship. Sometimes sites just need an answer, quickly and reliably. That’s where newer tools are starting to make a difference, whether it’s AI surfacing protocol details instantly or workflows that escalate unanswered questions automatically.
Part of site engagement, O’Donnell noted, “is just getting sites what they need.”
That may not sound revolutionary, but in an environment where sites are juggling multiple studies, sponsors, and systems, responsiveness becomes a competitive advantage.
The Human Piece Still Matters
For all the talk of data, AI, and automation, O’Donnell kept coming back to one point that hasn’t changed: There is still no substitute for being in the room.
That doesn’t mean every site wants the same level of interaction. Some prefer regular visits. Others would rather get a text and move on with their day. The problem is that sponsors often default to what’s easiest for them, not what works best for the site.
“We email, email, email,” she said, “and when that doesn’t work, we call. That’s not their preference.”
It’s a small example, but it reflects a larger issue. Even as the industry invests in more sophisticated tools and processes, it still struggles with basic alignment on how to communicate. Underneath all of this is a bigger shift in mindset. “In my opinion, the sites are customers,” O’Donnell said (again, maybe not surprising considering her job title). “The product is the experience of doing a study with a sponsor.”
That’s not how the industry has traditionally operated, but it’s where it’s heading. Sites have options. They choose which trials to prioritize. They decide how much effort to invest.
And increasingly, those decisions are influenced by how easy, or difficult, a sponsor makes their job.
The old model relied on relationships to carry that weight. The new model is trying to support those relationships with systems, data, and service, while somehow holding onto the human connection that made those relationships work in the first place. That’s a tough balance to strike. And right now, most of the industry is still figuring out how.