Before Burden Becomes Risk: Capturing Friction in Real Time

Clinical trials rarely struggle because of a single point of failure or protocol flaw. More often, risk accumulates through human moments: a confusing workflow, an uncertain answer, or a workaround that exists because the “official” process does not fit the clinic. These moments are the human side of clinical trials.

As a field, we view clinical research through the lens of protocols, platforms, and endpoints. But trials are carried out by people who interpret instructions, solve problems, and sustain attention over months or years. That human layer is where operational risk first appears. Long before a missed milestone shows up in a dashboard, someone may feel uncertain, underprepared, or skeptical that the process will work.

The challenge is that the signals we traditionally measure appear late, then we ask teams what happened, often weeks or months after the important moments occurred. By then, the story has changed.

Behavioral science helps explain why. Daniel Kahneman distinguished between the experiencing self and the remembering self. The experiencing self lives moment by moment. The remembering self organizes those moments into a story and decides how things went. Both matter, but they do not report the same thing.

Ask a site coordinator at the end of the month, “How is the study going?” and you may get a reasonable summary: “It is complicated, but we are managing.” Ask that same coordinator at the end of a difficult Tuesday, after three screening calls, a lab kit problem, an unclear eligibility question and you may hear something far more useful: “The process looks simple on paper, but every patient requires a custom solution.” That difference matters because clinical trial burden does not live in summary judgments. It lives in moments.

Behavioral science also provides us a solution. Experience sampling is a technique engineered to capture those moments. Experience sampling delivers brief, repeated questions while people are doing the work, or shortly after. Rather than relying only on retrospective surveys or lagging assessments, experience sampling collects signals from real-world context; it does not ask, “What do you remember about the burden?” It asks, “Where did burden show up today?”

For clinical research, this is powerful because readiness is not only a property of the site. It is more principally a property of the individual. A site may be activated, trained, and believed to be ready, while individual staff vary in confidence, motivation, clarity, perceived control, and readiness to change routines.

Over the past three years we have written about how initial start-up and training can assess whether individuals are ready to adopt trial behaviors. And we have demonstrated how readiness to change can serve as the earliest predictor of whether training will translate into action. Experience sampling provides the second signal: demonstrating how readiness holds up once the work begins. It becomes a secondary real-time risk indicator, complementing predictive readiness measures by showing how individual readiness is tested by real-world conditions: confidence, clarity about where to get help, intended workflow fidelity, and emerging workarounds.

First patient screened for your earliest sites is an obvious place to see this in action. A five-day experience sampling sprint could reveal ambiguous instructions, training gaps, technology friction, role confusion, and hidden assumptions before downstream sites inherit the same problems. Each day, site staff might answer: What created friction today? Where did you feel least confident? What should the next site know?

But the opportunity is far greater. In many trials, real complexities emerge over months or years. Recruitment strategies fatigue. Staff change roles. Amendments introduce new procedures. Minor inefficiencies compound. A study that felt manageable in month one may feel very different in month nine. Experience sampling can move from an early stage diagnostic to an ongoing learning system: brief monthly or milestone-based pulses during active enrollment, after amendments, first patient visits, data cleaning, or new staff onboarding. Not long surveys. Just a few well-timed questions designed to detect friction while it is actionable.

The goal is not to monitor people. The goal is to learn from and with them. If experience sampling feels like surveillance, it will fail. But if framed as a way to reduce friction, improve support, and protect performance, it can become both a measurement technique and an intervention.

Critically, experience sampling does not only create value for the sponsor, CRO, or study team trying to detect risk. It can also create direct value for the individuals completing it. A brief pause at the end of the day asks people to notice what actually happened, not just what they intended to do. That reflection can surface personal “say-do” gaps: the difference between the workflow someone planned to follow and the workaround they actually used, the support they thought they understood and the question they still avoided, or the behavior they endorsed during training and the habit that took over under pressure. In that sense, experience sampling becomes more than measurement. It becomes a small learning intervention, helping people see their own work more clearly while there is still time to adjust.

Site burden is not a static property of the protocol. It is experienced by people, in context, over time. Experience sampling will not replace feasibility assessments, readiness measures, monitoring data, or lessons learned. But it can complement them by capturing the immediate, situated, human experience of trial execution.

The remembering self can tell us the story of the trial. The experiencing self can help us improve it while there is time.

Brian S. McGowan, PhD, FACEHP, is chief learning officer and co-founder of ArcheMedX, Inc.; Kelly Ritch is chief operating officer of ArcheMedX, Inc.; and Jason Riis, PhD, is CEO and chief behavioral scientist, Behavioralize