Enrollment is the milestone everyone celebrates, and it's the moment the real risk begins. Patients who consented with genuine intent still leave, and they leave for reasons you can usually see coming, often in data you already hold. There's no single reliable industry average for dropout; attrition swings widely by therapeutic area, and one systematic review found half of trials were missing primary-outcome data for more than 11% of their participants. Retention isn't a matter of luck. It's a matter of noticing the fade before it becomes a dropout.
Designing Support Around Real Behaviors
Execution pattern friction is a history of inconsistent follow-through: missed appointments, refill gaps, abandoned treatments, short-term focus. It's among the strongest predictors of retention there is, for a simple reason. Past behavior predicts future behavior, and a patient's adherence and attendance history is the closest thing to a preview of how they'll handle a demanding protocol.
PRISM reads it across six models: treatment-adherence history (Proportion of Days Covered and Medication Possession Ratio from pharmacy claims), appointment-attendance history, planning and follow-through, persistence and resilience, short-term focus, and coping style. The response is built around the pattern rather than against the patient: high-touch reminders tuned to the individual, a missed-visit recovery flow, behavioral nudges aligned to short-term focus, and early-warning detection of disengagement. The point is to add support where the data says it's needed, not to exclude the patients who need it most.
Fitting the Trial to the Patient's Life
Schedule burden friction is when a patient can't sustain the protocol's visit cadence, overall duration, or dosing schedule given their life. A cadence that looks reasonable on the protocol can be impossible to keep up for a year, and the problem is getting worse: Tufts Center for the Study of Drug Development benchmarks show protocols growing more complex over the past decade, with the share of protocols requiring at least one amendment rising from 57% to 76%. The mismatch surfaces as attrition well into the study, when it's most expensive to lose someone.
PRISM reads it across visit-frequency tolerance, trial-duration tolerance, and pill-burden and dosing-schedule tolerance, comparing the protocol against the patient's attendance history, life schedule, and existing regimen. The response bends the schedule where the protocol permits: visit windowing, telemedicine for routine check-ins, simplified dosing reminders, and drop-off forecasting to pre-empt fatigue before it becomes a withdrawal.
They're afraid of a specific procedure
Protocol design fear friction is when a patient is afraid of specific protocol elements: placebo assignment, the washout period, biopsies, lumbar punctures, infusions, or overnight stays. Any one of them can end participation, and it often doesn't come up until the procedure is scheduled, when the patient simply declines or disappears. Placebo aversion is strongest exactly where you least want to lose people: in sicker patients who've already failed other treatments.
PRISM reads it across five models spanning placebo concern, washout concern, procedure tolerance, sample-collection tolerance, and hospitalization tolerance. The response is honesty plus choice: procedure-specific videos and explanations, placebo explained without minimizing it, a washout-bridge plan built with the patient's current care team, and choice architecture around procedures where the protocol allows.
They're worn out before they start
Treatment fatigue friction is when a patient is worn out by prior treatment, so a new regimen feels too hard even when they understand and trust the trial. For someone who's been through line after line of therapy, starting again is a genuine hurdle, and assuming fresh motivation misreads where they actually are.
PRISM captures it through the treatment fatigue model, reading prior treatment lines, years on therapy, recent changes, and self-reported exhaustion. The response meets the fatigue honestly: recognition and validation language, a simplified onboarding, a fresh-start framing, and paced check-ins that monitor energy rather than assume it.
Dropout is rarely random
The through-line across all four is that mid-study attrition is predictable more often than the industry treats it as. A patient's follow-through history, the fit between the schedule and their life, their fear of a procedure, their fatigue: each is knowable at or before enrollment. Scoring them turns retention from a post-hoc autopsy into something you manage in advance, while support still changes the outcome.
Sources and methodology.
This article synthesizes published research and public benchmarks with Jumo Health's PRISM readiness framework. Statistics are presented with their publication context; trial conditions and patient populations vary.
- Tufts Center for the Study of Drug Development, site enrollment performance benchmarks. View source
- CISCRP, 2023 Perceptions and Insights Study. View source
- Communications Medicine (Nature), 2025, trial representativeness of FDA-approved drugs. View source
- National Assessment of Adult Literacy (NAAL), US Dept. of Education / NCES. View source
- Paasche-Orlow et al., New England Journal of Medicine, 2003, consent form readability. View source
- Johns Hopkins Medicine IRB, informed consent readability guidance. View source
- Journal of Clinical Oncology (ASCO), 2022, travel distance in early-phase trials. View source
- Geographic access to NCI-funded cancer research sites (PMC). View source
- Patient-reported out-of-pocket costs in early-phase oncology trials (PMC). View source
- Pew Research Center, Internet/Broadband Fact Sheet. View source
- FCC, Broadband Progress Report (100/20 Mbps standard). View source
- Kogan et al., Psycho-Oncology, 2022, caregiver role in phase 1 trial decisions. View source
- Cerutti et al., Cancer Medicine, 2025, family system and trial retention. View source
- Tufts Center for the Study of Drug Development, protocol amendment benchmarks. View source
- Trials (Springer), 2025, retention and missing primary-outcome data review. View source
- US FDA, Diversity Action Plans (FDORA 2022). View source
