Industry estimates put average patient dropout across clinical trials somewhere in the range of 25 to 30%, with some individual studies reporting figures considerably higher. That headline number gets treated as one problem, but it's really describing two very different events with two very different cost profiles: screen failure, which happens before a patient is ever randomized, and mid-study dropout, which happens after baseline data has already been collected and real trial spend has already been committed to that patient. Conflating the two leads sponsors to under-invest in exactly the retention problem that actually costs the most.
A screen failure is disappointing but cheap — the trial loses the time spent on pre-screening and initial assessment, but no real study spend has gone into that individual yet. A mid-study dropout is the opposite. By the time a patient withdraws after baseline, after several visits, the trial has already paid for site time, assessments, possibly investigational product, and data management effort all of which is now a sunk cost with no corresponding data contribution to the final analysis. For a Phase 3 study enrolling hundreds of patients, even a dropout rate in the middle of that 25-30% range translates into a meaningful multiple of the per-patient cost in pure replacement expense, on top of the schedule delay required to re-enroll and catch up the lost participant-years of follow-up.
Most recruitment-focused retention strategy concentrates on the funnel that leads up to randomization — conversion, screen-fail rate, consent completion. That's genuinely important, but it addresses an entirely different risk than mid-study dropout, which depends on what happens to a patient's experience after they're already enrolled. A study can have excellent pre-randomization conversion and still lose a disproportionate share of its budget to mid-study attrition if nobody is specifically tracking and addressing disengagement once a patient is already in the protocol.
Comprehension at the point of consent is one of the stronger known predictors of sustained engagement through the rest of a study — a patient who didn't fully understand what they were agreeing to is more likely to disengage once the practical burden of visits and assessments becomes real. Visit and assessment burden accumulated over the protocol is another: the same instrument fatigue that drives ePRO non-compliance also tends to precede a patient deciding to withdraw altogether, meaning declining ePRO completion rates can function as an early warning signal for dropout risk, not just a standalone compliance metric. Protocol amendments add a further layer of risk specific to mid-study attrition — a substantial share of protocols are amended at least once after activation, and each amendment that changes visit burden or assessment requirements reintroduces a disengagement risk for patients who already consented to a different version of the study.
If mid-study dropout is tracked as part of an undifferentiated overall attrition number, it's invisible until the final analysis shows how many patients were lost. Treating it as its own tracked metric separate from screen-fail rate, tied specifically to post-baseline withdrawal — allows a sponsor to see a developing pattern (a specific site's post-baseline withdrawal rate climbing, a decline in ePRO compliance preceding withdrawal) early enough to intervene, rather than discovering the cost only once a patient is already gone.
Cloudbyz's connected Patient Recruitment, eConsent, and ePRO platform is designed to track post-randomization engagement signals ePRO completion trends, consent comprehension data, visit adherence — as part of the same system used for pre-enrollment funnel metrics, rather than treating mid-study retention as an entirely separate, unmeasured concern. A declining ePRO compliance pattern at a specific site becomes visible as a retention risk signal, not just a data quality statistic reviewed in isolation.
A 25 to 30% dropout rate is often treated as a single number to plan around. Separating out the mid-study portion specifically the costlier, sunk-cost half of that number is what actually lets a sponsor act on it before the cost is already locked in.
See how Cloudbyz's connected platform tracks mid-study retention risk, not just pre-enrollment conversion book a demo.