Trials Fail Because They Ignore Real Life.

One Idea Worth Acting On.

A KRAS-targeted oncology trial opens enrollment.

The incidence model says the patients are there.

Months pass. Enrollment crawls.

Nobody accounted for the fact that most KRAS-positive patients in the community had already been through multiple lines of therapy — and were excluded by the trial's own prior-treatment eligibility criteria.

The patients existed. The protocol had already ruled them out.

Execution breaks where life meets protocol.

What's Actually Happening

When trials struggle to enroll, organizations reach for familiar explanations — disease is too rare, patients are unaware, sites are underperforming. The protocol itself is rarely questioned.

But the protocol is often where the problem starts.

Eligibility criteria are designed to protect trial integrity and satisfy regulators. They are not designed against the reality of the patient population they will encounter. Prior-treatment exclusions, lab value thresholds, visit frequency requirements — each is individually defensible and collectively devastating to enrollment.

As Terri Conneran, patient advocate and contributor to Voices of Oncology's patient contributor, describes it: a patient with stage-four cancer and two children at home looks at a protocol requiring three clinic visits per week. She does the math — childcare, transportation, missed work, physical exhaustion. She does not enroll. She never calls to explain why. The trial records no screen failure for her. She is simply invisible to the enrollment data.

The trial does not know it lost her. The protocol never asked whether she could participate.

Why This Matters

Enrollment failure is the most expensive problem in oncology drug development — measured in months of delay, millions of additional cost, and patients who never accessed a therapy that might have helped them.

Trials that systematically exclude patients because of operational burden — not scientific necessity — produce datasets that do not reflect the population who will ultimately receive the drug.

The evidence generated in a trial of patients who could navigate the protocol is not the same as evidence about the drug's performance in patients who could not. That gap follows the therapy from approval into real-world practice — where prescribers discover the drug behaves differently in the patients who were always excluded.

Where It Breaks in the Real World

The KRAS enrollment failure was predictable. Disease incidence models count patients by diagnosis. They do not count patients by treatment history, by proximity to a trial site, by ability to meet a visit schedule, or by the eligibility criteria that will exclude them at screening.

Every one of those variables was knowable before the trial opened. None was built into the enrollment forecast.

The same gap appears in visit schedule design. A protocol that requires weekly in-person visits for routine blood monitoring is survivable for a patient at a major academic cancer center with a full support system. It is not survivable for a patient managing advanced cancer, a job, and two children in a city without reliable transportation to a trial site.

Both patients are eligible. Only one can participate. The trial does not distinguish between them.

What Needs to Change

Operational feasibility belongs in the protocol design conversation — not as an afterthought but as a primary constraint alongside scientific rigor and regulatory requirements.

That means enrollment forecasts that account for treatment history exclusions, not just disease incidence. Visit schedules reviewed against the daily reality of seriously ill patients, not just clinical monitoring requirements. Decentralized options — local labs, remote monitoring, home-based assessments — built in from the start for patients who cannot make repeated trips to an academic center.

The future of oncology research belongs to organizations that can design trials patients can actually complete. Not trials that are scientifically perfect and operationally impossible.

The Bottom Line

Trial design must fit the patient's life. Not force the patient to fit the trial.

Until operational feasibility becomes a core design principle — with the same standing as endpoint selection and statistical power — execution will keep breaking exactly where life meets protocol.

Add Your Voice — Where are you seeing this? What's working — and what's not?