Chapter 26 / Rescue, Crisis, Inspection, and Leadership Judgment / Paid beta preview

Chapter 26: When Trials End Early or Fail

How to end, pause, or learn from trials without abandoning participants, sites, data, or truth. ## Ending Early Is Still Trial Management RESP-640 did not end in a dramatic announcement. It ended through a series of disciplined meetings that felt heavier because everyone finally understood the facts. The enrollment hold had protected new participants while safety follow-up continued. Dose escalation remained paused. The amendment path could reduce some burden, but it could not erase earlier protocol drift. The data might still teach something, but not the clean answer the original trial had promised. The budget had become difficult to justify. Site morale was fragile. The sponsor could continue only by accepting residual safety, feasibility, data, regulatory, and credibility risks that several accountable functions no longer believed were justified. Dr. Daniel Liang said the thing the room had been circling. "Ending a trial early is not automatically failure," he said. "But pretending a trial can still answer a question it can no longer answer is failure." An early termination is a decision to end a study before the planned completion point. A study pause is a temporary stop in some or all trial activity. An enrollment hold stops new participants from entering while other activities may continue. A treatment discontinuation means a participant stops study treatment. Study discontinuation means a participant stops study participation. Participant withdrawal means the participant chooses to leave some or all study activities, according to the consent and protocol. These are not the same. RESP-640's decision was early termination for combined safety, feasibility, and credibility reasons. That did not mean every person on the team had failed. It meant the sponsor had to close the study truthfully, protect participants, preserve evidence, meet obligations, and learn. ## The CLOSE Frame Maggie Chen gave Lauren the frame for ending well: | Letter | Meaning | PM Question | |---|---|---| | C | Care For Participants | What safety, follow-up, transition, and communication duties remain? | | L | Lock The Facts | What data, records, decisions, and rationale must be preserved? | | O | Obligations | What regulatory, ethics, site, vendor, contract, and reporting duties continue? | | S | Story Without Spin | How do we communicate honestly without blame or false certainty? | | E | Extract Learning | What should the organization learn before moving on? | The frame reminded Lauren that stopping did not mean the project disappeared. In some ways, stopping created a new project: closeout. Closeout is the controlled work of ending trial activities, completing required participant, site, vendor, data, safety, documentation, financial, and regulatory tasks, and preserving the evidence trail. ## Different Kinds Of Failure Clinical teams use the word "failure" too casually. | Situation | Plain Meaning | Beginner Caution | |---|---|---| | Negative trial | Trial does not show the hoped-for treatment effect | Not automatically bad conduct | | Primary endpoint not met | Main statistical objective was not met | May still provide important safety or scientific information | | Inconclusive trial | Evidence is not clear enough to support a reliable conclusion | Often tied to sample size, missing data, variability, or early stop | | Futility | Evidence suggests the trial is unlikely to achieve its objective if continued | Should be preplanned or governed, not just discouragement | | Safety stop | Trial or treatment is stopped because participant risk may no longer be acceptable | Does not require perfect proof before protective action | | Operational infeasibility | Trial cannot be conducted as required with available sites, participants, systems, or resources | Can reflect design, assumptions, execution, or environment | | Business or strategic stop | Sponsor stops for portfolio, funding, commercial, or strategic reasons | Still must honor participant, data, regulatory, and ethics duties | | Failed execution | Trial conduct was not adequate to answer the question reliably | Requires honest accountability and learning | Some trials use a data monitoring committee, or DMC, also called an independent data monitoring committee, or IDMC, to review accumulating data and make recommendations according to a charter. Not every trial has one. DMC/IDMC recommendations, stopping rules, and unblinded review boundaries are not PM decisions. CARDIA-301, an earlier cardiovascular outcomes case Daniel mentioned, completed follow-up and missed its primary endpoint. That was painful, but not the same as a badly run trial. It answered an important question. RESP-640 was different: it ended before the original objective could be answered cleanly. Both demanded truth. ## Care For Participants Participant protection does not end when enrollment stops. RESP-640 had participants in different states: some were screening, some enrolled but not yet dose-escalated, some on treatment, some in safety follow-up, some discontinued from treatment but still in study follow-up, and some hard to reach. Dr. Samuel Reeves and Nina Patel led the participant plan. The team used the safety governance pathway to turn the details into operating choices the PM could assign, monitor, and escalate: Screening participants. Closeout Duty was Stop screening as directed, communicate respectfully, avoid misleading expectations. Enrolled not yet treated. Closeout Duty was Explain status,

...