You can finish a trial with every planned assessment completed and still be missing an observation your endpoint depends on.
Every activity was carried out as scheduled. The completion record looks exactly as it should, but the schedule never provided a suitable chance to collect the evidence you needed.
That’s the problem with treating completion as proof that a trial’s assessment plan was good enough. Completion tells you whether the work was done, not whether the right work was planned.
Quality by design has to answer that question before execution begins.
A full schedule can still miss the evidence you need
A useful schedule places suitable assessments where they can capture the required endpoint observations.
Start with the evidence. For each essential observation, the schedule must offer an assessment capable of producing it, placed where the result can meet the evidence requirement. Simply adding more assessments doesn’t establish either condition.
In a randomized trial of chemotherapy symptom monitoring, participants reported the severity of 11 symptoms daily through an automated system. One primary endpoint concerned the number of days with severe, moderate, mild, or no symptoms. [1] A schedule limited to clinic visits could have been completed in full without supplying the daily observations needed for that endpoint.
There are two ways a schedule can fall short. It can omit an assessment needed for a required observation or include that assessment at a position where the observation can’t be obtained.
Those are different planning mistakes. They lead to the same result: the team can follow the plan perfectly and still have no suitable opportunity to collect essential evidence. Performing the activities more diligently won’t change what those activities were capable of producing.
Every assessment was completed. the evidence is still missing.
When the trial moves into execution, the schedule determines the opportunities available to collect evidence. Completion confirms the planned assessments occurred, not that they could produce the required endpoint evidence.
A favorable completion result might be entirely accurate. The problem starts when it is treated as proof of something it never measured: whether the schedule provided suitable opportunities to collect the needed observations.
Suppose an essential observation is unavailable even though all scheduled assessments were performed. An omitted assessment leaves no suitable opportunity to collect the required observation; one scheduled at an unsuitable position cannot meet the evidence requirement, even when completed.
In both cases, execution reveals a weakness in the planned opportunity. It doesn’t show that the people carrying out the plan failed to perform the assessments. A clean completion record and missing endpoint evidence can both be true.
The most important schedule decision happens before execution
Before approving the schedule, clinical operations leaders must confirm that each assessment is clearly defined and evaluate whether its planned position allows it to produce the evidence the endpoint needs.
That takes a clear connection between every essential evidence requirement and a suitable assessment opportunity. An assessment with the right name isn’t enough if its scheduled position prevents it from obtaining the observation. And an additional assessment without a relevant evidentiary purpose does not provide the missing opportunity to collect the required observation.
This is the work of an evidence-led quality-by-design review. Judge a schedule by whether it provides a suitable opportunity to collect every essential observation. [2]
There’s a limit to that judgment.
A suitable schedule provides opportunities to obtain the required observations. It can’t guarantee that every assessment will actually produce usable evidence when performed. Missing evidence can have causes other than a flawed schedule.
But if the schedule never provided a suitable opportunity, that particular problem existed before anyone completed a single assessment.
Finding missing evidence doesn’t create the opportunity to collect it
Oversight can establish two facts at once: every scheduled assessment was performed, and an essential endpoint observation is still unavailable. Recording both facts gives an accurate account of what happened.
What it can’t do is replace an assessment opportunity that wasn’t planned. Describing the missing observation more clearly doesn’t make it available. An amended schedule with verified coverage of the required evidence corrects the original scheduling deficiency. Documentation of the shortfall alone does not.
That is the limit of checking execution after the fact.
You can prove that the team followed the schedule and identify the essential observation the schedule left unavailable. Neither finding proves that the original schedule was adequate.
If the evidence matters to the endpoint, the schedule has to give the trial a suitable chance to obtain it before execution starts.
Sources:
[1] Wiley. Automated home monitoring and management of patient-reported symptoms during chemotherapy: results of the symptom care at home RCT. Peer-reviewed randomized controlled trial published in Cancer Medicine, January 30, 2017. PubMed record
[2] International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH). General Considerations for Clinical Studies, E8(R1). Final ICH harmonised guideline, adopted October 6, 2021; editorial correction August 4, 2022. ICH official guideline PDF