A decentralized clinical trial activity reaches review with its completion recorded. The person responsible for verification can see that the activity occurred, but the activity itself is no longer available for direct observation. The review must proceed from the evidence preserved during execution, and that evidence may not show enough about performance to support acceptance. What appears to be a completed workflow can therefore arrive at verification with the central execution question still unresolved.
Verification begins after execution has left view
When an activity is performed without direct observation, execution is no longer directly observable at the point of verification. The verifier receives evidence that represents what happened rather than witnessing the activity itself. Documentation and system records can carry information from the point of execution into review. Downstream reconciliation may provide another indirect view, though its value still depends on what details about performance were preserved.
This changes the practical basis of verification. The verifier can only assess the elements of execution that remain visible through the evidence available after the activity. Information that was never captured cannot be recovered simply because a record exists. The handoff from performed activity to recorded evidence becomes part of the verification mechanism because the information preserved in that handoff determines what the later review can establish.
The distinction matters because evidence may be complete within its own design while still being insufficient for the verification decision. A record can fully show the data it was built to capture and still omit an aspect of execution that direct observation would have made visible. Record completeness and execution assurance are separate questions.
That visibility problem was addressed through contemporaneous remote oversight when the Nourish trial moved blood-pressure measurement into participants’ homes. Staff used live video to guide cuff placement and assess visible aspects of body position. After confirming premeasurement conditions verbally, staff timed the five-minute rest period. They then repeated readings before recording the values. The protocol therefore combined contemporaneous oversight of parts of the measurement process with recorded BP values. [1]
Completion can be clearer than performance
The most immediate friction appears when evidence confirms that an activity occurred but does not sufficiently establish how it was carried out. A completed entry can close one factual question, but the verification question remains unresolved because the manner of performance is still uncertain.
That uncertainty can produce two different operational outcomes. One is false assurance, where evidence of occurrence is treated as evidence that execution met the required standard. The other is unresolved verification, where the activity may have been performed correctly but the available evidence cannot demonstrate that fact. In both cases, the review process reaches its limit because the evidence does not carry enough information from execution.
False assurance produces confidence that exceeds what the evidence can support; unresolved verification reflects a different problem in which confidence cannot be established even when underlying execution may have been sound. The same decentralized activity can create either excessive certainty or persistent uncertainty, depending on what the evidence preserves.
This is where ownership of the verification decision becomes difficult. Accountability for verification can remain with a person who did not perform or witness the activity. That person is still expected to reach a judgment, but the judgment is bounded by evidence generated elsewhere. Responsibility remains intact even without direct visibility. [2]
Detection can change without execution changing
A further oversight problem appears when the quality of indirect evidence changes over time. Better evidence may expose more deviations or unresolved conditions even when the underlying manner of performing the activity has not changed, so a rise in detected issues may reflect stronger visibility rather than weaker execution. [3]
Observable findings can also show the reverse pattern. Sparse evidence can make activity performance look more stable because fewer problems are visible to the verifier, but that apparent stability does not establish stronger execution. It may only indicate that less of what occurred was visible to the verification process.
This distinction affects how leaders read operational signals. A change in downstream findings cannot, by itself, establish a change in execution quality. That signal reflects what verification detected, and its meaning depends on whether the evidence supporting detection also changed. Without that separation, a stronger evidentiary record can be mistaken for deteriorating performance, while a weaker record can conceal uncertainty.
Counting records or findings does not resolve this distinction. Leadership needs to know what the evidence can establish about execution at the point where the verification decision is made. More records can increase visibility without increasing assurance if they repeat the same limited view of performance.
Assurance is bounded by evidentiary reach
The governing constraint is the amount of execution detail preserved between performance and verification. Direct observation would make aspects of performance available during the activity itself, but without it those aspects must remain observable through indirect evidence if they are to support later verification.
This limits what oversight can establish. Verification cannot establish more about execution than the evidence makes visible. The presence of documentation does not change that limit, and additional documentation does not necessarily expand what verification can establish. Assurance increases only when the evidence preserves information that bears on the required manner of performance.
For leadership, the unresolved decision is specific. At the point of acceptance or further review, does the available evidence support a conclusion about how the activity was performed, or does it support only the conclusion that the activity occurred? If that distinction cannot be made, the review process cannot determine whether uncertainty reflects underlying execution or the limits of what was captured about it.
Sources:
[1] Wolters Kluwer. Implementation of an At-home Blood Pressure Measurement Protocol in a Hypertension Management Clinical Trial During the COVID-19 Pandemic. Journal article, published online 23 April 2022. Full-text journal article
[2] International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH). ICH E6(R3) Guideline – Annex 2. ICH Harmonised Guideline, Step 4, adopted 3 June 2026. Official ICH Annex 2 guideline
[3] SAGE Publications / Clinical Trials. Clinical trial monitoring effectiveness: Remote risk-based monitoring versus on-site monitoring with 100% source data verification. Peer-reviewed research article, first published online 1 December 2020; 18(2):158–167, 2021. PubMed record