External control · high severity
Confounding is addressed by a pre-specified method, with diagnostics
An external control differs from the trial arm in ways nobody measured and some ways nobody recorded. The method and the diagnostics have to be chosen before the comparison, or the adjustment follows the answer.
The check
Does the outline name the confounders, the adjustment method and the diagnostics that will show balance?
What the outline has to carry
- confounding
- a pre-specified method
- diagnostics
A check is met only when every element above is present. One element on its own is a partly evidenced check, not a pass.
What resolves it
List the confounders and how each was measured, name the primary adjustment method, state the balance diagnostics and the threshold for concern, and pre-specify at least one analysis for unmeasured confounding.
The source sentence
Make sure the analysis methodology appropriately addresses sources of bias like confounding.FDA Watch, 32:20 to 33:15
The quotation is short and is used to show the sentence the rule is read from. The rule itself is our statement, not the speaker's words.
The legal basis
The held frameworks that bear on this check, with the control each one contributes. The link opens the framework page so the reader can check the source.
Data management and external data
A data management plan covers external data, including real-world data, its quality, its traceability and its coordination with the analysis.
International · Read the framework
Try it on your outline. The free check applies this rule and quotes the line it read.