Source document and CRF updates: ensuring data collection reflects the current protocol version
Teaches RCs to coordinate data collection instrument updates across site-controlled source documents, sponsor-managed electronic systems, and hybrid local tools -- establishing the timing requirements per ICH E6(R3) Section 4.3.5, the verification step that confirms alignment before the first participant is seen under the amended protocol, and the data integrity risks of misalignment per Section 2.12.1.
The two-hour blood draw that should have been four
An amendment to a pharmacokinetic study changes the timing of a post-dose blood draw from two hours to four hours. The change is straightforward -- the sponsor determines that the four-hour sample provides more clinically meaningful concentration data. The amendment is submitted to the IRB, approved, and filed in the regulatory binder. The staff are trained. The sponsor updates the electronic case report form to reflect the four-hour collection window.
But nobody updates the site's local source document worksheet -- the paper form the research nurse uses at the bedside to record the time of each blood draw, the tube type, and the processing instructions. The worksheet still says "2 hours post-dose." The next participant arrives for the amended visit. The nurse, working from the source document in front of her, draws the blood at two hours. The data is entered into the sponsor's electronic CRF, which now expects a four-hour sample. The timestamp does not match the protocol window. The sample is collected outside the protocol-specified timeframe.
And here is the part that makes this scenario particularly damaging: the data cannot be uncollected. The participant's blood was drawn at the wrong time. No amount of retroactive documentation corrects the biological fact that the pharmacokinetic concentration measured at two hours is not the concentration at four hours. The protocol deviation is irremediable. The data point is, for the purposes of the primary analysis, unusable.
This is not a training failure -- the nurse was trained on the amendment. It is not a regulatory failure -- the IRB approved the change. It is a data collection instrument alignment failure. The protocol said four hours. The CRF said four hours. The source document said two hours. And the nurse, quite reasonably, followed the document in front of her.
I have seen variations of this scenario at every type of research site -- academic medical centers, dedicated research organizations, community practices. The root cause is always the same: the site treated data collection instrument updates as something that happens automatically when the amendment is approved, rather than as a discrete coordination task that requires its own workflow and its own verification step. This lesson teaches that workflow.
What you will learn
By the end of this lesson, you will be able to: