What Original Means When the Original Is Electronic

Part of Clinical Data Management and Integrity: Record Integrity and ALCOA+.

About 5 minutes. Free, in your language, no account needed.

Why this matters

Originality is simple with paper: the first thing written on is the original, and you keep it.

Electronically, the concept gets genuinely hard, and it is where more confusion exists than anywhere else in record integrity. A value can exist in six systems simultaneously, all identical, none obviously first. Getting this wrong means either retaining everything indefinitely at great cost or, far worse, discarding the thing that turns out to have been the original.

The rule, stated carefully

The original is the first durable recording of the information. Not the first place a human saw it, and not the system of record chosen for convenience.

Working through cases makes it concrete.

A measurement device that stores its readings internally and transmits them. The device's stored reading is original. What the receiving system holds is a copy, faithful or otherwise.

A device that displays a value which a person then types into a form. Nothing durable existed before the typed entry, so the form entry is the original. This is the case people find surprising, and it is why some devices are configured to retain readings specifically to create an original.

A participant completing an electronic questionnaire. The response as captured by that system is original. There is no earlier durable record.

A clinician who examines a participant and dictates a note transcribed later. The transcript is the first durable record unless the dictation audio is retained, in which case the audio is.

A paper worksheet transcribed into a database. The worksheet is original. This is the ordinary case and the one that goes wrong, because worksheets feel like scrap.

Direct entry, and why it is preferred

Where a case report form is completed directly, with no earlier record, the form entry is the source. This eliminates transcription entirely, which removes an entire class of error.

It also carries a specific obligation that teams miss: the electronic system now is the source document, so it must meet every source document expectation, including an audit trail from the first entry and retention arrangements that outlive the study's use of the system.

The trap is treating direct entry as a shortcut. It is a transfer of responsibility from the site's paper file to the sponsor's system, and if that system's retention arrangements end when the study does, the source has been destroyed.

The decision about which fields are direct entry must be made, documented, and made known to the site, before enrollment. A site that does not know a field is direct-entry source will keep its own parallel paper record, and two originals of the same fact is its own problem.

Certified copies

Where an original is replaced by a copy, the copy must be verified as complete and accurate, and the verification itself documented, including who verified it and when.

Scanning a file and shredding the paper is a certified copy process only if the verification happened and is documented. Without that, an original was destroyed and what remains is an unverified image.

In practice

A site uses an electrocardiograph that prints a thermal strip and stores the tracing internally for thirty days.

Current practice: the strip is printed, the clinician marks their interpretation on it, the interpretation is entered into the case report form, and the strip is filed in the participant's folder.

The data manager raises two issues.

The thermal strip will likely be unreadable within a few years, so the original will be blank while the case report form retains only the interpretation and not the tracing it was based on.

The device's internal storage holds the tracing for thirty days and is then overwritten, so the only durable original is the fading strip.

The recommendation is that tracings be exported from the device within thirty days and retained electronically, with the strip photocopied or scanned as a certified copy at the time of the visit while it is still legible.

This costs a few minutes per participant. Not doing it costs the tracings, and the tracings are what substantiate every cardiac safety statement the study will make.

What this looks like when it goes wrong

A study uses direct entry for vital signs. The site is told the form is the source and keeps no paper.

Two years after the study ends, the sponsor's contract with the electronic data capture vendor ends, and the data is exported to the sponsor's archive as datasets plus a separate audit trail export.

The exported dataset holds the values. The site holds nothing, because it was correctly told not to keep a parallel record.

An inspector at the site asks to see the source for a participant's vital signs. The site has no source. The source is a dataset held by the sponsor in another country, and the site cannot produce its own participants' source data at all.

Nobody did anything wrong at any single step. The direct entry decision was sound, the site followed instruction, and the archive is complete. What was never decided was how the site would retain access to its own source records after the system went away.

Key takeaways