Perfect Clinical Trial Source Documents

As a CRA, I regularly visit the sites to review the study conduct, encourage study participant recruitment, evaluate the ongoing suitability of the investigative site, provide protocol specific training, track investigative supplies, and to review source documents.  Source documents are the roadmap and clinical record to let me (or anyone who is reviewing) know what happened, when, how, and by whom.  When I look through source documents I am checking that they follow the principle of ALCOA:

• A-attributable, who wrote this?
• L-legible, can I read this?
• C-contemporaneous, was this recorded at the time of trial conduct or later?
• O-original, is this unaltered or copied?
• A-accurate, is this a correct reflection of the conduct?

The burden for documentation in a clinical trial is actually much higher than the burden for documentation in a standard clinical record or office/physician visit.  When you see your doctor for a flu he might write 3-4 lines in your chart documenting the visit.  If you were a clinical trial participant, the doctor or the clinical trial staff would likely need to collect and record a lot of specific information to satisfy the data collection requirements of the trial.

It is never acceptable to use
correction fluid in source documents.

The first time the pen touches the paper to record observations in a trial, this generates “source documents”.  If the nurse writes your blood pressure on her hand, her hand is a source document. If you stand on the scale and your weight is recorded on a post-it, that is a source document.  One of the most important rules of GCP is that the Principal Investigator (PI) is ultimately responsible to ensure that all source documents are maintained in their original form and that any alterations or corrections are indicated in a compliant fashion (changes should be striken through with a single line (so as not to obscure the orignal entry); correction fluid should never be used! The revision should be initialed and dated so there is a record of who made the change and when. An explanation of the change is optional but best practice).  The PI should instruct his staff not to write on their hands or sticky notes, to make observations using indelible ink (not pencil since that can be erased!) and to always observe the ALCOA principle when generating or modifying source documents.

Some sponsors actually provide source documents to the investigators to assist them to record all the required data during the study visits.  The source templates will typically have a blank header to record the subject identifying information and to indicate the date of the visit. Quality source will also include signature or initial blocks on each page so that the person(s) completing the forms can be identified later. If source documents are complete and accurate, the clinical trial data can be copied over and provided to the sponsor on Case Report Forms for analysis. I’ll provide an introduction to data capture/EDC and analysis in a future blog topic.

During source document review I always find items that are incomplete (left blank), inaccurate (such as blood pressures that are inconsistent with life, headers that indicate a different patient entirely, birth-dates that don’t match the age, units or counts that aren’t possible), or otherwise erroneous source.  I use small colored post-it notes (stickies) to flag my findings and ask the PI or the study nurse to review and update the source with notation.  I never “fix” the source myself.  If the source errors are very wide in scope, I will note it in my monitoring report and I may also ask the investigator to generate a signed and dated Note to File on the site letterhead explaining what happened with the source documents and the corrections completed during our visit.  Experienced study coordinators and investigators will not be surprised or frustrated when I request source corrections or clarifications.  Also, discovering errors in the source does not upset me at all, this is part of my job and I am partnering with the site to help them keep their documentation totally complete and an accurate reflection of what was done.

Source documents are created during the patient visit and should not be fabricated a few days before a monitoring visit.

Do I want to see perfect source at my visits?  Actually, absolutely not.  Original source documents will be working documents and I am expecting source documents to be a bit tattered, thumbed through with bent corners, covered in the occasional stain, marked with revisions, and otherwise imperfect.  Perfect source or typed source makes me especially suspicious and can be an indicator of fraudulent or sloppy record keeping.  I hope this discussion is helpful to you and that you will know what to look for when reviewing source documents at your monitoring visits.