28th Nov 2011, 11:18 AM
Source Data and CRFs
I have a question about source data and CRFs. I wasn't able to create a thread in the data management section. I work in the R&D department for an NHS trust and we have started sponsoring clinical trials. A few investigators of single-centred trials have come to us with the idea of using the paper CRF as the source data i.e. the first place they write down the information. The consent process and visit schedule would be entered into the subject?s medical notes but the details of the trial specific tests would only be entered on the CRF. The original CRF would be left in the subject?s medical notes at the end of the subject?s participation in the trial and a photocopy taken for the researchers use.
Is this acceptable? I would be really grateful for any advice.
30th Nov 2011, 04:19 PM
I am under the impression that as long as it has been clearly stated before the trial start where the source data is going to be recorded then this would be acceptable. I would be interested to see what other think. olivia
7th Dec 2011, 04:10 PM
We commonly use the CRF as source data for academic/non-commercial studies. This is acceptable by Ethics and R&D and we have been audited by R&D, external auditors and the MHRA.
19th Dec 2011, 11:01 AM
BARQA GCP Committee published a Q & A about this in Dec 2002:- ICH GCP 6.4.9 states that the protocol (or protocol referenced documents) should contain:- ?the identification of any data to be recorded directly on the CRFs (i.e. no prior written or electronic record of data), and to be considered to be source data.? This means that ICH GCP anticipates direct entry into the CRF.
The GCP Committee pointed out that it is important to cover all these aspects in the Pre-Trial and the Trial Initiation Visits. The investigator centre should identify what is the usual procedure for recording data and which data would have no source except for the CRF. This should be agreed with the sponsor before the study is initiated at this site.
The committee were not in favour of study specific forms that are copies of the CRF. This merely duplicates information and may lead to confusion as to which really was the source.
It was also a concern that there might be a danger of some ?back transcription? of data from the CRF to quasi source documents.
Although some data may be directly entered onto the CRF, there will always be a need for some independent source documents. It is important to be able to provide evidence of the existence of the subject; evidence of the subjects identity (confidential documents held by the investigator); medical history (particularly where the protocol may require certain historical information as an eligibility criteria); and medical details that will allow the study staff and other medical staff (including emergency staff) to provide for the good care of the subject both during and after the study (e.g. relevant adverse events, concomitant medications and other details).
19th Jan 2012, 09:00 PM
I see no problem with this.
We use an eCRF system and we create 'source data worksheets' that form part of the patients medical records ultimately. That way, the medical notes are source and contain all research data, it is then transcribed to the eCRF system online and no photocopying is needed. At the end of the trial, only the eCRF needs to be retained as research records and the source data worksheets can remain in the medical records at site - cutting down on the hassle of archiving them locally.
Seems to work well - we've been doing this since 2005.
2nd Feb 2012, 05:10 PM
Thank you for your responses.
I have had confirmation from the MHRA that it is acceptable to use the CRF as source for certain types of data as long as it is specified in the protocol.
Care must be taken by the investigator to ensure that information important for patient management is transferred from the worksheets/CRFs to the subject case notes, for example, adverse events, however in this case it must be clear which document is the source. Such information is vital for patient management eg when patients are managed outside the trial team by different medical specialities.
If your current protocol does not specify that the CRF will be used as the source (as was the case with our trial) then this should be updated. Since this is normal practice, a substantial amendment would not be required.