Electronic Clinical Trials and the Expectations of the Regulatory Agencies

November 21, 2012
Chris Cramer, Vice President, Clinical Data Management, Quanticate

Applied Clinical Trials

In the days of paper Case Report Forms (CRFs), the accurate recording and review of subject data entered into the CRF and diaries was based entirely on trust. The investigator would sign the final page of the CRF to confirm that he or she had reviewed all data contained therein and that it represented a true and accurate record of the state of the subject during the trial. The time and date of the signature could not, of course, be verified. If a dose escalation or change in procedure was expected, as a result of certain parameters or subject-reported outcomes, there was no way of knowing whether such a review had occurred. Neither was it possible to relate the review and signature(s) to the timing of these assessments, or the subject reporting the outcome. The advent of electronic data capture of both the assessments made at site visits, and reported independently (between visits) by the subject, has opened the door to checking the compliance of both the subject and investigator, as all such records now contain a data and time stamp.

This raises the question as to whether we should now be ’policing’ the site investigator just because we can now monitor activity electronically, or whether we should continue to rely on trust and that the medical professionals engaged in trials are working according to protocol?

When approaching database lock for a study where data has been captured electronically, data managers check that all investigator signatures have been captured within the electronic CRF (eCRF), to signify, just as with a paper CRF, that the data recorded for each subject at their site is complete and accurate, and they have reviewed it. Database closure is a busy period and much effort is usually spent in chasing recalcitrant sites to capture the investigator’s electronic signature. As most data in an eCRF is typically entered by other study site staff, this may be the only form an investigator is obliged to complete. To truly confirm that the data entered into an eCRF is a ‘true and accurate’ representation of the subject’s progress within the study, the investigator would be required to open each eCRF page (an extremely time consuming procedure) or, as a minimum, spend some time reviewing associated subject profile style summary reports. The frequent requests received for password resets around lock time tend to suggest that this is unlikely to be the case.

Recent inspections of electronically captured data have begun to probe this area with questions along the lines ”Can you confirm that the investigator reviewed this particular subject’s electronic diary at the required intervals during the study?” This is a possibility by viewing the audit trail on most systems, however this then raises the question “Who checks that the investigator logs into the system during the trial to perform these reviews?”

Raising questions in this way as to whether the electronic vendor (be it of an eCRF or electronic patient reported outcome (ePRO) system) can demonstrate conclusively that the investigator has been actively reviewing the electronic data for any length of time creates a dilemma for both the vendor and the company responsible for monitoring and managing the clinical data. A fully complaint system will capture, within its audit trail, a record of the date and time of each user of the system and the duration of their activity therein.

Should they be monitoring the activity of the sites through the creation and review of standard reports showing how much time an investigator spends reviewing the data prior to adding an electronic signature? Where does the accountability for checking at this level of intensity lie – with the sponsor or the system vendor? Will an increased level of quality be achieved or will this approach simply alienate the study site staff?

If it is decided that such a level of checking is necessary then the clinical trial investigators must be made fully aware that this will occur and what the expectations are. System vendors must also make it easier for electronic data review to take place by enabling assessments to be pooled across visits to show trends and outliers, perhaps in a similar format to a traditional laboratory report.

The FDA has released a draft guidance entitled “Guidance for Industry – Electronic Source Documentation in Clinical Investigations,” which states “investigators should review completed portions of the eCRF for each subject before the data are archived and released to the [third] parties. The investigator should indicate that he/she has reviewed the submitted data. For example, an investigator might initiate an electronic command to enable transmission of data to [third] parties, or append a data element identifier (with the date, time, and originator’s name), indicating that the investigator has reviewed the data element. This command or appendage would be applied to all the data elements belonging to the portion of an eCRF reviewed by the investigator.”(1) If this draft guidance becomes official, this will help strengthen the process, but still there is work to be done. CDISC published a set of 12 user requirements for source data in 2006 (‘CDISC Standards and electronic Source Data Within Clinical Trials 20 November 2006’) that apply to all source data “irrespective of the media or technology used to hold the data”(2) which is another step closer to where we need to be.

In summary: If the golden rule laid down by Quality Assurance of “if it wasn’t documented, it didn’t happen” is to be changed to ”if no one reviewed the audit trail of an EDC/ePRO system, then it didn’t happen” we need guidance from the regulatory bodies as to their expectations and assignment of responsibility. The overall responsibility of the trial lies with the sponsor, however do they rely on the vendor for the EDC system to monitor the investigators reviews of data, or does this responsibility lay with the Data Manangement function? Alternatively, do we not check, because in the days of paper CRFs this is how it was, and it was just accepted, despite now it being easy to actually monitor the investigators’ actions?

Refs:
1. http://www.fda.gov/downloads/Drugs/.../Guidances/UCM239052.pdf
2. http://www.phtcorp.com/pdf/EMAReflectionPaper.pdf


To learn more about Chris Cramer, and Quanticate, you can check out their website, Facebook, or Twitter.

Related Content:

News | Blogs