In this whitepaper I want to discuss the concepts of study deviations, violations, non-compliances, and amendments; and to compare the US and ISO positions with regard to reporting.
When I say ISO, I mean ISO/FDIS 14155 "Clinical investigation of medical devices for human subjects—good clinical practice" (2010) which may be purchased from the International Standards Organization in Geneva. When I say US, I mean 21 CFR Part 812 plus a collection of FDA guidances and IRB voluntary standards.
A poll—how many deviations do your colleagues make?
Who deviates? Who amends?
Sponsors may, themselves, contribute to deviations. Sponsors do incredible things such as: 1) writing protocols with such tight specifications they cannot be reasonably followed, 2) authorizing changes verbally, but not following up in writing, 3) approving changes they have no authority to approve, 4) inadequately training the study staff, 5) not insisting on IRB or EC approval or notification, 6) not insisting on FDA approval or notification, or 7) changing the device—such as it's configuration, formulation, materials, or software—without prior approval or notification, 8) selling devices without FDA clearance (tantamount to conducting a study without approval), 9) trying an unapproved device on a few patients to see if the design is right, or 10) a host of other things I haven't thought of yet.
ISO/FDIS 14155 has no definition for amendment, but it can be deduced that it means any intentional change made to the investigator's brochure, protocol, case report forms, informed consent form, or other study documents; that the change must have the approval of the sponsor and investigator, and be noticed to the EC and regulatory authorities. (Section 6.5.1.)
Part 812 has no definition of deviation, but it does say that the investigator is responsible to keep a log of deviations from the protocol (Part 812.140(a)(4)). Emergency deviations taken to protect a subject are reportable to the sponsor and reviewing IRB within 5 working days (Part 812.150(a)(4)). All other deviations require prior approval by the sponsor and possibly the IRB and FDA, and we usually consider these to be amendments. The only reference to a noncompliance is in regard to good laboratory practices.
The IRB's view
The IRB doesn't think in terms of deviations so much as unanticipated problems. And they merge the issue of adverse events with problems that are anticipated or unanticipated; in other words an adverse event is one more type of problem.
[ ] Deviation—a failure to follow the study plan. A deviation that occurs repeatedly should probably become an amendment.
And what about reporting
For ISO: The monitor is expected to identify deviations, discuss them with the investigator, and list them in the monitoring report (Section 184.108.40.206.a.) The sponsor is expected to list the deviations in progress reports as requested by reviewing ECs and regulatory authorities (Section 8.2.3.f). Investigators that cannot be compliant may need to be excused from the study (Section 7.1.1.)
For FDA: The US is different from the rest of the world in that we have a risk classification system for clinical studies. The rules are different for non-significant risk and significant risk studies, and this system is unique in the international community. For significant risk studies, the investigator is expected to keep a deviation log (Part 812.140(a)(4) and report deviations to the sponsor (Part 812.150(a)(4)). The sponsor is expected to take steps to bring the investigator into compliance either by withholding shipment of additional devices or by reporting the investigator to FDA. These requirements do not apply to non-significant risk studies.
For US IRBs: The rules vary from IRB to IRB, but if the board follows the OHRP guidance, unanticipated problems should be reported in the short term (my phrase) and anticipated problems in the annual progress report.
Workshop learning objective
You will receive
Who should attend
Date, time, registration