Good Clinical Practice: A Question & Answer Reference Guide - May 2018 - 89

Investigators/Sites
In a 2009 talk, a CBER compliance official noted that study site records and experiences that are too "perfect" will
also raise suspicions and may even make a site a "target" of an inspection. Adverse events or missed appointments
by study subjects are examples of issues that the agency would expect to see some evidence of, and the utter lack
of which might raise suspicions.
2.88 Q. Must an investigator gain IRB approval before implementing EVERY type of research change (except
those to eliminate immediate hazards)?
A. 21 CFR 312.66-Assurance of IRB Review states that an investigator must assure that he or she "... will not make
any changes in research without IRB approval, except where necessary to eliminate apparent immediate hazards
to human subjects." Although the FDA does not provide examples of research changes that require IRB review,
agency regulations make clear that even many "minor" changes in research require some form of IRB review and
approval. FDA regulations at 21 CFR 56.110, for example, permit IRBs to employ expedited review procedures
(e.g., review by IRB chairperson or experienced IRB member) in assessing "minor" changes in ongoing, previously
approved research as an alternative to full IRB review.
In addition, IRBs often impose their own reporting requirements that are more demanding than FDA regulations.
In these cases, under 21 CFR 56.113 (which grants IRBs the authority to suspend or terminate the approval of
research not conducted in accordance with IRB requirements), the IRB directives, in effect, have the same force
as FDA regulations.
Changes that are not minor, which the FDA says in a 1998 IRB Information Sheet would include the addition of
procedures involving increased risk or discomfort, require review and approval at a convened IRB meeting before
an investigator can implement.
As general guidance, the listing below provides several examples of study changes and their implications for
IRB review and approval:
* If a test or procedure is added or subtracted from a study, the IRB should approve this change.
* If one manufacturer of a test kit, say a urine pregnancy test, is substituted for another, then IRB approval is not
necessary, unless, for example, the sensitivity and/or specificity of the test is dramatically different.
* If the local laboratory changes its equipment for performing certain lab tests, then IRB approval is not
necessary.
* Changes in telephone numbers or site staff do not necessarily need approval, although the IRB should be
notified in such situations.
* A change from office-based visits to home visits for simple clinical assessments does not require an IRB
approval, although an approval may be necessary if tests or procedures such as blood draws are involved.
The clinical ramifications (so-called spillover effects) of every change should be assessed carefully by the
principal investigator. When in doubt, the investigator should always submit the change to the IRB for guidance.
An informal telephone call to the IRB administrator or chair might also prove helpful.
2.89 Q. In documenting a patient's eligibility for a study, must the investigator obtain the prospective subject's
previous medical records, or are history forms completed at the start of a study considered adequate?
A. An anticipated FDA guidance on informed consent could provide new agency perspectives in this area. In
September 2010, the agency noted that the new draft guidance will address several topics, including "review of
subject records."
In a recent informal response to this question, the agency's Good Clinical Practice Program noted that, "to
enroll a subject, the investigator must establish that the subject meets the eligibility criteria contained in the study
protocol. Taking a complete medical history is very important, but in general it would probably be coupled with
screening tests or procedures performed at the time of enrollment (e.g., x-rays or other scans, blood or other lab
tests), physical examination of the subject, etc., to ensure that the subject meets the inclusion and exclusion criteria
specified in the protocol. In other cases, establishing eligibility may be more difficult, and it may be necessary to
have documentation about the subject's condition prior to joining the study (e.g., through access to the subject's
prior medical records). For example, the study may exclude subjects who suffer from, or previously suffered from
depression, or used an antibiotic in the last 6 months. The subject's medical records would help to establish that
the subject meets such criteria.
89

Table of Contents for the Digital Edition of Good Clinical Practice: A Question & Answer Reference Guide - May 2018