FDA has published a proposed rule, “Reporting Information Regarding Falsification of Data” in the February 19 Federal Register. FDA wants to require sponsors to submit reports to FDA about false data, or the possibility of false data, within 45 calendar days after the sponsor becomes aware of the information. The proposed rule is open for comment until May 20, 2010. (A proposed rule is an FDA proposal to change an FDA regulation.)
FDA states that sponsors are in the best position to determine if there has been falsification of data used inclinical trials or in nonclinical (GLP) research. FDA notes that “through their monitoring, auditing, and reviewing of data,” sponsors can spot false data that FDA might not detect during a bioresearch monitoring inspection.
FDA issued the proposed rule because current regulations are unclear on the need to report falsification of data to FDA. The proposed rule states that sponsors are “reluctant or uncertain as to whether it is necessary to report the information (to FDA).” FDA notes that sponsors have excluded data of a clinical investigator suspected of data falsification or terminated their participation in a study without stating the reasons. FDA is concerned that persons suspected of data falsification then continue to participate in other studies. FDA states:
“Taking effective action in response to falsification could lessen the magnitude and impact of the falsification in a current study, reduce the potential for delays or compromise to other studies and applications (including studies and applications from other sponsors for whom such a person might also be working), and protect the rights, safety, and welfare of research subjects.”
FDA makes these comments in the “preamble” to the proposed rule. The preamble is FDA’s explanation of why the proposed rule is necessary with definitions and clarifications of those reasons. The preamble is followed by the actual proposed wording of the changed regulations.
A Closer Look:
“Such a person” is nicely put, but the proposed rule doesn’t address whether clinical trial vendors are included in their concerns. There is not an actual mention of contract research organizations, clinical laboratories, IVRS vendors, eCRF systems, etc. There is a reference to, “any confirmed or possible falsification by any person involved in studies conducted by or on behalf of a sponsor or relied on by a sponsor.” FDA is very clear they want to include the possibility of falsification.
The proposed rule discusses the serious reasons FDA is concerned with the falsification of data. However, thedistinct impression that I had from reading the proposed rule was that FDA is primarily concerned with the activities of a principal investigator at a clinical site. There is no discussion of the possible root causes of data falsification. During my career I have been asked to conduct “for cause” audits where data falsification was suspected. I primarily found a confused, undertrained staff member making unintentional mistakes.
Then there is the very real problem of overworked and/or undertrained staff deliberately altering data in an attempt to satisfy a supervisor and keep their jobs. This can happen not only at clinical sites or nonclinical laboratories, but at the many vendors participating in research that are seldom inspected by FDA.
FDA asks, “How does a sponsor become aware of data falsification?” They offer a few brief examples including the “monitoring review and evaluation of study data (e.g., noticing unusual data on case report forms and/or analytical reports).” After issuing Warning Letters to J&J and ICON Clinical Research, Inc. (see previous posts) it is clear that GxP professionals working for sponsors need to develop the techniques necessary to determine fraud. FDA told ICON:
“Study monitors failed to identify that on multiple occasions, site personnel documented administration of study drug to different subjects at precisely the same time.”Stop and think about this. It is entirely possible for monitors to review different charts at different visits andnot discover this possible fraud. When I conduct site audits I rarely compare things such as the time of administration of study drug across different charts. I usually review one chart at a time (maybe I should change).
These are issues that need review of data listings and biostatisticians who can look for funny numbers that aren’t necessarily evident to monitors and auditors (and FDA investigators). We need to utilize a variety of tools and not depend entirely on monitoring visits and site audits. One possible unintended consequence of this proposed rule is that clinical sites will be audited to the point it becomes a regulatory burden, not a safeguard for data integrity and the protection of study participants.
Consequences of Noncompliance:
It is interesting to note that FDA asks, “What are the consequences of not reporting confirmed or possible data falsification?” The answer should be sobering. It includes the possibility of a violation of 18 U.S.C. 1001. You will note that Title 18 of the US Code is not specific to FDA. It is the section of Federal law that includes the basis for felony prosecutions. Tough stuff.
Request for Comments:
FDA offers definitions of falsification of data, a definition of “data,” and a definition of unintentional errors. FDA also explains why they consider 45 days to be appropriate and a number of other issues. The comment period is used to clarify and seek input for these definitions and issues of concern to stakeholders and the public. The proposed rule lists several ways to comment. This is an important proposal for changing FDA regulations for food, drugs, biologics, medical devices, and veterinary medicines. It is an attempt to address legitimate concerns for public health and consumer protection. It deserves our attention.
Look for yourself at:
Remember- YOUR comments are always welcome