Continuing non-compliance should be reported, as well as serious non-compliance, CDER DSI officials say.

WASHINGTON -- Examples of clinical investigator non-compliance that should be reported includes continuing non-compliance, such as, where the principal investigator makes the same mistake repeatedly, particularly after the IRB has informed the investigator of problems, two Center for Drugs officials told the Public Responsibility in Medicine & Research (PRIM & R) annual meeting here Nov. 17.

David Gan, M.D., and Sharon Gershon, of the Division of Scientific Investigations (DSI) Good Clinical Practice Branch, said an example of serious non-compliance that should be reported is where the investigator "conducts non-exempt human subjects research without IRB review and approval, or without appropriate informed consent," or "where the researcher makes significant modifications to IRB-approved research without IRB approval."

Other examples of continuing non-compliance the officials provided at the recent PRIM & R meeting are when:

* An investigator has multiple problems with non-compliance over a long period; and

* An investigator has problems with multiple projects.

However, Gan stressed, it is important to remember that non-compliance reports do NOT go to the Office of Research Integrity (ORI). Only reports of scientific misconduct are reportable to ORI, including the following:

* Fabrication -- making up data or results and recording or reporting them;

* Falsification -- manipulating research materials, equipment or processes, or changing or omitting data or results; and

* Plagiarism -- appropriating another person's ideas, processes, results or words without giving proper credit.

Other types of non-compliance, the FDAers continued, are "isolated and minor failures to comply with human subject protections or FDA good clinical practices." Failure to address minor issues "may lead to future serious non-compliance or continuing non-compliance."

The DSI-recommended way for an IRB to handle investigator non-compliance, said Gan, is the following:

* Identify who will lead the institutional process;

* Conduct initial triage of non-compliance;

* Make required reports within and outside the institution;

* Investigate and audit;

* Address findings with corrective action; and

* Follow-up--make additional OHRP/FDA reports, and monitor implementation of corrective action.

"IRB policies and procedures must specify who (person or body) will receive reports of non-compliance and make initial determination of type of non-compliance," said Gershon.

If the non-compliance is serious or continuing, it must be reported to the IRB, appropriate institutional officials, OHRP and FDA, sponsor/CRO and the investigator. Remember that "if the IRB suspends the study while the determination of serious or continuing non-compliance is being made," it must be reported, Gan said.

The timeframe to report non-compliance depends upon the seriousness of the incident. The regulations state "prompt," but it can range from a few days to up to one month. Gan said: "Report as soon as it is certain that any non-compliance has occurred," and "a follow-up report should be filed when investigations of all allegations are completed or a corrective action plan has been developed or implemented."

Items that should be included in a report to OHRP or FDA include:

* Name of the institution where the research was conducted;

* Name of the principal investigator on the protocol;

* Name of the research project where the non-compliance occurred;

* Number of the research project;

* Detailed description of the non compliance; and

* Actions the institution is taking.

When it receives a non-compliance report, OHRP reviews for adequacy of information and corrective actions; responds stating that the report was adequate or requests additional information; may open a compliance oversight evaluation; and an incident report is sent to OHRP, which will be made available to the public via FOIA.

"OHRP looks closely at corrective actions and assesses whether or not they will help ensure that the incident will not happen again," said Gan. Corrective actions should be systematic in nature.

By Joseph Pickett, Managing Editor