GLP Quality Consulting



GLP Hot Topic- Report Generation Processes and a recent FDA Warning Letter

Scott C. Rumsey, RQAP-GLP

GLP Quality Consulting, LLC

It has been several years since an FDA focused upon the expectation of generating contributing scientist reports (covered in a separate White Paper). It is now clear that the agency expects a separate signed report from contributing scientists involved in significant sub-aspects of a GLP study (such as Pathology, Bioanalysis/PK, etc.), to be appended to the Final report. A good rule of thumb to follow is to generate a contributing scientist report if a scientist is involved in making any interpretations of the data. For example, a pathologist will make calls on the tissues observed on slides and then determine the relevance of pathology findings; a PK scientist will evaluate the test system exposure to the test article, etc.

The requirement for contributing scientist reports is covered in the US FDA GLPs where requirements for the final report are included in section 58.185(a) (12) where it requires, “The signed and dated reports of each of the individual scientists or other professionals involved in the study.”

Additionally, in the OECD Principles of GLP (section 9.1.2), it states that, “Reports of Principle Investigators or scientists involved in the study shall be signed and dated by them.” It is further suggested that a report from the Principle Investigator may be relevant in the OECD Multi-Site Study Guidance No. 13 where it states, “It may be useful for the Principle Investigators to produce a signed and dated report of the phase delegated to them for incorporation into the final report.” The Guidance provides further suggestions for content that largely mirror final report requirements. Of note is the OECD wording that appears to make the drafting of a Principle Investigator report optional. In practice, for multi-site studies, the FDA is expecting a report to be written. In cases where only tissue processing or other non-data generating activity takes place, it may be acceptable to include only a QA statement and a signed letter report that defines the work conducted as being in compliance with GLP.

However, the actual process of “how” reports are drafted reviewed and approved has not been a focus of FDA until recently. In August of 2010 a Warning Letter was issued to SNBL that included citations that impact draft report generation and review. The finding has implications for Industry practices and was cited as follows:

1. Testing Facility Management failed to ensure that personnel clearly understand the functions they are to perform [21 CFR 58.31(f)].

Specifically, Testing Facility Management did not provide personnel with proper instructions on the preparation of final reports. As a result, Study Directors routinely prepared, and the Quality Assurance Unit audited, final reports based in part on unsigned draft reports of contributing scientists. FDA found numerous examples of this significant deficiency during the inspections conducted in both 2008 (Studies SNBL.118.06, SNBL.210.07, and SNBL.053.06) and 2009 (Studies SNBL.258.10, SNBL.010.27, SNBL.289.03, SNBL.293.01, and SNBL.010.22). We note that the preamble to the 1978 Final Rule emphasized the importance of individually signed and dated contributing scientist reports in establishing the contributing scientists’ individual accountability for accurate reporting.1

 

Furthermore, the preamble to the 1987 Final Rule stated that “…only the signed and dated final report of the pathologist comprises raw data respecting histopathological evaluation of tissue specimens.”2 Because SNBL Study Directors prepared and submitted final reports to sponsors with draft contributing scientist reports that were not signed and dated, they lacked accountability and confirmation that the contributing scientists’ raw data was accurate. Furthermore, given that SNBL Study Directors invite the study sponsors to comment on and edit both the contributing scientist reports and the final report, bias in data interpretation and reporting cannot be ruled out.

Your December 9, 2009, written response claimed that SNBL misunderstood this issue as it was cited on the Form FDA 483, Inspectional Observations, dated August 1, 2008. Your claim has no basis for the following reasons:

 

• Inspectional Observation Item 7 clearly documented the deficiency found during the 2008 inspection thus: “The Study Director prepared the final report using unsigned draft reports of Contributing Scientists.”

 

• The SNBL written responses dated August 20, 2008, and October 2, 2008, stated that “Study Directors will be instructed to ensure that the final versions of contributing scientist reports are utilized during final report preparation.”

 

Consequently, your claim of misunderstanding the issue is not justified and lacks merit. In this regard, SNBL offers no assurance that SNBL will effectively correct this ongoing deficient practice and will adhere to GLP regulatory requirements for reporting nonclinical safety data.

This finding seemingly implies that the study director cannot use draft contributing scientist reports to create a draft final report. It also implies that a sponsor cannot review draft reports due to a perceived ability to bias the report with comments.

The finding also implies that the Quality Assurance Unit cannot audit the final draft report using draft contributing scientist reports.

It has been industry practice to create draft reports for QA and sponsor review. These reports would then be updated based upon any errors or findings prior to being signed final. QA would review and changed in the draft reports for accuracy prior to signature. For contributing scientist reports, these same procedures, with the addition of a study director review would typically be performed. Some sponsors typically request that a QA audited draft report be submitted for sponsor review.

While it is understandable that the FDA would be concerned with potential for undue influence and errors in the review of draft documents, it is not realistic to expect that some of these practices should be curbed. Ultimately, the study director is responsible for the accuracy of the reports. Quality Assurance is responsible for citing any discrepancies. Draft reports are reviewed for changes prior to signature. Checks and balances have been put into place in the GLP regulations that would require collusion between personnel to willfully falsify data or results. Practices identified in the SNBL Warning letter (working on draft final reports with draft contributing scientist reports, sponsor review of draft reports, etc.) have been ongoing for years with no instances of serious problems identified.

Some of the Implications:

1) Study directors would have to wait until signed contributing scientist reports were available to begin drafting a final report.

2) More report amendments would need to be generated as a review process for draft reports by sponsors and study directors would be curbed.

3) Sponsors would not be able to review and comment on work that is in draft form. How this can be accommodated in light of the fact that sponsors need to monitor their contracted studies is unclear.

4) QA could not audit draft reports, but would need to wait until after a report is signed. This would generate report amendments to accommodate any changes resulting from a QA audit. This would also remove that ability of QA findings to be confidential, as any resulting report amendments would define the findings of the QA Unit.

5) Timelines for the conduct of GLP regulated studies would need to be expanded as each segment of a study would need to be signed-off before the next segment could be started.

The takeaway: At this time no changes should be made to report generation and reporting practices until clarification of the FDA position on the issue is provided. A letter was sent by the Society of Quality Assurance (SQA) on October 8, 2010 objecting to the apparent change in Agency policy on drafting reports. No additional examples of FDA findings related to this issue have been identified as of this writing.

References:

FDA Good Laboratory Practices Regulation 21 CFR Part 58:

OECD Principles of GLP: (98)17&doclanguage=en

Application of the OECD Principles of GLP to the Organization and Management of Multi-Site Studies: (2002)9&doclanguage=en

FDA Website-Warning Letter to SNBL Dated August 9, 2010:

-----------------------

Quality Solutions for the Life Sciences Industry

Quality Solutions for the Life Sciences Industry

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download