Frances Liem, Head of Laboratory Data Integrity Branch, Office of Compliance, US EPA, presented these items in two talks.

Her first presentation summarized EPA's inspection program and items noted during inspections over the last year.

In FY99, 113 laboratories were audited; 58 had not been inspected previously. She presented three enforcement actions:

Colgate Palmolive - product chemistry studies; the Sponsor conducted the studies. Fine = $10,000

Rockland Corporation - product chemistry studies; Chemtest Laboratory in New Jersey conducted the studies. Fine = $15,000

Wexford Laboratories - antimicrobial studies; the laboratory was not identified since the laboratory had done the studies properly. Fine = $200,000 This was a unique instance where the Sponsor (Wexford) had conducted several studies using an antimicrobial product at an independent laboratory. The Sponsor did not agree with the study results and modified the reported results prior to submitting the reports to the EPA without following the proper procedures (i.e., preparing a report amendment). She did not give many other details.

She did not list specific issues for any one enforcement action, but indicated that in general items included: erroneous reporting, lack of a QAU, no QA statement in the final report, SOPs not approved by management, lack of an approved protocol, data not recorded, protocol lacked elements required by GLPs, final report lacked elements required by GLPs or did not contain a compliance statement, failure to retain raw data.

There were 21 studies rejected by OPP (scientific reviewing branch) for GLP issues. She summarized these as follows:

Antimicrobial studies; seven studies from two laboratories

Product identity/chemistry; three studies from one laboratory

Toxicology; 11 studies from three laboratories

The reasons given were selective reporting (did not report all data); data not recorded and maintained; and the compliance statement indicated only that the study was not GLP and did not give the differences from the regulations. She went on to say that, historically, studies have been accepted with compliance statements indicating the study was not conducted in compliance with GLPs. However, OPP is beginning to require that the compliance statement indicate the specific items that were non-compliant (e.g., no QA audit, no SOPs, etc.). This follows the requirements stated in the GLPs (see 160.12).

Some additional items presented by Frances were what she termed as "Puzzling Encounters" and "Industry Misconceptions". These included:

Final reports were revised after QA had performed an audit and no follow up audit was performed.

Data were recorded to reconstruct a study (e.g., critical data had not been recorded at the time of the observation, but was added at a later date).

Required QA records were not maintained. There has been an idea from some in industry that all QA records could be disposed of after the study was completed.

QA personnel making changes to raw data.

QA personnel performing study related facility activities (e.g., some QA are involved in equipment calibration scheduling and conduct).

EPA can request to see the written QA procedures (SOPs), but not QA findings and responses.

The second presentation by Frances was a question and answer session. The following is a summary of the items presented and discussed during this session:

The Study Director signs the protocol after the Sponsor approves it (i.e., the SponsorŐs representative should sign first).

The stability of each batch of test substance used in a study has to be determined. This does not need to be a separate study, but can be as simple as analyzing the test substance before and after use in a study or series of studies.

All solvents must have expiration dates. She suggested having a SOP to define how these dates are assigned.

Reference substance characterization must be conducted according to GLPs.

The EPA prefers to see the use of "same as above" rather than dittos or arrows. She said that no matter what is used, it should be specified in a SOP.

SOPs on computer system should have adequate security to avoid intentional or unintentional changes (e.g., be read only) and there should be a contingency plan if the computer is "down". If they are printed, she indicated that she would like to see a date/time stamp indicating when they were printed and the SOP specify that they can only be used on the day of printing.

When describing changes to the protocol in the report, EPA prefers these to be noted in the text where the issue occurred rather than a listing an appendix. In her opinion, the best would be to do both. However, as long as the differences are described in the report, the requirement has been met.

If the raw data is unstable (e.g., thermal paper, electrophoresis gels), something should be done to preserve the data. For example, make copies, take photographs, etc. The original data should be archived and retained as long as the quality affords evaluation. The method for addressing this issue should be indicated in the protocol or a SOP.

No studies destined for EPA are exempt from compliance to GLPs.

- Mary E. Lynn