FDA's CLIA waiver application draft guidance could offer test sponsors added flexibility in designing clinical studies.

The ^FDL-1 document, slated for publication in the Sept. 7 Federal Register, follows years of deliberation between the Clinical Laboratory Improvement Advisory Committee, AdvaMed and other stakeholders.

Currently, manufacturers primarily rely on a 1995 joint CDC/CMS proposed rule to determine whether to pursue a CLIA waiver for a diagnostic. However, stakeholders have long been interested in making the criteria more flexible (¹ Also see "FDA Draft Guidance Will Aim To Increase CLIA Waiver Flexibility" - Medtech Insight, 18 October, 2004., p. 11).

In April 2004, FDA officially took authority from CMS for categorizing the complexity of diagnostic tests according to the Clinical Laboratory Improvement Amendments of 1988.

The agency determines whether a test can be waived or is of moderate or high complexity. CMS certifies laboratories and other medical facilities to perform tests within each of these categories. A waived product can be used in the broadest array of facilities with the least oversight. Thus, the designation might confer marketing benefits (see ² Also see "BioSite Triage Becomes The First BNP Test To Receive A CLIA Waiver" - Medtech Insight, 5 September, 2005.).

However, some hospital and healthcare system representatives have expressed concern that too many waivers could lead to lapses in patient safety (³ Also see "CLIA-Waived Test Market Increasing As Public Health Concerns Loom" - Medtech Insight, 9 May, 2005., p. 20).

Waived tests, as defined by CLIA, "employ methodologies that are so simple and accurate as to render the likelihood of erroneous results by the user negligible."

The CDC/CMS proposed rule stressed the use of a reference material to prove test accuracy; however, the new draft guidance appears to provide more leeway to sponsors regarding what firms can use as a comparative method.

Industry stakeholders have expressed concern about the difficulty of proving accuracy for new analytes that lack well-characterized reference materials.

The draft recognizes "that reference methods may not be available for every device type." It notes that a reference method that has been "thoroughly investigated" is preferred, but it offers a number of alternatives relying on mathematical "traceability" techniques for arriving at an acceptable comparative method.

The new draft also emphasizes that clinical studies should recruit "intended users" and that actual patient samples should be used whenever possible.

Sponsors have often hired non-trained personnel to use the test kits with mock specimens in clinical trials, under the rationale that CLIA-waived tests should be easy to perform (⁴ Also see "CLIAC Waiver Criteria Input Readied As FDA Gears Up To Draft Guidance" - Medtech Insight, 23 February, 2004., p. 12).

However, the draft guidance advises that "you should enroll individuals who represent anticipated operators of the device," which would include medical assistants, nurses or doctors.

FDA suggests that samples be obtained at a minimum of three clinical sites and tested by a minimum of nine users. For kits that yield a quantitative result, at least 360 samples from the intended-use population should be tested. However, in cases where this is impractical, up to 60 patient specimens can be substituted with contrived samples.

For tests that provide a qualitative "yes" or "no" type of response, "the goal should be to have a minimum of 120 samples positive by [the comparative method] and a minimum of 120 samples negative." For low prevalence diseases, sponsors should consider geographical and seasonal variations to maximize the positive samples and, if necessary, supplement studies with archival or contrived samples.

The draft provides detailed guidance on statistical analysis and performance criteria assessments for clinical studies to support a CLIA waiver.

The document also stresses the importance of "failure alerts and fail-safe mechanisms," including "lock-out" functions that block results from appearing if the device is mishandled or if system checks are not completed. This requires a thorough hazard analysis that includes "flex studies," in which the kit is tested under conditions of operational stress.

A sponsor should also develop external control samples that can be included in the test-kit packaging. Product labeling or package inserts should include step-by-step instructions on how and at what frequency controls should be run, the guidance suggests.

FDA will accept comments of the draft until Dec. 6. The Clinical Laboratory Improvement Advisory Committee has a meeting planned for Sept. 7-8 in Atlanta.

Long-Awaited CLIA Draft Guidance Offers Some Leeway For Study Design

More from Archive

More from Medtech Insight