What You Need to Know about FDA’s Draft Guidance on Writing Test Reports

New draft guidance recommends formats for test and summary reporting.

Image source: AdresiaStock/Shutterstock

On May 31, FDA released the draft guidance (DG), “Recommended Content and Format of Complete Test Reports for Non-Clinical Bench Performance Testing in Premarket Submissions.” The DG addresses both individual test reports and summary reports that are intended for regulatory submission. Such reports might be different from other internal test reports that are not prepared for submission purposes. Non-clinical means bench tests and animal work are excluded from the DG, as are other biocompatibility and sterilization testing.

The recommended test report has six sections:

  1. Description of test.
  2. Objective of test.
  3. Description of test methods and procedures.
  4. Pre-defined pass/fail criteria.
  5. Test results.
  6. Discussion of the conclusions.

Item 3 includes test sample information, test sample size/selection, and the test protocol. It is noted that the tested devices or parts should represent the final, finished device that has been subjected to all manufacturing processes (including sterilization), environmental conditioning, and simulated transportation. (For example, having lived in Texas for many years, I found it interesting that the temperature inside a truck trailer in that environment can exceed 120° F. A similar effect may be experienced in the case of devices that are stored in the trunks of representatives’ cars.) For sample size, it is suggested that there be a scientific rationale. This is apparently instead of the sample size being random and capricious.

As item 4 suggests, the type of testing envisioned here is performance against specifications, rather than open experimentation. Having pass/fail criteria in advance is quite different from performing a test, looking at the results, and then deciding whether those results are acceptable. However, pass/fail criteria may have evolved earlier in the design cycle, and more specifically in the pre-design research cycle. This is why some people carefully separate R from D, noting that “research” is not a regulated activity while design is (and noting that “development” is perhaps somewhat vague). Another way to consider this point is that verification testing, and in particular testing for the purpose of a regulatory submission, should not produce any surprising results, and you clearly can’t include in a regulatory submission that your device failed your own test requirements. This is further emphasized in item 6, for which it is recommended that “For 510(k) submissions, you should discuss how the conclusions demonstrate substantial equivalence of your device to the identified primary predicate device based on known clinical performance, device performance specifications publicly available, and/or relevance of your specified acceptance criteria to the intended use of the device.” Note that this is specifically a regulatory conclusion, as opposed to what might otherwise be an engineering testing conclusion for verification, i.e., we met the specification. Notably the term “verification” does not appear in the DG. One might also note that FDA wants to know, and wants you to tell them, what the test results mean rather than just what they were.

Summary report recommendations are similar to the above, except they will appear in the body of the submission rather than in appendices. Summary report items are also similar to those above, except in item 5, where “Results Summary” replaces “Test Results.” In addition, an item 7 is added, which is a table of contents for the individual test reports included as appendices.

In some ways, it is helpful to have the specific format and content that FDA expects (or just suggests), if not quite requires. But to follow the DG, it might be necessary to reformat and modify current practice into FDA’s recommended style. This can be at least an interim burden to companies with established practice. Here, being a newcomer has an advantage, because you might not have an existing practice yet. It can also be noted that there is nothing surprising here, and this DG follows what is FDA’s usual approach—to adopt what is otherwise reasonable practice, rather to make up entirely new practices.

William A. Hyman

William A. Hyman is a professor emeritus in the department of biomedical engineering at Texas A&M University and adjunct professor of biomedical engineering at the Cooper Union. Reach him at [email protected].

500 characters remaining