Review of the FDA’s draft guidance on ‘Content of Human Factors Information in Medical Device Marketing Submissions’

This article provides three observations of the FDA’s recent draft guidance on the ‘Content of Human Factors Information in Medical Device Marketing Submissions’ that manufacturers need to be aware of when preparing a Human Factors and Usability Engineering Report (HFE Report)

The Background

In December 2022, the FDA published draft guidance on the ‘Content of Human Factors Information in Medical Device Marketing Submissions’. The guidance provides a risk-based approach to guide manufacturers and FDA staff on the human factors information that should be included in a marketing submission to the Center for Devices and Radiological Health (CDRH).

When finalised, this guidance will supersede the FDA’s guidance on the ‘List of Highest Priority Devices for Human Factors Review’ and will also complement their guidance on ‘Applying Human Factors and Usability Engineering to Medical Devices’ (referred to as the Human Factors Guidance). Also, the FDA plans to use this guidance to update the definitions in the Human Factors Guidance, as well as the descriptions of expected content in each section of the HFE Report.

The risk-based approach involves using a flowchart with decision points to determine which of three submission categories is relevant. The flowchart is applicable to both new and modified devices. A key aspect of the flowchart is whether critical tasks exist for new devices or whether critical tasks are impacted for modified devices. This has been depicted in the following slide taken from BERGO’s training course on medical device usability.

Slide on HF Submission categories

The slide also shows which section(s) of the HFE Report should be prepared based on the category of HF submission. The remainder of this article will discuss three key observations of this draft guidance.

Observation 1: Greater harmonisation of definitions

The Human Factors Guidance will likely incorporate the following:

• Revised definitions for terms formative evaluation, hazardous situation, human factors engineering, task, use safety, and user interface.
• Added definitions for terms harm, normal use, residual risk, serious harm, serious injury, use environment, use-related risk, and use-related risk analysis.

A key observation is the FDA’s motivation to align their definitions to those in IEC 62366-1:2015+AMD1:2020, ISO 14971:2019, and ANSI/AAMI HE75:2009 (R2018).

One missing and desirable update, not mentioned in their guidance, would be to harmonise the terms ‘human factors validation testing’ (in the FDA’s guidance) and ‘summative evaluation’ (in IEC 62366). Those terms are synonymous therefore one term would be preferred.

Also, further clarification would be desirable on the definitions of ‘serious harm’ and ‘critical task’. In the draft guidance, ‘harm’ and ‘hazardous situation’ are defined to include “damage to property or the environment” and how those aspects “are exposed to one or more hazards”. The definitions of ‘serious harm’ and ‘critical task’ do not currently incorporate those aspects.

Lastly, it would also be helpful if the draft guidance provided a clearer example of how to use the definition of ‘serious harm’ when linking it to a classification of severity. Typically, on a 5-point scale of severity, a task would be deemed critical if the severity was estimated to be 3 or above.

Observation 2: Greater clarity on documentary expectations for modified devices

The draft guidance has provided much needed clarity on not just how to tailor the documentary effort for satisfying the Human Factors Guidance but also on the documentary effort needed for modified devices.

While the risk-based approach is something already in practice – tailoring the effort based on critical tasks – the linkage to different submission categories is helpful, especially for modified devices where the documentary expectations were unclear.

When comparing the modified device with the existing device, greater clarity has been provided on the content and format of information to be documented in the HFE report with respect to:

1. the user interface (section 3 of the HFE report)
2. use-related risk analysis (referred to as a ‘comparative use-related risk analysis’ in section 6 of the HFE report)
3. the identification of critical tasks (section 7 of the HFE report)

With respect to (1) and (2), example tabular formats are provided (Tables 3 & 4). However, in table 3, a column would be useful to record the outcome of the comparisons to determine whether critical tasks have been impacted.

Observation 3: Revisions to expected content of HFE report

The first observation is that the order of sections 5 & 6 of the HFE report have been switched in the draft guidance. This appears to be more logical as the summary of preliminary analyses and evaluations (section 5) helps lay the foundation for a more comprehensive use-related risk analysis (section 6).

Further in section 7 on ‘Identification and description of critical tasks’, clearer instructions are given on how critical tasks are determined. Manufacturers need to provide a table of severity levels with descriptions of each level. However, as noted in the last paragraph of observation 1 of this article, an example of how to use severity levels for determining critical tasks would be helpful.

Lastly, in section 8 on ‘Details of HF validation testing of final design’, further clarification is given on what is expected when residual risks are found to be unacceptable in a HF validation test. This guidance sets the expectation that if other means of risk control and mitigation measures are required to reduce risk, then those measures need to be re-evaluated.

Conclusions

The risk-based approach and different categories of HF submissions provides much needed clarity on the FDA’s expectations on the content of human factors information in medical device marketing submissions. Particular emphasis has been on modified devices.

There is also greater alignment with recognised standards and clearer instructions on how to identify critical tasks. However, as noted in this article, further refinements would be desirable. Also it would be helpful to provide a clearer illustration on how the HFE report links to deliverables of the Usability Engineering File.

Lastly, it is not clear why a separate guidance document is needed to present the risk-based approach. Given that a significant proportion of this draft guidance will be incorporated into the Human Factors Guidance, it would seem more useful to simply update the Human Factors guidance than to have two documents containing the same information.