As we resume our discussion of Human Factors Engineering (HFE) for medical devices, it's worth repeating that HFE should be thoroughly incorporated into the development process. HFE is about working ahead to catch and prevent potential user errors before they happen. Ideally, iterative testing against anticipated use-based risk would be scheduled in lockstep with development phases.
The FDA recommends prioritizing safety measures during development as follows:
1. Safety Through Design
2. Protective and Preventative Measures
3. Safety Information and Training
Notice how this is order is reversed in comparison with a project that lacks HFE processes? Without HFE, the common (and, in the case of medical devices, flawed) approach would be to create the product, then provide documentation and training, then implement protective and preventative measures and redesigns in future versions / upgrades based on testing of the finished product and post-release feedback.
That seems like a rational approach because it lets you gather data on user errors based on interactions between real users and testers and your finished product. That data can then be evaluated and ranked based on the probability of occurrence, and the results used to inform requirements going forward.
For many products, this is the correct approach. But it's unacceptable for medical devices because the potential severity of harm from user error is much higher. That’s why the FDA recommends that in medical device development, use-based hazards should be ranked and preempted according to severity, rather than likelihood.
Again, seems backward! But I'm sure you see why: without the kind of probability data that you would typically get from testing, you can't reliably anticipate the likelihood of any given error; but without anticipating and preventing potential errors, you can't ensure that testers and early adopters won't be harmed during your device's initial deployment. HFE's “backward” approach to medical device development is your ticket out of this catch-22 conundrum.
Here's how it works:
First, as you begin each phase of development, your team must work to identify ways that users might interact incorrectly with your device, based on its anticipated final form and any changes made during the previous phase. Having feedback from a representative focus group of targeted users may help with this process.
Next, rank the possible incorrect interactions you've identified by severity. For example, using the wrong colors to indicate an increase or decrease in stimulation therapy vs. an improperly placed On/Off button that can be accidently pressed. If a given error occurs, what might the consequences be? Resist the temptation to under- or over-emphasize specific errors based on gut intuitions about their probability.
Then, identify ways you can prevent or mitigate these potential errors, starting with the most severe (unacceptable risk) and working down to the least critical (an acceptable risk). For every mistake, seek a design-based solution first (examples: connectors unable to connect to the wrong component; automation of tasks prone to error when attempted manually), then protective or preventative measures (ex.: security features; warning screens and alerts), and rely on safety information (ex.: labels; manuals) and training only as a last resort.
The reasoning behind this design>prevention>training philosophy is straightforward: people neglect information and people forget training. Sometimes people even try to circumvent preventative measures, if they think they're unnecessary or working incorrectly. But product design guides users intuitively into specific modes of behavior or paths of interaction. Even if the information is absent, training is forgotten or unavailable, and preventative measures are bypassed, the design presents the same nonverbal suggestions for user interaction every time. That's why engineering design for human factors is so important: it's the most effective means of reducing and eliminating use-based hazards.
Velentium does HFE not only to ensure patient safety but also to catch design flaws early. When faults are identified during the requirements and design phase, fixing them is orders of magnitude less expensive than fixing them after release. If you also account for the cost of product redesign post-release, not to mention the cost of possible fines or lawsuits caused by poor design, an integral HFE process is not only the right choice for patient safety but the smart business choice as well.
In our next post, we’ll look ahead at the final phases of HFE integration into medical device development: Human Factors Validation Testing. Knowing what this is and whether it will be required or beneficial for your project lets you plan for it as early as possible; failing to plan might mean blowing your project budget and schedule as you scramble to comply.