Adopting neuroscience-based human factors methods could yield better outcomes for patients, improved communication design and engineering teams, and maybe even help companies build IP.
The methods behind assessing human factors in medtech are outdated—and they’re due for an upgrade, said Charles Mauro CHFP, of Mauro Usability Science. “The type of research employed to determine if drug delivery devices are usable for the intended patient population is heavily mired in methodology that’s at least 100 years old, which is basic human observation of people interacting with technology combined with the personal opinions of those individuals involved in such studies.”
He describes a typical human factors research lab as follows: Say a team is testing an injector pen. The group will organize a study and use a room that is not much more than a market research room. The team will bring in appropriate respondents (usually those that profile as theoretical patients) to look at several similar drug delivery devices and perform action tasks associated with the device (e.g., injections into a pad). The tasks are followed by preference questions such as, “Which one of two devices do you prefer? Why do you like it better?”
This method is simple observational research and it does have value. However, Mauro said the personal preference questions, in particular, are fraught with problems of inefficiency, personal bias or bias generated by how the moderator asks the questions and, most importantly, an inability to measure actual performance of a device in ways that can be adapted by engineering teams to produce objectively better drug delivery devices. Today too much human factors research is really little more than traditional market research with a task sequence added, Mauro said.
What is missing from the scenario is structured, actionable data. “If you’re comparing two auto-injectors, you may get the respondent saying, ‘Oh, I like A over B,’ but in fact, when you look at the actual human factors research data ‘A’ may have had a much less effective delivery outcome, when measured in true human factors engineering terms,” he explained.
For example, perhaps the angle at which the device was held was not right. Perhaps the respondent didn’t hold it long enough to ensure correct dosage. And what if the respondent had micro-tremors (e.g., due to age or nervousness, etc.) as they used the device, creating greater pain. The same types of problems surface when testing the instructions for use that come with a drug delivery device. The FDA requires both drug delivery devices and instructions for use be testing for understandability and error reduction. Again, using traditional observational research approach does not provide the company hoping to improve the instructions for use with critical micro-level information on what is confusing and or too complex. Instead, the research team must rely on verbal feedback from the study participant who will often have no idea what is actually causing them to be confused when reading and viewing instructions for use.
A sea change
Compare those data with those generated by new advanced neuroscience-based data capture tools such as 3D spatial tracking. Spatial tracking can get down to sub-millimeter levels to assess whether that individual pushed the auto-injector deep enough into the injection path, whether they held it at the right angle and whether the injection time required to complete the injection was fulfilled
Combined with other advanced measurement tools, such as Newtonian force measurement, high-definition electromyography, high-res eye tracking, and micro-facial expression analysis, automated task analysis and data capture, cognitive workload analysis, and information foraging theory, all of which are employed by Mauro’s company, human factors teams have the power to provide data in critical new ways. For example, high-resolution eye-tracking when testing instructions for use or package designs for drug delivery devices makes it possible to determine exactly what the patient is viewing. It illuminates detail such as the order they look at the instructions, for how long, and how often they return to certain information that is confusing. This makes improvement of instructions for use far more productive and saves having to perform expensive and repeated retests.
Mauro explained that the FDA mandates designs intended for use by specific patient populations whose cognitive and physical capabilities may vary widely from the healthy population. But there is a serious lack of information on what those parameters might be “There’s no concrete data on the strength of an 80-year old woman with osteoporosis, so when we recruit the respondents based on their actual drug profile, we find that current devices require excessive muscular effort or force activation.”
He noted his belief that an entire generation of drug delivery devices may not be sufficiently human factored for certain patient populations. There are new technologies for measuring these critical variables and developing engineering design specifications for device operation.
Mauro said, “The ability to objectively view how much muscular effort is being applied vs. how much force is required opens up a whole new way of thinking about human factors.”
“It takes human factors engineering from basically a craft-based observational methodology to a very strong science-based methodology where you’re actually monitoring the physiological and psychological response to the device and you’re also monitoring the actual manipulation of the device itself in 3D space,” Mauro said.
From a design perspective, it converts traditional usability testing data into hard engineering terms. “This creates a much better link with the development team internally, said Mauro. “Instead of the human factors engineering research team presenting highly subjective (soft) data, the engineering and product and device development teams really adopt this new engineering-focused performance information much more directly.”
This new form of human factors research produces robust data on device design that provides clear direction for producing device enhancements that will benefit the user directly. When teams apply the more rigorous scientific approach, they can figure out which design aspects can be combined, enhanced, or discarded. The upshot is that developers end up with more precision, much better devices, and improved clinical outcomes. These new methods also create an opportunity for creating patentable innovations that relate specifically to human factors performance, a highly desirable benefit very difficult to achieve with traditional human factors research methods.
Drug and biologics
Mauro said even simplest syringes can be significantly improved. He also pointed out that the therapeutics within such devices can have an effect on usability. He noted that the new biologics coming down the pike are all based on extremely large molecules, which will require a different thought process in terms of syringe design. “When you have a very large molecule as the basis for your drug, you have very high viscosity and with high viscosity comes an entirely new set of complex human factors engineering problems such as a dramatic increase in forces required to deliver a dose effectively, extended drug delivery time and onset of hand tremors caused by high activation forces. As a result pain during injection goes way up.”
Mauro’s team recently completed a number of large studies for devices administering large molecule therapeutics. He said “It’s very clear there’s a big opportunity for the design of new prefilled syringes or new syringes that are properly human factored for these new drugs. Nothing on the market today that solves that [viscosity] problem.”
The demand for advanced human factors
As drug companies move away from simple auto-injectors and syringe-based drug delivery for large molecule biologics and develop new auto-injector style devices the need for advanced human factors research increases significantly. Mauro believes the new devices will require more advanced computer-based delivery mechanisms which will by necessity deliver to the patient visual and audio status indications that will produce much more complex device interfaces.
Such devices will place higher demands on patient populations with diminished capabilities. A new generation of on-body drug delivery devices introduces an entirely new set of complex usability problems. Because such devices must remain affixed to the patient for long periods of time the devices must keep the patient updated on drug delivery status and alert the patient when delivery has failed to meet clinical requirements.
“Only through the application of advanced human factors engineering design and testing methods can complex on-body devices be made truly useable for intended patient populations. The time to adopt advanced human factors research methods is now,” Mauro said.