Adults over 65 use more medical devices than any other group in the country, and they are the group device design most often fails. Not because designers are careless, but because devices are designed and tested by people in their thirties who can read six-point type, open a foil blister with their fingernails, and remember a four-step priming sequence a week after being shown it once.
The result is predictable: devices that work perfectly in a lab and get used wrong at home. Missed doses, unheard alarms, a cuff applied upside down. Those are not inconveniences — they are use errors, a recognized source of patient harm that regulators expect you to design against and prove you have addressed.
What actually changes with age
Designing for older adults starts with specifics, not sympathy. These are the changes that most often decide whether a device is usable.
| Change | Typical effect | Design consequence |
|---|---|---|
| Presbyopia and reduced contrast sensitivity | Needs 2–3 times more light; struggles with low-contrast text | Larger type, high contrast, no gray-on-gray, no glossy screens |
| Lens yellowing | Blue and violet hues darken and desaturate | Never code information with blue vs green alone |
| High-frequency hearing loss | Alarms above about 2 kHz become inaudible | Lower-frequency or multi-tone alarms; add visual and tactile cues |
| Reduced grip and pinch strength | Peak pinch can drop by a third or more | Lower actuation forces, larger grip surfaces, no twist-and-pull combinations |
| Reduced fine dexterity, tremor, arthritis | Small targets missed, small parts dropped | Bigger controls, wider spacing, nothing that requires a fingernail |
| Slower working memory and task switching | Multi-step sequences fail under stress | Fewer steps, visible state, forgiving error recovery |
These are averages across a heterogeneous group — a 70-year-old may be a marathon runner or may have advanced arthritis, and your intended use population probably includes both. Design for the harder case.
Design rules that follow
Make the state visible without a screen
The most valuable property in a home device for an older user is that its status is obvious across the room. Is it on? Is it charged? Did the dose go in? A physical indicator — a window showing plunger position, a mechanical flag, a large illuminated ring — beats a menu on a small display. Every question answered by navigating is one they may answer wrong.
Design forces, not just shapes
Specify the actual force to open a cap, press a button, or connect a tube, and test it against realistic strength values rather than your own hands. A useful working target for a control an older user must operate reliably is well under 10 newtons for a button and under 20 newtons for a cap or a connector. Avoid compound motions entirely — push-and-twist child-resistant closures are a genuine barrier for arthritic hands, and if you are required to have one, design the rest of the interaction to compensate.
Alarms that are actually perceived
A loud beep at 3 kHz is inaudible to a substantial share of the population you are designing for. Use lower fundamental frequencies, harmonically rich tones rather than pure sine beeps, repeated patterns rather than a single event, and always pair sound with a visible and where possible tactile indication. Consider whether a caregiver or a phone needs to be alerted too.
Legible labeling
Type size, contrast ratio, and matte surfaces matter more than typeface choice. Avoid text molded into plastic with no contrast, which is invisible to anyone with reduced acuity, and make sure safety information survives cleaning. The regulatory requirements are in FDA medical device labeling requirements, but meeting them legally and being readable are two different tests.
Errors should be impossible, not just discouraged
Where a wrong action causes harm, use physical prevention rather than instruction. Asymmetric connectors that only mate one way, a cuff that cannot be fastened backward, a cartridge that will not seat if it is the wrong one, a lid that cannot close unless the part underneath is correctly placed. This is the strongest available risk control and it is exactly what a reviewer wants to see, because instructions and training are the weakest controls under ISO 14971 risk management.
The instructions nobody reads
Assume the manual is read once, badly, or not at all, and that the person who read it is not the one using the device six months later. Practical consequences:
- Put the critical instruction on the device. A short, permanent label with the two steps that matter beats forty pages in a drawer.
- Use a quick-reference card with large type and images, designed to be kept next to the device or stuck on a refrigerator.
- Write at a sixth-to-eighth grade reading level, in short sentences, with one action per numbered step and an image for each.
- Show the failure cases. What a correct and an incorrect setup look like, side by side, is worth more than a paragraph of prose.
How to structure the full document is covered in how to write instructions for use. Note that IFUs for home-use devices are themselves validated in usability testing — if participants cannot follow them, that is a finding you have to act on.
Design for the caregiver as well
Many devices used by older adults are set up, maintained, or monitored by somebody else: an adult child, a home health aide, a spouse. That second person has different needs — they want to confirm the device was used, change settings, and troubleshoot remotely — and they are often the one who decides which product gets bought.
Practically, separate the everyday user interface from setup and configuration, so daily interaction stays radically simple while the caregiver gets access to more. An adherence view or an alert to a caregiver's phone can be the feature that sells the device — though a clinical alert claim changes your regulatory position. The general design context for the home setting is in home-use medical devices, and the broader consumer version of this market is covered in product development for older adults.
Test with actual older adults
This is where programs get caught. Human factors validation for a device used by older adults must include participants who are genuinely representative — not staff, not the founder's healthy 68-year-old friend who happens to be an engineer. If your intended use population includes people with arthritis, low vision, or mild cognitive impairment, your study population has to include them too.
What to run:
- Formative studies early and often. Five to eight participants per round, with rough prototypes. Change the design, run it again. This is where the value is.
- A summative validation study with the production-equivalent device and the final IFU, with participants drawn from each distinct user group, run in a realistic environment and without training beyond what a real user would get.
- Realistic conditions. Dim light, background noise, an armchair rather than a lab bench, their own reading glasses or none at hand.
Every use error and close call gets analyzed for potential harm and traced back into the risk file; the formal method is in usability engineering for medical devices. Budget genuinely for recruitment, which teams routinely underestimate.
One last note. Nothing in this article requires the device to look medical or institutional. Older adults reject products that announce frailty, and a device that gets hidden in a cupboard because it is embarrassing has failed as surely as one that is unreadable. Dignity is a design requirement, and it is compatible with every rule above; the wider case for that is made in inclusive design.
Projects House develops medical and home-use devices for older users, including the human factors work that has to hold up in a submission. Tell us who will be using your device through the contact form.