As someone who’s spent years working on usability in drug delivery, I’ve seen how easily “smart” can become complicated. Connected auto-injectors promise a lot, but are we designing for real people or just building what tech allows? This article challenges our assumptions and calls for putting patients back at the centre of our thinking.
They flash. They ping. They upload.
They connect to apps. They talk to clouds.
They promise to improve lives, one notification at a time.
In the race to innovate drug delivery, connected auto-injectors have become a symbol of progress. They promise better adherence, richer data and seamless integration with digital health systems.
But here’s the uncomfortable question we need to ask:
Are connected auto-injectors actually helping patients… or just helping us feel like we’re innovating?
Because while we build systems to monitor, track and optimise, many patients just want something that works – simply, intuitively, reliably – without feeling like they’re being monitored, gamified or judged.
This article is a challenge to our assumptions. It questions the idea that more tech equals better care. It draws from real-world usability studies to highlight a growing disconnect between the needs of real people and the digital ecosystems we’re designing around them.
If you’ve ever observed a patient with arthritic hands fumble with tiny Bluetooth pairing buttons…
If you’ve ever heard someone say, “I just want to take my medicine, not pair it with my phone”…
If you’ve ever witnessed the anxiety of a user, terrified of breaking expensive equipment they never asked for…
You already know where this is going.
So, let’s have the honest conversation:
Are we building for people… or platforms
Are we making things better… or just more complex?
This isn’t a call to blame tech and reject innovation. It’s a challenge to rethink what it’s for. Because the future of healthcare doesn’t need more connectivity. It needs more clarity, more simplicity and more confidence at the point of use.
It’s about asking better questions.
Like:
What problem are we really trying to solve?
If you’re up for a bit of constructive rebellion, read on.
The dream of the smart injector
Let’s start with the vision.
Smart injectors will solve adherence, or so we are told. They will ping your phone, notify your clinician, log every dose to the cloud and perhaps even sense your emotional state and recommend relaxing music.
These devices can track, time, optimise, perhaps even gamify the act of self-injection. Sounds Impressive. Until you ask a simple question:
Who is this actually for?
Are we solving real problems? Or are we designing clever technologies and then working backwards to justify them?
Because if you have ever observed a usability study with real patients, not fictional personas, not polished marketing videos, you will know something important: connectivity does not always help. And sometimes, it actively gets in the way.
“But it improves adherence” – Doesn’t it?
One of the most common arguments in favour of smart injectors is that they improve adherence. And yes, the data suggests connectivity can support improved adherence for some users, in some contexts, over a short time period.
But let’s examine that more closely.
Initial studies might show a boost in adherence in some cases quite significantly, but from a low base, in some case as low as 39%1. But how much of that is truly down to connectivity? Could it be the appeal of a new device? The smoother injection mechanism? The fact that the needle is hidden? The novelty of owning a high-tech product?
In my experience, most patients do not forget to take their medicine. Quite the opposite, many build routines, even rituals around it. A coffee first. A favourite chair. A moment to prepare mentally.
And sometimes, they choose not to inject. Because they feel fine. Because they are tired. Because, frankly, it gives them back a sense of agency.
We see this with inhalers. I have asthma. I know I should take my inhaler morning and night. But sometimes I do not. And it is sitting right there next to my toothbrush. I will spend two minutes brushing my teeth and then skip ten seconds of using the inhaler. If it had a flashing red light, I would simply turn it around so I can’t see it!
I am not proud of that, but I know I am not alone. Inhaled medication adherence is notoriously low. Injectable adherence is often not much better, and not for reasons related to forgetfulness. It is about anxiety, denial, fatigue, or simply being overwhelmed.
A push notification does not fix that.
Who are we designing for?
The tech confident early adopter from the lifestyle video? Or the single parent juggling three jobs? The older adult with arthritis and limited mobility? The person who never set up their smartphone and is afraid to touch it?
Some people enjoy the reassurance of connected devices. Others find them alienating or intrusive. I have interviewed patients who see their mobile phone as something they did not ask for, do not trust, and rarely use.
Even for those who are more digitally confident, the idea that doctors are reviewing your injection data in real time is unrealistic. Clinicians do not have the time, the interest or the capacity to monitor this data regularly. In fact they actively don’t want real-time alerts, because then they would have real-time responsibility! At best, it might be looked at every few months in the regular check-up. At worst, it is ignored.
So, all this data – who is it actually for? Maybe the Pharmaceutical companies or Healthcare providers, but not the patient. It’s great to support a clinical study, but this is not a patient benefit.
For many people, the core challenge is not recording the injection. It is getting the injection done in the first place.
Usability isn’t one-size-fits-all. For practical guidance on designing for a wider range of real-world users, read our white paper: Prioritising patient usability in home-use medical device design.
Simplicity as a strength, not a compromise
Here is where I believe we need to shift our thinking.
Electromechanical autoinjectors can be transformative, especially for users with reduced strength, visual impairment or reduced dexterity. But even these devices are not immune to the J curve of usability.
The idea is this: when we introduce a new technology, usability often gets worse before it gets better. We add complexity. Training. Interfaces. Connectivity. Before long, the simple “click and hold” of a classic manual autoinjector starts to feel like the good old days.
Ideally, we climb out of the dip and reach a new level of usability. But in reality, some devices never do. And the longer they remain stuck, the more they begin to look like a sunk cost.
What we should be aiming for is something intuitive, consistent, and easier than a manual device. No special skills. No charging. No pairing. No configuration. Just something that works.
It sounds idealistic. But it is not impossible.
We just need to start with the user, not the technology.
Are we designing for the patient or the platform?
As designers and developers, we risk becoming locked into technology ecosystems. We optimise for Bluetooth range, app functionality, screen size, cloud compatibility.
But ask yourself: has the shape of the human hand changed recently? Are patients asking for more ecosystems?
We often assume that syringes are effective because we are familiar with them. But for many users, they are not. We are drawn to miniaturisation because it sounds impressive, but larger, more robust devices often perform better in usability studies, especially in home settings2.
And then there is longevity. These products are often designed to last 10 to 20 years, which is an eternity in technology terms.
In 2005, we had Nokias and iPods. Fast forward and those ecosystems have disappeared. USB ports have changed. Wireless standards have evolved. AI has emerged at astonishing speed.
Meanwhile, your connected injector is supposed to stay current? Sure, it works today on the current Phone OS and Bluetooth version, but how do we push the upgrades for this in ten years’ time? Or do we just accept the lifespan of our devices now are measured in consumer driven lifecycles?
We need to acknowledge that this is not sustainable, neither technologically nor environmentally.
Understanding real world use
We often say that Human Factors should lead the way. But in practice, it is frequently squeezed between technical constraints and marketing ambitions.
If we start with real users – older adults, carers, people with low literacy or impaired mobility – we find very different priorities.
I have seen countless requirements documents insisting on the smallest and lightest designs, usually from marketing teams. But usability studies rarely place size at the top of the list. For home use, larger and heavier can actually improve handling confidence.
The priority should be the hand using the device and where on the body it is placed.
Performance should match usability, not the other way around.
The AI wildcard
No conversation today is complete without AI.
We are told it will revolutionise drug delivery. Virtual coaches. Predictive reminders. Personalised feedback. Dosing advice. Symptom tracking. In theory, AI will be a digital nurse in your pocket.
But it also introduces serious questions, like what happens when it gets it wrong? Who is responsible?
Would you change your routine based on an AI suggestion? Would your clinician back up a recommendation they cannot see or understand?
This is the danger of the black box – systems we trust, but do not fully grasp. In medicine, that is not just a technical issue. It is a human one. And there is a risk that we create systems that disempower users rather than support them.
Part of the answer might be transparency. Maybe using AI in ways that lead us to trust the initiative it is taking; we are not there yet!
AI can be part of the solution – but it must be understandable, explainable and maybe optional. Because the smartest injector in the world still needs one critical component: a person willing and able to use it.
So, what should we do?
Okay, I’m being deliberately provocative, there are a lot of great smart auto-injectors on the market. My point is that the aspects of them that are brilliant are the things that work for the user benefit, and this is where our focus should be. As someone working in design and usability, I don’t want to confuse smart with connected! I also want to ensure we are taking all the users along for the ride and not decide some people are outliers because they can’t use modern technology.
Here is my challenge to the industry:
Let us not be seduced by shiny features. Let us be obsessed with usefulness.
Before we add another chip, another app, another update, ask: What problem are we solving?
Let’s return to the fundamentals:
Usability. Simplicity. Reliability. Empathy.
Let’s design for people, not platforms. Let us prioritise clarity over cleverness. And let us recognise that the most sustainable, usable device may be the one that does less, not more.
Final thoughts
To finish:
- Smart can be brilliant, but only when it is solving the right problem
- Complexity should have to earn its place
- Real users are messier and more diverse than our assumptions
- The future will move faster than we think, so we must design with humility
Let us not confuse innovation with complication. Sometimes, the smartest device is the one that does not need to be smart at all.
I would love to hear your thoughts, especially if you disagree! Because that is how we move the conversation forward.
References
- Portaccio E, Zipoli V, Siracusa G, Sorbi S, Amoto M. Long term Adherence to interferon ß therapy in relapsing-remitting multiple sclerosis. Eur Neurol. 2008;59:131-5
- Antalfy A, Berman K, Everitt C, Alten R, Latymer M, Godfrey C. The Adherence and Outcomes Benefits of Using a Connected, Reusable Auto-Injector for Self-Injecting Biologics: A Narrative Review. 2023 Adv Ther (2023) 40:4758-4776