Let’s look at some examples as my short answer above was rather abstract.
Example 1: App For Psychotherapy
Everyone seems to be developing apps for psychotherapy-like use cases nowadays, so this could be a widely applicable example. Assuming you are developing software which should treat a mental disease (e.g. depression, anxiety disorder, etc.), well, that’s clearly a medical device, because you’re doing therapy of disease, among other things.
Okay, so far so good. But if you change your content and description of your app from “treating depression” to “reducing stress”, then, suddenly, you’re no longer a medical device, because “stress” is not a disease (at least in the ICD-10 code.. I think?). In other words, you’re simply an app like all those meditation apps out there.
Obviously you shouldn’t hide things here – if you claim to only reduce stress, then all your content inside your app should mirror that message – you shouldn’t have content in your app which suddenly claims to treat depression (duh).
Now, does this even make sense if you have to change so much in your app? I’d argue that, yes, in many cases, this might make sense, because from a product perspective, you can test many things: You can test whether users generally like your app, whether they would be willing to pay for it if you offer additional purchases, and whether it actually delivers value (do people like it because it truly reduces their stress levels?).
All of these things are very hard to test when developing a medical device because you’re essentially blind until you release your first version, and that might take quite a while because you’re slowed down by regulatory compliance work. And, arguably, there’s an overlap between the lifestyle app and medical device app of things you might want to test: Do people regularly use these sort of apps or do they uninstall them? What amount of money are people willing to pay? etc.
Of course this doesn’t work for all scenarios. If your software is meant to treat people with schizophrenia, then, well, I don’t know how a watered-down lifestyle version of that might look like. Maybe you’re more creative than me here.
Example 2: Sleep Tracker
This is a really cool example. Tracking someone’s sleep with software does not automatically make your product a medical device. It depends on your intention behind that sleep tracking (your intended use, as regulatory people call it). If it’s for optimizing your performance as an athlete, it’s not a medical device. If it’s for improving your meditation practice, it’s also not a medical device. If it’s for detecting sleep apnea, well, then it’s a medical device (= diagnosis / monitoring of disease). You get the idea?
Now – if your software can detect sleep apnea, does that make it a medical device? No! It depends on what you display to the user. If you e.g. only display “sleep interruptions” to the user and don’t mention sleep apnea anywhere, then it’s still a lifestyle device. If you display “sleep apnea events” to the user, well, then you’re a medical device again.
Why all these specific examples? Because that’s pretty much what Emfit, a Finnish company, did, and it’s perfectly legal (as far as I can tell, not a lawyer, obviously): You could buy their “personal” sleep tracker for “hobby” use to track your sleep data (like for those quantified self people), or you could buy their “certified” sleep tracker for detecting sleep apnea events. As far as I could tell, the devices looked really, really similar, or were even identical – only the software would differ slightly for the certified version. The hobby version would not include any mention of sleep apnea.
(Their product lineup seems to have changed in the meantime – maybe there were hit by the MDR class I up-classification fiasco, no clue)
Conclusion
So there you have it. Is bringing a medical device to market as a lifestyle device always possible? No, definitely not. But, in many cases, you can bring a similar product to market to test your business and product assumptions, and that might provide your company with valuable user feedback.
Again, it’s all about not fulfilling the definition of a medical device.
One More Thing
Ah, and one more thing: Don’t do shady things like “our app treats depression but it states it’s not a medical device everywhere” – no, dude, just because you state it, your app won’t magically become a lifestyle device. It’s really about what it’s used for – that’s what counts.