Digital health - Strategy or Fantasy?

The worst person you can fool about your product’s value is yourself.

You’re a startup. You have a fantastic plan to use digital technology to solve a huge problem in the management of a given disease. The only thing stopping you is building it. You can hardly wait. You know it will work because you spoke to your data scientist co-founder, who assures you that the novel deep learning algorithm they lovingly crafted can classify patients with a precision never seen before. It makes sense that this would be best implemented as an app - after all, you want a quick route to market and you want your product in the hands of users to maximise the chance of improving health outcomes. Thankfully you know a bit about app design, you taught yourself Swift and you know your way around a wireframe. Bootstrapping is hard work, but this sacrifice now will pay off for you and your users in the future. 

So is it time to start building?  

No. No. No. No!

4x “no” because you’ve overlooked the four big risks of product development. The digital health startup space is awash with good ideas. Of course it is - as we careen towards the singularity it’s surely no trouble to drag healthcare along for the ride? But good ideas, like seeds, are cheap; it takes wisdom and persistence to actually start harvesting fruit. 

Lucky for you, Marty Cagan and the Silicon Valley Product Group have already distilled down the wisdom you need to avoid following your nose straight into a wall. It takes the heavy lifting of good product management to ensure you launch (and land) your product successfully, but you are very welcome to cut corners if you don’t want to succeed. Let’s now explore these timeless principles in the context of digital health.

Value risk

Beware the difference between “I really think this product is going to help users” and “I have evidence that users have problem x and would pay for solution y”. The assumption behind the first statement is naive at best, hubris at worst - both are lethal. 

Are you providing users (e.g. patients/clinicians) with something that will measurably benefit them, or that the payer (usually not the patient/clinician!) will want to purchase?

The value-based care equation comes in useful here:

Value = Health outcomes that matter to the person paying / Cost of delivering those outcomes

The worst person you can fool about your product’s value is yourself. So if the only metrics you can boast about are AUC and model accuracy, make sure you haven’t forgotten to assess how this is going to improve either concrete, measurable health outcomes, or the cost of delivering those outcomes. In a system like the NHS this can get very complicated, and this is why health economists are needed to assess the value of interventions that have diffuse impacts. 

Usability risk

I know you think that your product’s good, but what about your end users?

Even if you have developed a very powerful engine (which you assume will be an important differentiator from competition), if the car still hasn't been built you have no idea if people will like driving it. The same applies to digital health products: even if your model accuracy is unheard of, your end users might hate/ignore/misinterpret the overall product, meaning the real world impact will be zero (or worse, causing harm). 

In the parlance of medical device regulation, you would do well to adhere to the international standard IEC 62366-1, which details the importance of embedding design and user research into the assessment of the product performance within its intended use environment, not just in a cosy lab setting.

Whilst we’re at it with funky international standards, we should mention IEC 82304-1, which complements the “what you want to achieve digital product-wise” with “how you should approach the whole digital product software development life cycle” to show that your product is safe, from all points of view, not just usability. 

Feasibility 

Are you biting off more than you can chew? I’ll answer this question with two different focuses: project delivery and regulation.

Project delivery: Are there any assumptions/unknowns that could sink the entire project? Who knows, maybe it’s great that you want to reinvent the wheel, but if your product involves complex technical integrations across multiple convoluted user journeys then maybe your project is at risk of taking longer than you can easily estimate yourself. Early discussions with the people who will actually build and use your product are key to minimising this risk.

Regulation: It is incredibly easy to build a product, release it and iterate. But if you want to make one which diagnoses/treats/manages/predicts a disease process then stop! Simply to do the above you have to adhere to medical device regulations, GDPR and a handful of other delightful surprises - for instance, in Europe the idea of having the instructions for using your medical device on your website, so that your users can find it easily after their dog has eaten their paper copy, has a regulation all of its own. Make no mistake, regulation is a good thing for our collective health, safety, privacy and security. But when staring at the shining prize of being the next digital health unicorn, these are often afterthoughts. 

Business viability risk

At some point you will need to make some money from what you are doing. But the larger the company becomes, the greater the chance that the sales and business development teams don’t have a deep understanding of precisely what your product can and cannot do. At startups this is inevitable at some point, but beware, with medical devices there has to be evidence for every claim you want to make.

When you are a young digital health company trying to lock down some of your first sales it’s rare to have an off the shelf product ready to sell, so deals are made where both parties co-design and deliver the product. Great! You’ve got your first deal! Sadly you’ve also signed up for the threat of scope drift perpetually being around the corner. 

As your business evolves you will have very different aims from each of the different opportunities that come your way. Earlier on it might simply be credibility or a success story, whereas later it might be revenue critical for extending your runway. It takes good leadership (and well written SoWs/contracts), communication and clear understanding of both parties' different aims and definitions of success to maintain your business viability. 

Conclusion

Uncertainty is inevitable - the mistake is to either be paralysed by it, or to brazenly overlook it. Sadly, there are no simple answers here, but it helps to try to differentiate between known unknowns and unknown unknowns. 

Value risk is probably the most important nut to crack. Everything else can be solved by adjusting resources or efforts, but if you aren’t providing value then your house of cards will eventually tumble. 

Ultimately the above is a useful framework for de-risking your work. It helps prevent the kind of oversight that is very easy for people new to something like tech, digital health or medical devices. The different skills required to build these products are usually not all contained within one person, or even within a group of co-founders, so it’s important to understand where you might need to ask for help. 

Understanding your blind spots is not always the first place the mind goes when full of the hope and enthusiasm of refining an early idea, but this list should hopefully remind you of the need to zoom out and remind yourself to de-risk what you can.

Hardian Health is clinical digital consultancy focussed on leveraging technology into healthcare markets through clinical strategy, scientific validation, regulation, health economics and intellectual property.