Today is my last official day under contract to Fosanis GmbH. I had my first encounter with the founders following my talk at the Digital Health Forum in March 2018. Following that initial meeting I became an advisor, writing a major funding proposal, bringing scientific techniques to the core of the product. In November 2018, following the closure of my own company, I became a full-time member of staff – as Head of Data Science – and led the project on the basis of the ideas contained in my funding proposal.
11 months is a long time in start-up time. The highlights, for me, have been:
- Hiring at least 14 people.
- Building a really great data science team.
- Seeing my ideas come to life.
- Working in a highly collaborative environment.
The team, in particular, was hugely satisfying for me. I was able to attract top-class performers and put them to work on a meaningful topic. Working with them is what I will miss the most.
In a very condensed period, we built a full data science infrastructure at the core of the company. I hope that the company will be able to leverage this architecture going forward.
So I’m leaving. I have spent almost 2 years mentoring other start-ups, advising on projects and generally working in the Berlin Digital Health scene. From that time, I have a number of observations, which are general and do not necessarily apply to my former employer, but I think they are a relevant summary of my time in this community.
The great white hopes for digital health in Germany are certain upcoming legal changes. The two opportunities are the hoped for appearance of an eHealth record, and the mandatory reimbursement of Apps for Patients. Every founder I know is salivating at this opportunity. I am rather more pessimistic.
I have talked with key decision makers, and they do not see the appearance of a realistic eHealth record in Germany any time in the next 20 years! Founders will focus on what they want to hear in order to keep their investors happy. But the cultural issues around privacy in Germany lead even the most highly educated into emotionally-charged rants opposing health record digitalisation.
Reimbursement of Apps is a more interesting issue. Germany has a history of reimbursing treatments which are medically non-recognised. So maybe they will reimburse Apps without further due dilligence. However, medical costs are already spiralling and uptake of digital solutions by consumers in Germany is on the low side by international norms – so I think that there will be less pressure from consumers, than expected, for reimbursement of these Apps. In the short term, government subsidies will be created to allow start-ups the breathing space to prove that their apps work. But validation will be the long term key to reimbursement.
In the greater question of whether digital pills will ever work, I am on the fence. In terms of absolutes, yes such a solution will eventually ‘work’. But what does work mean? Right now, we are automating away tasks which can easily, and cheaply, be performed by humans. This does not seem like a scalable use case. The constant argument that there are not enough expert practitioners (eg psychotherapists) is laughable – this is due to an existing unwillingness to reimburse these people, why should an app be any different? The idea that the app development is a one time cost which can be amortized over the next XXX years is not credible – any investor willing to believe this nonsense should contact me, as I would love to play with your money.
The idea of genuinely trying out solutions via digital means and seeing what works remains hugely attractive to me. However, (i) the Nudge consortium in the USA have had zero success with their approaches to date, and (ii) finding real solutions requires a lot of hard work, and this is something which most players in this space are not serious about engaging in. Ironically, it is the idealists, who lack funding, who are most interested in putting in the hard work, but they are not selling themselves so well to investors.
From my perspective, start-ups in the USA are leading the world in digital health. They are my final hope at this point. Developing a solution, at a minimum, requires insights into (i) algorithms/automation, (ii) design, (iii) human behaviour, (iv) medical reimbursement. I look forward to seeing what appears.
For the next few months, I will be focusing my energies on two projects. I am beginning a software project on Causal Inference. I hope to produce a basic libarary in the Julia language for interacting with data sets and constructing causal queries. This is a topic I’ve been researching for quite some time now, so I want to have a bit of time to play with the concepts before I move on to my next job. The other project is a paper on how to Develop and Validate AI/ML applications for Medicine. Based on our experience mentoring AI spin-outs, my coauthor and I already have a workable framework; now we want to write it up and share it with the world.
Since I am going to be asked.
1. I will be ’employed’ as a Visiting Researcher at Berlin Institute of Health for the coming months.
2. I am open to offers on what I will do next, but understand that I have already been approached with an offer that I Head a new Research Institute, so I’m not interested in start-ups which pay neither serious amounts of cash nor hard equity. #expectation-setting!