What's up doc? 4 key learnings from our Healthcare Innovation Tour
First mover attack
Like in a chess game, can't we beat the disease by thinking and acting it one step ahead?
During our most recent Innovation Tour, we had a meet-up with Riley Ennis the co-founder of Freenome. In March of this year, they secured a 65million dollar investment...basically showcasing that VCs know these guys are on to something. Not only have they found a new way to derive from simple blood test whether you might have cancer, they are also capable to predict which drug combination is likely to work against that particular cancer. Earlier detection of cancer and higher chances of surviving: finally, some new ways to fight one of the deadliest diseases on the planet.
Of course, it wouldn't be Silicon Valley if people wouldn't try to take this one step further, into the seemingly impossible: by trying to defeat “death”. I met with BioAge Labs and their mission is to basically prevent us from dying. Although, they are not there yet, of course, and are (for now) trying to prolong our life in a healthy way by 20%. They claim to have found a biomarker for aging allowing them to push back diseases like Alzheimer. The ethical question is of course...who wants to live forever?
My virtual coach Joe
Instead of focusing on people who are ill, why don't we try to prevent disease?
Currently, predictions say that about 75% of Americans will become obese by 2020. I spoke to Omada health which wants to enable people everywhere to live free from chronic diseases... Pretty cool mission, right? Their first battle is to defeat diabetes.
Instead of you working with a physical person who coaches you with your diet and whom you only see once a week, you get teamed up with a digital therapist and some other team mates. Every day you weigh yourself on a smart scale that automatically syncs with your account to help you track progress over time. The scale comes with an app where you can register every meal. There’s no counting of calories but, you simply fill in 2 questions.
- did you eat a small bowl, medium or did you have a 'craving' and had a giant, big bowl
- was that bowl super healthy, just regular stuff or maybe ...a special treat,
That way, you learn to understand your behavior and that is exactly the key in making someone changing their diet. Instead of 1 encounter with your physical coach to discuss your diet plan, there’s permanent monitoring - more than 30 times a week - as you fill in these questions over and over again. Based on all the data you provide to Omada health, you get individualized praises and comments. Being peered, you can also follow up the status of your teammates and get encouraged by them equally.
No cure no pay
Influencing the behavior of patients is key because studies show that the number one reason of failed treatment is that people are not taking their medication on time or are not using the medical devices correctly. So, what if we would pay our doctors, the drug companies based upon whether their drug or device is really working? No cure no pay. We would see a completely different focus. I know that the outcome of a cure is also very much dependent on whether you are using the medication well or the devices well. So how can you intervene as a doctor better during the treatment?
Aluna is a small start-up that delivers solutions for patients with asthma. While you blow into their connected device, the app shows you if you are doing it right or if you need to blow harder to get the right results. The device will provide feedback on how good you are using it, so you can be 100% sure of the results and based on that you can adapt the medication accordingly.
I understand that as a doctor you might not be in for this 'value based pricing' after all. What if you get all the ‘worst’ patients, those who simply will never follow your prescriptions rigorous? Through machine learning, Evidation helps with segmenting people into adoption categories, giving you an idea on how likely this patient will follow the procedures. Taking into account all the data a person produces you can predict how routinized a person is and whether treatment will have any effect. Very chaotic patients will need different communication methods then those who are extremely organized ones.
Me, myself and I... precision medicine
We all know that you and I are very different: we have a unique microbiome in our guts, we react differently to food, to drugs and yet we all receive the exact same treatments and pills when we are faced with a certain disease. Luckily, this is about to change. I met with several companies predicting which drugs or diet would be a better fit for you. Ubiome is absolutely fascinating. Through a DIY kit, you take a sample of your gut microbiome. And the returned results provide you with a unique comparison of your gut versus those of other people. Did you know your microbiome can tell you how likely you are to become addicted to drugs or what the side-effects of certain drugs might be?
Fighting against breast cancer, Ourotech is also a highly personalized company. Taking a sample of the tumor and using a machine learning algorithm, they are able to shrink the timeframe dramatically from almost 1 year to detect whether a certain drug cocktail is successfully fighting your cancer … to 7 days!
The ultimate example of precision medicine is of course just printing an individual medicine based to fight your disease. Sounds utopic, right? Well, not really. Autodesk which you probably all know from their AutoCAD software, now uses its engineering knowledge for designing and simulating proteins and analyzing the CTGA codes (codes out of which your DNA is build) needed to tackle the exact disease that is in your body. The product is not yet on the market for humans but they have been able to print a virus and cure a dog from cancer. What an exciting future!
Of course, we are not 100% there yet. There are different challenges we still face within health care.
Firstly, we will need to learn to trust the machine learning algorithms. Skepticism in health care is similar to the viral reactions when a Tesla or Uber autonomous vehicle causes an accident. We might have thousands of safe trips but if we fail just that one time we do not trust the machine learning algorithms any more. What is the cost of errors we are willing to accept as a society?
Secondly, the environment in which we need to train these machine learning algorithms is very complex. There are a lot of causalities in the full chain of how a disease came to cause. Let alone the legal, data quality and trustworthiness of the systems.
So, yes, there is still some way to go. But, judging from all the companies we visited during our Innovation Tour, we are getting closer by the day.