“It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.”—Sherlock Holmes via Arthur Conan Doyle
It’s been a while since I sat and read a book cover to cover in one day, but yesterday was that day. I was reading a book called, The Remedy, by Thomas Goetz and it was un-put-downable. I particularly like non-fiction that tells a story so well that it feels like the best kind of novel. The Remedy definitely gave that sense while telling the true story of how scientists and physicians battled it out to prove that there was such a thing as bacteria and viruses; that those things could in fact cause contagious disease; and, that sometimes, even though starting with the best intentions, those same scientists can go off the reservation and create bad from good as their motivation evolves from discovery to fame and fortune.
Specifically the book tells the story of Dr. Robert Koch, one of the true pioneers of germ theory, discoverer of the Tuberculosis bacteria and the creator of the first attempted cure for the disease. It also tells the story of Louis Pasteur, Joseph Lister (after whom Listerine is named) and many other scientists, contemporaries of Koch, who added to the body of knowledge we all take for granted today. And lastly, it tells the story of Arthur Conan Doyle–yes that guy who wrote the Sherlock Holmes stories–but in this case his story is how Conan Doyle’s own medical background (he was a physician) and journalistic aspirations led him to identify that Koch’s cure for tuberculosis was a fraud. Notably all of this took place not even 150 years ago in the late 1800s.
What struck me about the book, aside from how interesting and fun it was to read, was how relevant so much of it is to what is going on today in the evolution of digital health, and particularly in the area of digital diagnostics and therapeutics. As Thomas Goetz writes, this time quoting Conan Doyle on the role of a physician, “Someone described our condition as that of a blind man with a club, who swung it at random. Sometimes he hit the disease and sometimes the patient.” Goetz amplifies on Conan Doyle’s delight in the adoption of the scientific method, just coming into use in his time, which could prove definitively the difference between correlation and causation and deliver meaning to patients. As Goetz characterizes Conan Doyle’s view he writes, “ Scientific medicine didn’t just offer a new way to look at disease, it offered doctors a weapon to combat it. Science made it possible to discern facts out of ambiguity…” and moreover solved the problem of patients knocking on the door “with the hope of a cure but never the expectation of one.”
One of the biggest challenges we have today in digital health is the absence of scientific proof that products do what they claim to do. There is an amazing abundance of discovery, invention, and marketing, but an entirely unsatisfying amount of evidence—the data that would prove the claims are fact-based. I see this almost every day in my consulting business where I spend a lot of time helping very large pharma, medtech and other healthcare companies that have lots of clinical assets and massive patient and provider customer sets find their way to the world of digital health. These companies, which are filled with smart, experienced and science-oriented people, keep hearing that they have to get with the digital health program or die, so they hire me to help them figure that out. But I can tell how ambivalent these large companies are about the exercise. They love the idea of the innovation but they recoil from the lack of rigor they see out there in the digital health world, a rigor that typically defines how their own organizations operate. The challenge becomes particularly poignant when faced with their partnering choices—too often the concepts logical for adoption are cool and interesting and even scalable and growing, but severely lacking in proof points. The questions I often get from my clients are these: How do we know this thing actually works? Where is the data? Where is the clinical trial? Where is the evidence?
And you know what? Those are great questions.
“It’s possible to see why the first wave of _____ theorists failed to convince their peers of the certainty of their ideas. In each of these cases the argument for ____ was based on association and proximity. Each proposed a theory, offered some evidence, and made his argument. But there were too many holes and gaps, too much theoretical hand-waving….It was too easy for the scientific status quo, not to mention the public at large, to reject the ____ theory outright.”
Kind of like Mad Libs, that sentence above had its blanks filled in with the word “germ/germs” by Thomas Goetz in The Remedy, but I could just as easily fill them with the words “digital health.”
I had an interesting debate this week with some one who started a sentence by saying that one of the great things about digital health is that it doesn’t require the kind of rigor that necessitates FDA oversight. Isn’t it great that the FDA doesn’t want to regulate much of what’s going on here? My reaction to that was actually quite the opposite, and will probably annoy many of my friends, but I feel that this is a real negative. I’m not saying I want to give the FDA free reign and make it hard and costly to get digital health products to market, but I’d be lying if I said it’s an entirely bad idea that the FDA oversee those products that make significant medical claims. If digital health companies are going to be delivering diagnostics and therapeutics that affect people’s medical condition, they should be treated just like the other guys who do the same, aka those companies who are my clients.
I wouldn’t want to be given an untested treatment and take the risk that it actually killed me not helped me, would you? That’s exactly what happened with Dr. Koch’s tuberculosis remedy in Goetz’s story, by the way. He was such an esteemed scientist with such an amazing track record of true discovery that when he brought his remedy to market people used it in droves. But sadly the treatment was worse than the disease and either did nothing or worse. And the reason it failed was in his haste to find a cure and his desire for global recognition, vanquishing his competitors and financial reward Koch discarded his usual scientific method and delivered a therapeutic built on assumptions and hype, not data. Too many digital health companies have so far taken a similar road and not done what it takes to demonstrate safety, efficacy and value. And while this may save time to market and money waiting for studies to prove out, I think this will really hurt those companies in the end.
One of the odd things about the field of digital health (at least compared to other parts of medical innovation) is that entrepreneurs and investors want it to behave more like tech than healthcare, meaning awareness and adoption at lightning speed, rapidly displacing all that has come before. And my guess is that this isn’t really going to happen here. As Goetz writes, quoting studies that show how medical advances take 17 years to spread to common practice and Everett Rogers specifically saying, “…a new innovation doesn’t simply erupt into widespread use all at once. Instead it gradually spreads through a professional class, as some adopt the new tool with enthusiasm, while more reluctant actors sit back and let their peers take the risks and work out the kinks. It’s a steady process, but it can be a slow one.”
The adoption of digital health products that have meaningful therapeutic and diagnostic value are probably going to follow this slow path while evidence is amassed proving it’s value. An advantage the digital health world has that products from the prior 150 years didn’t have is the advantage of social media and other means of communications that spread awareness of advances much faster. The spreading of that knowledge to the patient population is a very new thing and is itself a mixed blessing. Of course it is wonderful that patients are getting engaged and seeking whatever they can to help improve their own conditions. But when they see the slick stories in Tech Crunch and Goetz’ own Wired and on TV and Twitter and there is little data to back up claims, it is hard to separate the flash from the trash.
As Goetz says, “The laboratory is a long way from the doctor’s office.” And if digital health companies want to take Elon Musk-style rapid transit along that route, they are going to have to get comfortable with the scientific method. Pilots at a client here or there are only a piece of the proof. As history shows us, just because a few hopeful and forward-thinking risk-takers will try your remedy doesn’t mean it actually works. Real proof comes from well-designed studies that demonstrate clearly the impact of the new intervention in multiple settings used by multiple providers and patients.
We are starting to see the clinical community demand more rigor and there are definitely companies out there who get how important this is. While not a majority to be sure, numerous companies have spent the time undertaking serious clinical studies and seeking FDA approval for their digital health products; among these are AliveCor, Propeller Health, Proteus, CardioMEMS, Gauss Surgical, to name a few. But for every one of these there are hundreds of upstarts trying to figure out how to build their products so it is just short of needing that approval and so they can minimize the clinical evidence they need to start selling. I suspect their adoption curve is going to look a lot more like an adoption roller coaster and that they will burn through a lot of capital relying on hope as a strategy.
There is a recent explosion of energy from organizations willing to help digital health companies prove out their value propositions and that is great. UCSF’s Center for Digital Health Innovation and Partners’ Center for Connected Health are but two academic programs that have set themselves up to help digital health entrepreneurs know and love the scientific method. GE and Stanford have just co-created a for-profit company called Evidation Health that was established with a similar mission and there are others.
And thank goodness for that, because digital health holds incredible promise to expand access to care for more people, deliver it inexpensively and deliver it well. We as an industry must ensure that scientific rigor comes to this field so we can realize all of the good it has to offer and that it doesn’t get bogged down by hocus pocus and hype. As Goetz writes in The Remedy, Conan Doyle had watched the masses of people streaming into Berlin, where Koch had allegedly created a cure for tuberculosis and to him this demonstrated, “how much people now expected of science, how much the public was open to its power to improve their lives.” This is not only the same today, this desire and expectation and openness has only grown as consumers become actively engaged in their own healthcare. Let’s make sure our digital health companies give the people what they need, not just what they want.
Ps—To buy The Remedy on Amazon, click HERE; for a very good and differently focused review of The Remedy, click HERE.
PPS–The title of this blog post liberally borrowed from the lyrics of Jason Mraz’ song The Remedy
Rick Lee says
I refuse to accept your thesis that digital medicine should be measured by tried and true medical yardsticks, like the Random Controlled Trials. Rather, I support the word digital and the yardsticks for measuring success in a digital world, deployed by Google, Facebook and Apple – iterative, A/B testing. I truly believe that what the FDA would like to impose on digital medicine, the approach that sort of worked once for devices and pharmaceuticals, no longer works. It’s too slow, cumbersome, not in the real world and dinosaur-like. For instance, we cannot tolerate a case/control study that has some drivers texting while they drive, while others don’t. Compare and contrast the results. Change won’t allow us to put the genie back in the bottle – i.e., digital technology. Hence, we have to be more creative about the yardsticks we use to measure digital medicine.
Lisa Suennen says
Rick, I agreed there could be faster and better ways, but they should at least be valid! Lisa
Adam Chopko says
Nice post. A number of thoughts here, so I’ll try to keep it brief (ish):
1) I see a bit of a “chicken-and-egg” situation for funding. Investors are unlikely to pay for trials to prove the medical efficacy of a pre-launch smartphone app, so companies have to do what they can and iterate without trials and then, if proves necessary for the product to move into the mass-market, decide to fund a study using the money that they have presumably either earned or raised. (A company like Proteus where someone is swallowing a tiny sensor is very different than a common digital health smartphone app – scope of “digital health” here is broad). I support your push towards trials and the scientific method, but this must be financially viable. For the sake of innovation, I don’t want to see a world in which small companies in digital health must be acquired by larger companies pre-commercialization in order to fund trials.
2) What about having some type of optional FDA / Government Approval? More cautious consumers who want to wait-and-see can wait for the FDA to give a “stamp of approval”. Others can use the products pre-approval. Companies can decide if it is worth it or not to spend the time, money, and headaches getting this government nod. (Not talking here about invasive surgical devices or anything like that).