Every day we enjoy the fruits of the tech revolution. On our recent family vacation to San Juan, Puerto Rico, we stayed in an Airbnb apartment, took Uber everywhere, shared photos in real time with family and friends, ordered in food on my phone when my daughter was ill one night, and entertained the kids on the long flight with apps and movies on our tablets.
But when it comes to healthcare, we call a receptionist to negotiate an appointment some number of months in the future; fill out countless forms to disclose our health history and medication usage (which, by the way, the doctor will ask about again, anyway); and carry silly, little paper insurance cards to convey our coverage information that almost never matches what’s in the “system.” (The pharmacist is confused every time I fill a prescription because my pharmacy benefit manager insists that I am female.) It makes my head hurt and makes me feel like I’m in—gasp—1990.
Where are the healthcare equivalents of Uber, Airbnb, Google and OpenTable to solve these mad inefficiencies? The reason that they don’t exist surely isn’t for lack of venture capital funding and bright ideas. A fascinating report by Rock Health shows that 7% of all VC funding in the U.S. is aimed at digital health. It’s bound to happen and bound to create some disruptions, but what’s taking so long?
I believe that there are five big reasons for the slow pace of tech disruption in healthcare. Understanding them can help smart companies navigate, mitigate or negotiate their impact.
- Cybersecurity: The first big barrier is cybersecurity. To people watching the industry, this isn’t a new threat, but it has gotten some recent attention. While the fictional vice president met an untimely end on the TV series Homeland a few years back after his cardiac device was hacked, the show has seemed pretty far-fetched until recently when a short seller claimed that St. Jude’s devices could be hacked in a similar fashion. Even if you question the motives of the short seller and its claim that all of these devices should be immediately explanted, there is something to this. Even monitoring applications like capturing and displaying blood glucose values on a smartphone can lead to dangerous consequences if the numbers have been tampered with.
- Liability and regulation: Speaking of dangerous consequences, there is a very murky landscape of what should be regulated and how. The FDA has issued guidelines on what is classified as a medical device (keeping in mind that software is a medical device under the right circumstances), which are pretty difficult to interpret, even though they try to be user-friendly given that digital health is a key priority. Another critical question for many solutions that would cross state or even sovereign borders is how to license healthcare providers. Here’s yet another challenging issue: If life sciences companies have the data on how their products are performing—or not performing—do they have an obligation to act or face liability considerations? Many pharmaceutical companies have been scared out of direct involvement in the space as a result. The startups within Silicon Valley simply lack the capability, governance and experience to deal with these hurdles. It’s now impossible to refrain from bringing up Theranos in this context.
- Privacy: A close relative of cybersecurity is privacy. HIPAA generally defines how we think about healthcare information privacy in the U.S. Nearly every visit to a new doctor or pharmacy requires new forms to sign. Substantial penalties await violators. But this sprawling act determines what can and cannot be done, very often creating large hurdles. In addition to being baffling for industry insiders, Silicon Valley has a reputation for playing fairly loose with their own definition of privacy. Wikipedia has a page dedicated to privacy issues with Facebook, which is possibly the longest entry on Wikipedia that I’ve ever seen, let alone read.
- Payer provider fragmentation: The team behind any potentially disruptive technology needs to understand funding flows. Simply put, digital health needs to figure out how to get paid. Many solutions would work best if the patient was paying, but they often aren’t. If the insurer is the payer, there’s a lot of work to do to find reimbursement. If the doctor still needs to be involved, as in prescription setting, then it can be difficult to disrupt the existing payment and incentive schemes. A provider must also contend with multiple platforms—as we see in diabetes with Diasend, Glooko, Tidepool and CareLink—since no single standard has yet been established. Even when a solution is particularly well-targeted and focused, there’s also the problem of market sizing, and because funding in healthcare is so famously short term in its thinking, the benefits need to accrue to the current payer, not at some abstract time in the distant future. My contention is that these sometimes perverse and always complicated funding flows also create complex and sometimes counterintuitive incentives.
- Proof: Eventually to buy into any solution, the payer will want to see the outcomes. Real-world evidence is a huge buzzword, but it’s still hard to come by. We have seen that many barriers still exist in outcomes-based contracting. While this is the latest thing that everyone wants to prove, many organizations are getting “pilot fatigue.” This means that many great ideas will need to wait in the queue.
Although the five barriers above are daunting, substantial investment continues, and I would contend that the big unicorn disruption is out there somewhere. Smart startups are finding ways to partner with established manufacturers or providers, tapping into cash and critical advice to navigate these unfamiliar waters. To many of the readers of this blog, this is the best news yet because it helps create a path to delivering the benefits to patients, while also emphasizing the critical role that established companies in the space can play to shepherd investment forward. The tech disruption in healthcare is long overdue, and hopefully not too far away.