Those Digital Health IPOs—Flipping the Stack & Filling the Gap


I’ve been driven steadily nuts by a series of recent articles that are sort of describing what’s happening in health tech or (because the term won’t die) digital health, so I thought it was time for the definitive explanation. Yeah, yeah, humility ain’t my strong suit.

It won’t have escaped your attention that, after five years during which Castlight Health more or less single-handedly killed the IPO market for new health tech companies, suddenly in the middle of July 2019 we have three digital health companies going public. While Livongo, (FD-a THCB sponsor) Phreesia and Health Catalyst are all a little bit different, I’m going to use them to explain what the last decade of health tech evolution has meant.

Don’t get carried away by the precise details of the IPOs. Phressia is already out with a market cap of $845m. Yes, it’s true that none of the three are profitable yet, but they are all showing decent revenue growth at an annual run rate of $100m+ and Livongo in particular has been on a client acquisition and annual triple digit revenue growth tear. It’s also the newest of these companies, founded only in 2014, albeit by buying another company (EosHealth) founded in 2008 that had some of the tech they launched with. Going public doesn’t really mean that the health care market will swoon for them, nor that they are guaranteed to change the world. After all, as I pointed out in my recent somewhat (ok, very) cynical 12 rules for health tech startups, UnitedHealth Group has $250 Billion in revenue and doesn’t seem to be able to change the system. And anyone who remembers the eHealth bust of 2000-2002 knows that just because you get to the IPO, it’s no guarantee of success or even survival.

But just by virtue of making it this far and being around the 1/10th of 1% of health tech startups to make it to IPO, we can call all three a success. But what do they do?

They are all using new technologies to tackle longstanding health care problems.

  •  Phreesia gives provider organizations tablets which their patients use to fill in that clipboard information, pay their bills, and get to see a little (pharma-sponsored) health content.
  •  Health Catalyst delivers data warehousing and analytics for some of the biggest provider systems in the country. Its technology is delivered on-premise for enterprises but it’s increasingly moving to the cloud (which is more scalable and more profitable). Very unusually for a pure tech company Health Catalyst also goes at risk for its clients’ outcomes.
  •  Livongo helps people manage their chronic conditions (mostly diabetes, but also high blood pressure, obesity, and some mental health issues) delivering a combination of products like infomated glucose meters, services including coaching, and data analytics.

The type of problem that they’re individually going after tells you about the major problems in health care.

A. Clinical care delivery in the current system

First, there’s the mess that is clinical care delivery at the coalface. We just spent $40 odd billion of the Chinese taxpayers money on putting in EMRs. We’ve paved the cowpath. In fact we have created a hidebound referral structure that locks in place the dominance of the current delivery systems. That’s not to say that EMRs haven’t improved clinical care. I’d argue they have, even if they’ve driven clinicians crazy in the process, but they’ve cemented in place what we did, and made it harder for more innovative care patterns to be introduced. That has to change and it is changing in three main ways.

1) The data in the EMR is slowly being opened up via API access (FHIR, SMART on FHIR, TEFCA and all that), leading to the ability to use that data in new tools and services. More and more app stores and interfaces are being introduced, and more companies like Xealth and Unite.us are building access directly into the EMR workflow.

2) The second main trend is the need to create a way to incorporate more and more data that isn’t in the current clinical workflow. Phreesia sits in this space. In their case, they collect patient administrative information, patient surveys and pre- and post-visit information. This all ends up in the patient record. Phreesia also gets administrative data off paper and deals with payment. Finally it returns information back to the patient. All of this was previously done on paper, or not at all, and was done badly. Now this patient generated data, which will soon include more and more data generated outside the clinician’s office, will be part of the record. It will also improve administrative efficiency.

3) The third trend, is the ability to analyze this data to improve what we know and change workflows to improve outcomes. Health Catalyst, which started as a new type of data warehouse under the EMR, is now providing more and more analytics and, as I mentioned, is even going at risk for the resulting outcomes. They’re by no means alone, with startups like Qventus and Ayasdi using data to change workflow and clinical patterns across hospitals and systems.  It’s part of a much wider move to use data, analytics, AI and algorithms to understand what works and what doesn’t. Of course the big question is whether this will change outcomes and reduce costs. But irrespective of that, the availability of data will lead to much greater use of analytics across health care, and more and more venture dollars will be invested there.

B. The New New Thing

For those who remember Jim Clark and Healtheon, the similarity of today’s “New New Thing” to that of 20 years ago is that new players are trying to go around the system. Unlike 20 years ago, it’s not so much about putting an intermediary between the insurer and the provider, rather it’s attempting to get at patient care at the source. The source is of course chronic illness. Now Al Lewis may claim that, especially for the under-65 population, chronic illness isn’t the driver of costs  for inpatient care that you’ll hear about at conferences and from the CDC, but there’s no question that difference in cost between a controlled vs uncontrolled chronically ill patient is significant.

That means a raft of new service businesses incorporating devices, technology, coaching and analytics to try to track and change the behavior and hopefully the outcomes of those with chronic disease. That starts with diabetes, moving onto heart disease, high blood pressure, mental health and respiratory conditions (asthma and COPD). Dozens of companies are focusing on all of these and Livongo is squarely in this space.

Indu Subaiya and I have called this “flipping the stack.” Instead of starting with the care encounter and layering services and tech on top of that, this new approach is starting with technology (particularly at home tracking of the chronically ill), then layering on services, with face to face clinical interventions only being used when needed.

Dozens of companies are putting together this in-home layer and many more are coming in as IOT infomates the bathroom and the bedroom. But the one area has been a little separate is telehealth. That’s because it started as a substitute for minor acute care issues for healthy people, rather than a way to care for the chronically ill. But that’s already changing. Doctors on Demand now claims it does chronic care management, Teladoc has invested in coaching platform Vida, and I would-be very surprised if Livongo doesn’t bump up the acuity level that it can deal with—probably by  buying a telehealth service and partnering with a (or starting its own) medical group. Its keto-diet based competitor Virta, already has its own doctors—even if Livongo’s Glen Tullman is not a fan! (He probably likes his ice cream as much as I do…)

The end result is that Livongo is the first of a new type of care management company out of the gate. Don’t forget that Lee Shapiro and Glen Tullman bought dozens of companies while they were running Allscripts and they have already put together either external or internal tech services for diabetes, pre-diabetes, high blood pressure and mental health. They will certainly add technologies for tracking and monitoring, more behavior change tools, more telehealth services, and probably more pharmacy/medication tools and more home visits. Next of course is the move from focusing on the under-65 population to the really expensive folk in Medicare and Medicaid

So the real question that emerges is what is the future of health care delivery?  And who is in charge?

Right now 99% of care is delivered through traditional health care systems. They in turn are connected to their physical plant–hospitals and clinics. But everyone knows that the health system of the future will be much more about meeting patients where they are. Will the current players extend out to these new locations? Will specialist new companies like Livongo take that role? Or will the consumer tech giants that already access the home like Amazon, Google, Apple, Comcast et al end up delivering the devices tech and services for the chronically ill?

And of course what do the other giants, the insurers who are adding technology and delivery capability–notably United/Optum and CVS/Aetna–end up doing in-house and what do they outsource?

The IPOs this week are part of a significant shift in the health care ecosystem. Of course it doesn’t mean that Phreesia will integrate all patient data into the current delivery system, that Health Catalyst will revolutionize delivery system analytics, or that Livongo will change the location of care management. But these are core parts of the next generation of the health system, and by going public they are both signalling that potential to the market and putting themselves in position to be significant players in the future.

Matthew Holt is the publisher of THCB and co-founder of Health 2.0

7 replies »

  1. My dear friend, would you (and your editorial team) be interested in a rejoinder from one of those who — for reasons still misunderstood — have made positive unit economics a priority in building their health technology business?

  2. Think many/most of these startups are going to be either relegated to the small number of “simple” disease management topics or will ultimately find a role as providing inputs to human-driven processes. As an investor in one startup and a reviewer of others I consistently see two fatal flaws: tech folks are accustomed to operating in an environment where the structures and processes can be well-understood which doesn’t describe biology. We’re still trying to figure out the roles of signal processing and associated genetic mutations in cancer, for example. Next, the human factor is daunting; when I see claims about improving compliance for patients I always have to wonder what the persistence looks like. How many patients who actively engage with an app or some other disease management strategy are actually still engaged after one year? Two years? And is there a lingering improvement in outcomes even after engagement ends or does all the benefit just evaporate?

  3. If you’re short Castlight Al, why are you here and not on the beach? Unless you’re doing this from the beach!

  4. Hmmm…my name is being taken in vain for a change (that never happens…not). Here’s the scoopage. I went short Castlight the Monday after their IPO, though I could only find 1000 shares. I have no problem with “digital health” IPOs. What I do have a problem with is sketchy outcomes. Here are a few questions that Livongo has refused to answer about their own. Always, always insist on validation by the http://www.validationinstitute.com since they are the only people who seem to be able to spot questionable claims. https://validationinstitute.com/the-livongo-study-cora-pro-interpretation/