Opinion

Beware Amazon and the other Sirens of Telehealth

Ido Schoenberg, M.D., the Co-founder and CEO of Amwell, said this recently about Amazon’s entry into the telehealth space:

"We're not going to opine about what Amazon is doing. I would just say welcome to the swamp...

We always talked about the fact that care is moving into the home, and you need a very deep degree of integration with devices, better analytics, new care plans, better integration with multidisciplinary team, deep understanding of the clinical flows and the financial flows and so on and so forth...

It's very hard. We spent a decade and a half building it. It's not the DNA of big tech. And big tech did not have a track record of doing that in world technology... 

So, we think that what they bring to certain employers could be very helpful, and yet does not compete directly with most of the value proposition that we offer at this time. When you come to see Converge, you'll realize the enormity of complexity that you need to truly serve complicated patients with multiple existing relationships and financial arrangements and regulations in their most convenient location, which is their home. We certainly totally agree with that notion. We just think it's really, really hard to create a solution."[1]

Where to start…

First, telehealth is not “very hard” in 2021. Telehealth was very hard in 1996, when a few of us were implementing teleradiology across the nation pushing CT scans over POTS (plain old telephone service) to deliver board-certified radiology reads in rural communities. Telehealth in the early days of AOL? Very, very hard.

“Hospital at home” was very hard in 2011, when a few of us tried to deploy at scale the concepts developed by Bruce Leff, M.D. at Johns Hopkins and successfully implemented by Presbyterian Healthcare Services.[2],[3] “Hospital at home”, which is what Dr. Schoenberg seems to reference, is still difficult, but much less so than it was in 2011. But telehealth is not hospital at home, and in 2021 telehealth is pretty simple.

In 1996, telehealth was difficult for three reasons: technology limitations, prohibitions against practice across state lines, and credentialling. Technology wasn’t a barrier before 2020, but customer adoption was a challenge. The pandemic demonstrated in a few days that millions of people could and would adapt to video technology for many needs, including health care. CMS eliminated long-standing (and anti-competitive) prohibitions against practice across state lines for Medicare patients in April 2020, and most states and payers quickly followed suit.[4] Today, the only significant challenge to telehealth is credentialling, which is an insurance-created “friction cost” that delays reimbursement, which is perhaps why it remains.

If the telehealth landscape is, as Dr. Schoenberg says, a “swamp”, then perhaps every health system should exercise caution when considering whether to invest significant resources in a telehealth strategy.

First, if a person didn’t try telehealth in 2020, then it is not clear what will motivate them to try it in 2021 and beyond. What is lost in press releases about the mergers and IPOs of telehealth companies and investor calls forecasting increasing demand is that approximately 290,000,000 Americans did not use telehealth in 2020.[5] Why? I don’t know all the reasons, but there are many reasons.

Second, telehealth has limited utility for most care delivery. Telehealth is amazing for behavioral health, highly efficient for radiologists, a viable alternative for some critical care medicine and home-based monitoring, and frequently useful for dermatology. Notably, health systems don’t generate margin on any of those services.

Third, and most importantly, telehealth is increasingly a commodity, which Amazon’s market entry demonstrates. Amazon’s core business model is to match demand with supply and then distribute that supply to the demand with ruthless efficiency and massive scale. Technology is not a problem for the company that created a multi-billion-dollar technology vertical – cloud computing – out of nothing. With CMS effectively eliminating the barrier of providers practicing across state lines, a centralized physician network can treat patients anywhere through technology and be reimbursed by the largest single payer in the US. As for credentialling, Amazon has a division called MTurk, “a crowdsourcing marketplace that makes it easier for individuals and businesses to outsource their processes and jobs to a distributed workforce who can perform these tasks virtually.”[6] Credentialling is a task that seems amenable to mechanical turks. What remains is matching supply – a network of technology savvy physicians, i.e., all of them – to demand – American consumers, 150,000,000 of which are already Amazon Prime members.[7],[8] Delivering a physician via a computer to chat with an Amazon Prime member will not be an inordinate challenge for a company that can deliver your groceries from Whole Foods two hours after you place the order.

It is also not a coincidence that large payers have entered the telehealth market directly. Although United and Cigna and Aetna lack the ability to match demand and supply as well as Amazon, they can unilaterally eliminate the friction cost of credentialling while using benefit design to steer their members to their captive platforms.

If the telehealth market is a “swamp” filled with Amazon and large payers, then how is your health system going to differentiate itself and compete? Why would you want to?

What separates hospitals from all other providers is the ability to deliver complex healthcare services, not commodity health services. Health systems don’t advertise their excellence in skin tag removal because hundreds of thousands of providers can deliver that care. Health systems market their excellence in complex clinical services because that is what they can uniquely do, which is why those services are marginally profitable.

Why are health systems throughout the U.S. following the siren song of telehealth? For some, because of the incessant media attention on telehealth; for others, because of spending too much time and money with consultants; for others, because of their fixation on their own internal growth in telehealth delivery from 2019 to 2020, aka the law of small numbers, confusing a market aberration caused by a global pandemic with a trend.

38,000,000 Americans used telehealth last year, but 292,000,000 did not. Do you want to focus your strategy on 12% of the population that used a commodity service (once or twice), or do you want to focus on the 100% of the population for whom Amazon will not provide cardiology or orthopedics or oncology care?

As you consider whether to invest limited financial resources in an increasingly commoditized market, ask yourself which consumers you are attracting. Are they the consumers your health system needs to win in the negative-sum game of the hospital business? Do you know who those prized consumers are? Do you know your true “share of care”?


[1] American Well Corporation (AMWL) Q4 2020 Earnings Call Transcript | The Motley Fool

[2] https://pubmed.ncbi.nlm.nih.gov/16330791/

[3] https://www.healthaffairs.org/doi/abs/10.1377/hlthaff.2011.1132

[4] https://telehealth.hhs.gov/providers/policy-changes-during-the-covid-19-public-health-emergency/telehealth-licensing-requirements-and-interstate-compacts/

[5] Trilliant data analysis, March 2021.

[6] https://www.mturk.com/

[7] https://blog.doximity.com/articles/physicians-iphones-a-data-love-story#:~:text=Over%2085%25%20of%20physicians%20are,that%20chooses%20Apple%20over%20Android.

[8] https://www.statista.com/statistics/504687/number-of-amazon-prime-subscription-households-usa/