#28 - Amazon’s Powerlifting Routine

Amazon’s unique strengths that will help it enter and win in healthcare

Things That Happened

IBM Watson is a piece of shit. You don’t have to trust me on this one. That’s according to one doctor at a Florida hospital where they’ve been using IBM’s tech for oncology decision support. Stat News reported this week that Watson often recommends unsafe and incorrect treatments, which may have to do with the fact that it wasn’t trained on real patient data, but instead hypothetical cases created by doctors.

The government did a thing that’s good for patients. CMS proposed a new rule that would normalize pricing for outpatient care, ending the current practice where care delivered in a hospital setting costs more that the exact same care delivered in a private practice. This will erase hundreds of millions in costs for both patients (in the form of reduced copays) and CMS. For an in-depth analysis from someone who knows a lot about this stuff, check out this tweet thread: 

Farzad Mostashari@Farzad_MD

It's strange that everyone on health policy twitter is not going apes over the Hospital Outpatient Prospective Payment System (#OPPS) rule released today

It contains one hugely significant proposal that hospitals will fight bitterly...but could actually help them in the long run

July 26, 2018
While you can criticize the administration for it’s whack-a-mole intimidation approach to regulating drug pricing, CMS is quietly doing some good things for patients. 

Google got a doctor. Dr. Toby Cosgrove, who formerly served as the CEO of the Cleveland Clinic will be serving as Executive Advisor for Google Cloud Healthcare & Life Sciences. Google Cloud is Alphabet’s answer to Amazon Web Services and Microsoft Azure, and this unit specifically targets bringing healthcare entities into the cloud and offering them additional data services like AI-based insights. And don’t forget, Microsoft got their own doctor a few weeks ago when they announced they’d hired Jim Weinstein, former CEO of Dartmouth-Hitchcock, to be their Head of Innovation and Health Equity for Azure. It seems Google Cloud is playing catchup with Microsoft in more ways than one. 

What does it say about this new era of tech and healthcare that the next career step for the CEO of a prestigious health system is working for a tech company? 

Things Worth Understanding

What is Amazon going to do in healthcare? Probably a little bit of everything. 

The hard thing about analyzing Amazon’s move into the space is that it’s happening very quickly and on many fronts. They’re an entertainment company, a consumer products company, a digital services provider, an online marketplace, an AI R&D shop, not to mention the single largest online retailer. The reasonable expectation is that they will push into healthcare through every one of these businesses, and add opportunistic healthcare plays as they arise (like the collaboration with JPMorgan and Berkshire Hathaway).

For more in-depth exploration of the effect they could have in each channel, check out Nikhil Krishnan’s recent briefing on the subject. 

Rather than delve into each vertical, I think it’s a better use of our time here to discuss Amazon’s overarching strengths and how they’ll help them in healthcare. 

Investor Trust

Amazon earned as much profit in its first 14 years combined as it did in the fourth quarter of 2017 alone, as Quartz pointed out. Jeff Bezos’ greatest trick was convincing Wall Street to be patient and accept a policy of reinvesting all operating margin into growth initiatives rather than realizing it as profits. And here’s how Bezos rewarded the street for their patience: 

If any big tech company has the ability to spend a long time unprofitably learning to operate in healthcare, and having big failures along the way, it’s Amazon. 

Consumer Trust

For consumers, Amazon is often in a league of its own when it comes to trust and reliability. This stems from an early realization Bezos had back when they were still only selling books. A publisher sent him an angry email because he saw bad reviews of one of his books on Amazon. Forced to defend the practice of letting users post reviews, Bezos reached a moment of clarity: “… we don’t make money when we sell things. We make money when we help customers make purchase decisions.” 

This ethos, as far as I’m concerned, is what has made Amazon the first-look location for all online purchases. Amazon has such a positive and trustworthy relationship with their customers, it makes other retailers look almost untrustworthy in comparison. 

As Amazon gets into more direct consumer healthcare areas like pharmacy (via their PillPack acquisition), consumer health information (via Alexa), and perhaps even benefits and insurance, they do so from a starting point of consumer trust. This already puts them ahead of all other players. 

Krishnan makes a very good point here: 

Nikhil Krishnan@nikillinit

I think an extension of this idea additionally is whether new companies can develop a consumer friendly brand within these legacy sectors faster than existing players can improve theirs and develop a consumer first ethos internallyhttps://t.co/kwSbhVjXzC

July 18, 2018
For example, is it easier for Amazon to build a health plan than it would be for a health plan to build the consumer trust Amazon already enjoys? 

We shouldn’t undervalue Amazon’s consumer-focused orientation and the special relationship it has with its customers. 

An Ecosystem

One benefit of having a constellation of products and services is that the value you derive from any one customer relationship grows with each new offering. 

Health insurance, as a business model, is undergoing a change. Companies are now forbidden from altering pricing or being selective based on risk, while the tools to predict future spending and care utilization are getting better. MLR rules also stipulate how much profit these firms can earn. As insurance increasingly becomes a more challenging way to create reliable profits, I believe it will become more of a channel for customer acquisition and retention. 

The Aetna-CVS merger creates a number of synergies for the two companies, one of which is certainly that Aetna policyholders will be given good reason to patronize CVS pharmacies where, in addition to prescriptions, they’ll purchase candy, use minute-clinics, and print their photos.

If Amazon were to build or acquire a health plan, they would stand to gain in much the same way, but on a vastly larger scale. In a nutshell, what if your Prime membership now included health coverage? 

Their Structure

Amazon is famous, organizationally, for its “two pizza teams.” This is the rule that ensures teams will never get so large they’ll need more than two pizzas between them for dinner. Benedict Evans, a partner at Andreessen Horowitz, writes

The obvious advantage of a small team is that you can do things quickly within the team, but the structural advantage of them, in Amazon at least (and in theory, at least) is that you can multiply them. You can add new product lines without adding new internal structure or direct reports, and you can add them without meetings and projects and process in the logistics and ecommerce platforms. You don’t (in theory) need to fly to Seattle and schedule a bunch of meetings to get people to implement support for launching make-up in Italy, or persuade anyone to add things to their roadmap.

This means not so much that products on Amazon are commodities (this is obvious) but that product categories on Amazon are commodities.

Amazon is a company built around innovation. This atomized structure, as Evans calls it, enables the Amazon organization to essentially act like a microcosm of the free market, where entire teams can be created to meet a new opportunity, and destroyed if it they flounder, all without upsetting the larger organization. 

This organizational framework will allow Amazon to experiment and deploy in healthcare until they find products and services that work. It’s only a matter of time. 

Things To Read (and Watch)

The story of how the robot came into healthcare is the story of what's wrong with medicine in America today,” says Dr. Martin Makary in Netflix’s new documentary The Bleeding Edge. He’s discussing the da Vinci Surgical Robot, one of the numerous medical devices the movie covers with a critical eye. The overall takeaway: innovation cannot be separated from quality, and outcomes matter more than technique. Oh, and it also paints a grim picture of how our regulatory structure for medical devices is failing. 

“…one in six Americans have past-due health care bills on their credit report, a debt totaling $81 billion in all,” Laura Santhanam writes for PBS, summarizing a forthcoming study from Health Affairs analyzing medical debt. While our conception of medical debt is that it can wipe out the uninsured when they have a catastrophic event, this study shows that 53% of outstanding bills are actually for $600 or less. While those with insurance are less likely to have medical debt, with rising deductibles it’s easy to rack up $600 in charges (in a single urgent care visit, for example) that the policyholder must pay in full. This sort of financial exposure for low-acuity care, as the story notes, discourages people from seeking care when they might need it. 

“If the industry does not behave properly, the government is going to step in,” says Leonard Schleifer, CEO of Regeneron Pharmaceuticals in a profile in Forbes. This story, from Matt Herper and Ellie Kincaid, is an interesting read if for no other reason than offering a pharma-eye view into drug pricing and its complex system that determines what you actually pay at the pharmacy. 

Startup Of The Week

Tel Aviv-based Healthy.io just received FDA clearance for its smartphone based Dip.io urinalysis test, which means you can now test your urine, with clinical-grade accuracy, from the comfort of your own home. 

The Opportunity

Urinalysis testing is a common diagnostic procedure. Pregnant women regularly test their urine to screen for gestational diabetes and preeclampsia. Hypertensive patients may screen their urine to assess kidney function and blood sugar levels, as would pre-diabetic and diabetic patients. Chronic kidney disease, which is asymptomatic in its early stages and causes heart disease in its later stages, afflicts an estimated 30 million Americans, 90% of whom don’t know they have it. CKD can be diagnosed using urinalysis. 

We’ve all done it; you pee in a cup, and then awkwardly hand it to a nurse. Behind the scenes, they’ll send it to a lab, or more likely just dip a coated stick into the sample, then move the stick over to a machine that reads the results. 

The Solution

Healthy.io’s new kit, Dip.io, lets you do all that from the comfort of your own home. No nurse required. The enabling piece of tech is the combination of a refined computer vision algorithm and the fancy-looking color board, pictured above. They can effectively color correct and boost the precision of your smartphone’s camera to produce clinical grade readings in a varied light conditions and across different lenses. 

The Prognosis 

Dip.io is already in-market in Israel and Europe. The company claims it’s on track to have 100,000 patients using the tool in the UK alone this year. 

There’s no doubt that an at-home urine test is more convenient for patients, more cost-effective for the system overall, and much more effective for broad-based screening. If the current purchasing behavior for urine tests was consumer-driven, Dip.io would be a sensation. But because physicians have to recommend the kits to their patients, Healthy.io is essentially in a B2B2C sales funnel. These can be tricky, especially when you’ve got a physician in the middle. If it doesn’t make their life especially easier, it’ll take some convincing to get them to change their existing behavior. 

That said, in the world of remote monitoring and diagnostics, this test has a better shot than most. As opposed to devices that spit out reams of data, urinalysis testing gives physicians a simple binary outcome; either things look normal, or they don’t and they need to follow up with the patient. There’s no issue around characterizing and contextualizing data. 

Dip.io is a slam-dunk tool for population health, especially where long-term costs are considered. Spending <$20 per patient to catch CKD early and mitigate the disastrously expensive ramifications of the later stages of the disease certainly has a strong ROI. Making it easier to perform regular urine testing will also improve adherence to treatment plans. 

As for the company, Healthy.io looks to have a bright future. At their core, they’re a computer vision company that’s pursuing a horizontal strategy within healthcare. But they’re also building vertically-integrated solutions - they didn’t develop a CV algorithm for reading urinalysis and sell it to Quest Diagnostics. No, they’re going to sell you the test, and use the data to further refine their algorithms. This will also help them as they move horizontally into the skin and wound care vertical, and beyond. 

Things I Listened To 

This week was very much an Arcade Fire week, with most of my listening going to 2017’s Everything Now (spotify)

Don’t Forget…

This machine is powered by sharing. If you found this week’s post informative or interesting, please do share it with a friend. I’ll try not to embarrass you.

Until next time,

-Isaac