Why haven't new models of primary care become mainstream already?

I recently read a nice piece about why CareMore Health and Iora and One Medical and all the other groups out there trying to reinvent office-based primary care haven’t “become the new normal.” CareMore was founded in the 90’s and has now ~130,000 patients and has changed hands multiple times most recently selling for $800M to Anthem. Iora has raised ~$225M and I haven’t been able to find how many practices they currently have. One Medical has raised ~$230M and has ~55 or so practices scattered throughout the country. I’ve written a ton about why digital health hasn’t taken off and become the new norm (here and here), but not so much about traditional old-fashioned primary care with a modern twist. I’ve also got a few years of experience and ~$40M of VC money spent building both brick and mortar and completely virtual new models of primary care. So here’s my best guess.

They’re glacially building Blockbuster in the era of hyper-scalable Netflix. Let’s say there’s ~125 brick and mortar practices between CareMore, Iora, and One Medical. Let’s say each practice can take care of 2,500 patients. If they were all at capacity, that would be 312,500 patients (this is a guess, not an actual number). And those patients would have to live and work within a 20-30 minute or less travel time to the brick and mortar practice. They are limited by geography, population concentration, and advertising to potential patients who live in those catchment areas. Growing from 0 to 312,500 via brick and mortar took ~25 years across the 3 of them. It’s delusional to think that growing from 312,500 to 330M will be what happens in the next 25 years. Something unprecedented and massive will have to appear out of thin air.

New models have focused on power-users, which are 5-10% of a potential population. New primary care models are heavy. There are care teams treating lots of people who fall on the sicker end of the spectrum. There are one hour appointments. You can get your COPD treated and do yoga all in the same place. They have to be heavy because they’re meant for very sick people, which is ~5 to 10% of our total population or 5-10% of employees who work in the same area as the employee on-site clinic. And if the service you offer is only meant for 5 to 10% of the population, 90-95% of people who experience the service will think it’s overkill and not for them. Also, I don’t know if the early adopter types think it’s cool to do yoga in a doctor’s office with a bunch of high medical needs people. So, the attempts to make these primary care facilities cool can fall flat. This limits their broad appeal.

Power users, by definition, are already plugged in to a whole traditional team of PCPs and specialists and that new primary care model has to steal them away from the devil they know. Many power users have shopped around and found a decent team they like and they’ll often treat all these visits as social opportunities to connect with their doctor and nurse friends. To these patients, it’s often more about the relationships with their “doctor friends” than it is about the disconnected, inefficient, services they’ve always used.

Unlike a product, a service is hard to standardize and it must be experienced. To experience one of these new services, it takes a leap of faith. Primary care is a human-powered service. These kinds of services are at the mercy of the diversity of human personalities. While you can standardize a process, you can’t standardize personalities. That being said, making a service enticing is the difference between the sexiness of Apple’s iPhone product page (tech specs are tech specs) vs. pick any of your favorite spa or hotel websites (oooh..."simple luxury"). It’s extremely hard to communicate the nuances of an experience that makes one spa fundamentally different from another and, honestly, that experience can immediately fall apart if one person on the team is having a bad day and takes it out on a guest. And since most people are used to primary care functioning more like the DMV, we’re all conditioned to be skeptical. 

Incumbents pay lip service to innovation, but actually antagonize it. There’s a ton of money and revenue in inefficiencies. When you can still bill for 99% of inappropriate visits or admissions, hospital networks don’t really want to get in bed with you. Insurers kind of do, but that’s not to pass on the savings to the customers, that’s just to maximize their profits. The whole point of an insurer is to perfectly predict next year’s expenses and adjust premiums accordingly. Innovation throws a wrench in their data-driven prediction models, especially if the innovation doesn’t do what it says it’s going to do. And as long as employers and the government continue to pay whatever they are billed, traditional fee-for-service care will be fiercely protected by hospitals and insurers.

New models are far more expensive for payors than the status quo. Kevin O’Leary wrote the definitive piece on this. I’ll let him summarize this: Traditional primary care costs $25 PMPM, new models cost somewhere around $50 to $75 PMPM. For insurers or employers to be interested they probably have to at least breakeven or demonstrate cost savings, so they need to demonstrate $25 — $50 PMPM of cost savings to break even, which is a large challenge.

There’s no funding model for this type of slow growth. Primary care is a ridiculously low margin business. And most don’t break even nowadays, especially after being bought by the local hospital network and turned into a referral base for all the far more lucrative specialist and imaging visits they get from having in-house primary care doctors. So, after taking in hundreds of millions of dollars of a mix of VC, then PE money, that’s a massive hole you’ve got to dig yourself out of. These new models of primary care haven’t typically chosen to go direct to consumer, so they must find employers or insurer partners to build these heavy brick and mortar spaces. Brick and mortar practices then become relevant to people who live or work within a convenient travel time away from these practices— think a catchment area of 10 miles or 20 minutes drive away. This then limits their usefulness to high needs people (5-10% of people) who live close to a brick and mortar practice who are convinced to switch their care to a new model. Typically, any human-powered service open to the general public takes about 18-24 months to reach capacity. This model applies to doctors, hair stylists, etc. There are things you can do to speed up this race toward capacity but there are also things that make getting to capacity harder. Given that the sales cycle into large employers (with enough employees to justify a brick and mortar practice for the 5-10%) and insurers is typically 2 to 3 years and capital runways between VC rounds is 18-24 months, this presents unique challenges. After Sherpaa’s VC-pocalypse in July 2016, one of our investors who recognized this unique challenge asked me how much time my co-founder and I spent appeasing our investors. His guess was 40% of our time was spent appeasing current investors and also hunting for and pitching new investors. I’d say that was accurate. When an entrepreneur is that distracted by ensuring a financial lifeline, that makes them ~40% less productive at actually building the business. But no matter what, the VC model mandates a 7-10 year 10x return, which is 3 to 4 healthcare sales cycles. The incentives are diametrically opposed to primary care reality.

There’s no clear target market and advertising is handicapped by employers or health plans. The iPhone launched with a very clear target— young, tech-savvy, wealthy folks with an insatiable interest in an increasingly mobile lifestyle. New primary care models target “Medicare Advantage” or “Boeing Employees.” Think about what would have happened if the iPhone would have launched to a subset of companies Apple convinced via conversations with HR directors to let them market to their employees. This leaves new primary care services at the mercy of corporate or Medicare plan communications, which is hardly sexy.

New models must use tech to streamline operations, but not too much tech. Since new primary care models are “tech-enabled” to maximize efficiencies and streamline processes, targeting the Medicare population doesn’t enable you to use a ton of sexy tech, because, well, Medicare folks aren’t known for their tech savviness. So, you can only ask so much of them which will likely result in handicapping your service to fit the tech ability of your target market.

It’s about politics, not about good ideas. Good ideas in healthcare are a dime a dozen, hence the massive influx of VC money into healthcare. But throwing money at good ideas diametrically opposed to the business model of healthcare and the incumbents hell bent on increasing their revenue to appease their shareholders is a recipe for a massive amount of acquihires and wind downs over the next few years. The federal politics of healthcare prevent healthcare business model innovation. Well, not so much as prevent, but allow enough of it to happen to ensure it doesn’t get too big, hence the o to 312,500 in 25 years.

Patients don’t value what innovators value. Most healthcare innovations are top-down. Patients don’t create innovative healthcare services. Innovative doctors do. No other industry uses video chats with strangers to solve problems. But healthcare entrepreneurs created them because it wasn’t so much of a stretch to have a 10 minute transaction in an office vs. via a video. It was the classic “look for a reimbursement code or an established concept and create a incremental improvement business around it.” Who cares if this innovation is actually what consumers want or is based on how they currently behave? And if these new primary care models delivered a service valued by 330M Americans, you would think they would have expanded like wild fire. But maybe these new models only jive with a very small subset of patients? Maybe the vast majority of patients don’t want such intense care and relationships? Even the sicker ones don’t want that? Which brings me to my final point…

Where has an innovation in primary care spread like wildfire? By far, this is Urgent Care. From 0 to 7,000 in the last 3 decades. Say each urgent care can do 50 visits a day, at capacity, these 7,000 centers can do ~128M visits per year. Out of the 1.1 billion doctor visits in America per year, I’m sorry to say it, but this is the new primary care. It’s not the primary care doctors and policy wonks define as primary care, but it’s primary care as defined by patients. They want accessible, on-demand, local, in-n-out, easy quick fixes that takes their insurance and they’ll even pay a slight premium for it because it’s cheaper than the ER. The masses don’t care about relationships or continuity or integrated care from teams. They haven’t had a primary care relationship like they had decades ago in decades. They just want the quick fix. The next phase of healthcare entrepreneurship will come from those entrepreneurs who straddle the line between giving the masses what they want (not what doctors want the masses to have) and discovering how to fit their innovation into insurance reimbursement.