Why Has DIGITAL HEALTH Been a DISAPPOINTMENT, and How Can It Be Fixed?

Digital health has, for 10 years, promised to turn health care into a data-first, low-cost industry worthy of the 21st century, acting as its supposed savior .

Consequently, since 2011, investors have invested in digital health 30 billion Dollars .

But after the past 10 years, what visible changes can we point to in health care costs or the average patient experience? Are there major companies capable of disrupting the market?.. According to Neal Khosla, founder and CEO of two digital health companies, Curai and Totemik, and according to us, no! Neal Khosla calls this “The Digital Health Conundrum” .

The following is a translation of Neal Khosla's cnbc.com article dated 15 December 2019, “Why digital health has been such a disappointment, and how to change that”:

First, let us consider the few digital health companies that could be considered successful.

Teladoc? A publicly listed telemedicine company? Adoption and use of virtual doctor visits remain below 10 percent. Despite marketing and rapid growth, Teladoc remains an infrequently used method of addressing urgent health issues. Nothing is wrong with that, but it is not transformative. Ask any insurance executive whether telehealth has affected insurance costs or patient access; not to spoil the surprise, but they will say no, and may even say it increases costs.

What other health company could qualify? Livongo? Livongo, which lost half of its 2 billion Dollar market value after going public in July 2019! It has only just crossed the 1 billion $ threshold. These companies pale beside massive software successes such as Stripe, Uber, Airbnb, Square, and others. They incrementally improved the efficiency of existing resources, but in my view they failed to effectively advance health care expertise . This is actually the most critical issue: clever software developers with bright ideas and deep-pocketed investors alone are not enough for health care's digital transformation. Health professionals are insufficiently involved, and medical informatics expertise is becoming increasingly important.

Perhaps Livongo is digital health's biggest success, but given the investment, we seem to need more industry change.

If health care is a huge market ripe for technology disruption, where are the winners? Why do the few successes offer incremental rather than transformative approaches?

Let us face reality: Digital health has been one of the most disappointing investment areas of the past 20 years. Venture capitalists still have not developed a strategy that works in health care. Technology entrepreneurs' approaches have failed to meet expectations.

Most fall into either technology for technology's sake —in other words, first-generation digital health ventures such as Google Health—or those overly focused on suggestions from health care people.

My favorite examples of the second group are companies working on artificial intelligence-assisted charting. Yes, for doctors, data visualization is a major problem. Eliminating it would be a major achievement. But that does not transform health care. In my view, it is putting a golden saddle on a donkey. Insider approaches offer incremental solutions; insiders do not create real innovation but point to what is known. Incremental solutions do not create venture-scale change.

This article concerns neither pharmaceutical companies nor biotechnology , but reimbursement organizations such as the Social Security Institution and the health care provider world.

Why Is Digital Health Not Succeeding?

The technology approach and venture model have been highly successful in digital health. What surprised me was the lack of original thinking about approaches to health care. Despite financial incentives to find home-run “startups”, venture capitalists want to return to predictable market strategies. A startup is the name given to new ventures and different projects with both growth and development characteristics. Very few people want to take risks in their approach, preferring painful sales processes and organizational resistance to change .

Venture capitalists avoid new strategies because they fear a 20 percent chance of change and an 80 percent chance of getting nothing. Why? Investors lack a healthy relationship with “risk” and want predictable progress. This is why they love the software-as-a-service (SaaS) business model: easy to understand and repeatable.

The most common reason technology people hesitate to enter health care is “learning the financial incentives.” That is not the key issue. Learning any player's financial incentives ignores that player's place in the rest of the system. Even within one organization, many players have their own incentives. Existing health care systems and organizations make go-to-market opportunities almost impossible.

The real challenges when working with existing health companies are:

1. Provider Systems

Provider systems, including physician offices and hospitals, have two main barriers to digital health adoption: misaligned financial incentives and a culture avoiding change.

The fee-for-service modelis considered the primary obstacle to digital transformation in health care, but is not the only misaligned financial incentive. Provider systems generate substantial revenue at thin margins. Median operating margins for health systems in 2018 were 1,7!

Imagine generating 1 billion $ in revenue and taking home 10 million $ profit. A startup approaches you wanting to change operational processes. Unless you are %100 certain it will succeed, why take the risk? These organizations have optimized their processes to the limit to achieve that % 1 margin. A 250.000 dollar contract significantly hurts that profit. There is no tolerance for trying new things, nor must there be. Health systems generally live month to month. Consider the list of the last 19 hospitals to close in the USA. Now try explaining why they should adopt your solution. Your solution's impact creates a dilemma: if too small, they have bigger fish to fry; if too large, it creates too much risk.

Health systems also have change-resistant cultures with complex internal relationships. Primary care physicians and specialist care staff have conflicting wishes. Medical informatics specialists and capital owners have conflicting priorities. Executives prioritize different things from both capital owners and informatics staff.

Someone from a large academic health system once told me, “If the faculty senate votes 100 yes and 1 no, we will cancel the project because of a tie,” he said. Because of their training, physicians avoid mistakes at all costs. What results? Stagnation. This change-averse culture spreads through health systems, as it should. On what planet could venture-scale change happen in such an environment? If you limit your product to administrative processes, you are better off, but affecting services is difficult.

2. Payers

Insurers are more open to change but difficult customers because they cannot change provider behavior, for example hospitals, lack direct patient relationships, and again have misaligned financial incentives.

One main problem with working with payers is their inability to influence provider behavior. Providers have little incentive to adopt anything new. This ends many promising businesses.

Call9 is one startup to consider. Call9 had a fantastic proposition: deliver telemedicine to patients in skilled nursing facilities to prevent unnecessary emergency department visits. The value proposition is enormous! Nurses assess the patient and contact the doctor. The nurse knows the patient's history. Avoiding an emergency visit can save tens of thousands of dollars, while telemedicine costs only a few hundred. In June, Call9 closed. Why? Among other reasons, despite savings for payers, finding skilled nurses and getting them into the system was difficult.

Even when companies generate revenue from payers, patients and people wishing to protect their health, they struggle with customer utilization. Teladoc markets its % 8 utilization rate as 4 times the industry standard. Imagine Slack boasting that %8 of its customers used its product!

Employees of a successful company selling to Self-Insured Employers told me they ran marketing campaigns for every new customer. This adoption challenge becomes layered risk alongside a long enterprise sales process and consumer marketing challenge. Health Care Consumers Are Hard to Reach: Insurance databases holding customer information are outdated, with incorrect addresses and contact details and duplicated/deleted patient records. Good luck!

Finally, among other factors, medical loss ratio restrictions create negative incentives for insurers to reduce costs. For more information, Sidney Primas's post explains why rising health care costs benefit insurers. Here is a clear explanation of the “80/20 rule”:

The 80/20 Rule of Health Insurance

Insurers may [legally] retain at most %20 of premiums for revenue and administrative expenses. Consequently, to increase revenue, premiums must increase. But they cannot simply raise premiums because insurers must also price them competitively. The solution is an increase in health care charges.

What Is Missing in Health Care Digitalization?

What exists in other technology-driven industries but is missing in health care?

In health care, go-lives, implementation periods, and pilots stall adoption. In contrast, the software-as-a-service world, easy to understand and repeatable, teaches us that the best software companies start self-serve. Self-serve software is never developed in health care. Why not? What is missing to implement self-serve or other forms enabling rapid progress? There are 3 things: A lack of early adopters, a lack of platforms for low-cost experiments, and integration difficulties.

1. Few small organizations adopt new products early. Health care consolidation worsens this. Surgical practices, where present, are not set up to adopt new software tools like small businesses. On the insurance side, the small early adopter is almost unimaginable: insurance is a scale business. Oscar Health and Clover Health may come closest, but they are still dealing with insurance blocking as they grow.
   
2. Continuity of care and integration into existing health systems are almost impossible. Few platforms allow health care “applications” to be built as we use APIs such as AWS, Plaid, and Stripe. This is changing through platform layers created for digital health applications by companies such as Eligible and TruePill. Of course, the biggest problem is interoperability of health data and it remains unsolved. This would enable data exchange between digital health applications and health systems. Integration requires long development processes because of unnecessary endpoints and reluctance around sharing data.
3. Distribution channels obstruct consumer-focused solutions. Enterprise players' incentives differ from users' incentives, whether users are patients or doctors. The result is slow user experiences that are never used and consequently low utilization. Health care needs applications developed with the consumer obsession seen in technology, rather than feature checklists designed for enterprise sales, as in electronic health record companies.

What are my takeaways? Everything in health care is messy, and digital health is no exception. Linear approaches do not bring change. We need radical new approaches or black-swan events. We need people who think differently. Is that not what technology has done best?

What Have We Learned from Our Failures? How Can We Improve Digital Health

For over three years, I have spent much of my time thinking about how to create major change in health care. I realized that initiating change requires us to think at a systemic level rather than focus on a particular player and its needs. Interactions between health care players have caused this approach to fail. We call it a health care system, yet solve individual small pieces.

Creative solutions to this problem exist. Perhaps the most successful go-to-market approach has been Self-Insured Employers (SIEs), pioneered by companies such as Grand Rounds and Livongo that recognized incentive alignment. This worked very well for a while, but as the market became saturated, we saw how difficult targeting an SIE as a customer had become.

Large employers “We will use data and technology to improve health care efficiency” have heard hundreds of times. CFOs and HR directors are inexperienced at analyzing health care costs and internal incentives to avoid paying for external revenue streams that could get them fired. The burden of saying “I will reduce costs” has become heavy. They ask third-party administrators to handle these problems, and then you continue dealing with large insurers.

Full stack approaches have also reduced the complex external revenue burden of incentives. In technology, “full stack” means developing the entire computer system or application, from user interface to database and internal architecture. Although early to say, some of health care's most groundbreaking companies will be these. Digital health companies such as Omada Health and One Medical overcame provider adoption challenges by hiring their own providers and selling directly to self-funded employers and insurers. One major challenge is simultaneously running an enterprise sales process and a consumer marketing task. More importantly, because of their enterprise focus, these companies advance by emphasizing sales and marketing rather than product.

Digital health has clearly also struggled to find a balance between virtualization and customer contact . One group focuses on efficient, low-cost services for low-complexity patients requiring no unusual care, such as direct-to-consumer companies Hims and Roman. Another focuses on entirely manual, expensive services for high-complexity patients with multiple chronic conditions requiring extensive treatment and management.

Consider the new class of primary care and medical management companies: ChenMed, Iora Health, and Landmark Health. Although among the most interesting in their fields, they succeeded by becoming more intensive and manual in health care workflows. Instead of reducing manual services, we are moving toward more! These companies deserve credit for innovative care models improving quality and cost, but while this is a step forward in one direction, it is a step backward in another.

Other full-stack companies, Virta Health, Omada, and others, are also largely manual. In my experience, the industry's secret is: Technology-enabled service means “we developed software that makes very little difference to the service”. They still have someone interacting directly with patients behind the scenes. Nearly every company does this. It is another form of rampant incrementalism. People have not charted a way to effectively change the underlying cost structure of care delivery. A fundamental question is: how do we turn health care into a product that can be mass-produced? This is the key to an equitable, efficient future economy: “improving services.” We must scale care effectively, especially for high-complexity patients needing greater access. We need medicines with near-zero marginal costs.

But there is a lesson: You can create change in health care if you are willing to think about how the pieces fit together. A fundamental misunderstanding underlies claims that health care is slow. Timelines do move slowly because of health care's local nature and natural pace of change. But when disruption comes, it comes fast. Can we find examples? Understand their underlying causes? Recreate these changes to move further forward?

1. Policy Changes

Back to reality. Policy changes can make enormous differences in health care, and market-disrupting companies have benefited. Good examples are electronic health record companies Epic Systems and Cerner. Cerner's market value is 20 billion $ , while privately held Epic is estimated at half of Cerner's value. Both have operated since 1979 but were not enormous until the last 10 years. This timing reflects electronic medical record industry growth over the past 10 years. Look at the chart of hospital technology adoption over 6 years. Enormous change.

Why did this happen? In 2008, the Obama administration enacted HITECH, providing financial incentives to provider groups using electronic medical records. Many companies turned health policy changes into major economic opportunities. Oscar Health and the many Medicare Advantage startups arising during Obamacare also deserve mention.

2. Existential Threats

If you do not want to wait for the federal government to understand things, good news: another way exists to change the system. I call these existential threats. Another way people built venture-oriented companies was by frightening the health care establishment. As mentioned, the system's incentive is to resist change, allow health care costs to increase, and profit over time. Thin-margin companies with large revenue streams resist change unless they encounter a black swan posing an existential threat, real or perceived, to their business. When they do, they move quickly. My favorite example is urgent care clinics.

Since the early 1990s, America has seen around 9,000 new standalone urgent care centers. How many Starbucks appeared in a similar period? About 14,000. Retail health care grew at 65-70 percent of coffee consumption. Let that sink in. It is extraordinary. How did it happen?

It began as incentive alignment: consumers and insurers loved urgent care convenience and cost. What sealed it was health systems realizing these small centers were emptying their own revenue-generating emergency rooms. Today, every major US health system has its own urgent care clinics or an external affiliation. When these players are frightened, they act.

The Quiz Show Champion's Health Care Failure

Not every example of fear-inducing action has produced successful products. The best example was IBM Watson. Watson put advertisements in New York subway stations claiming cancer diagnosis and treatment had changed forever. Combined with venture narratives, it became a perfect storm. Watson Health partnered with leading US oncology centers: Memorial Sloan Kettering, MD Anderson… MD Anderson invested 62 million $ in its collaboration. All collapsed when people realized Watson was marketing hype. I see Watson as a selfish, unethical, incompetently executed plunder that caused long-term technological skepticism among doctors and health systems.

Another good example is Intuitive Surgical. Intuitive has a market value of 60 billion $ and develops robots to improve surgical success. A wonderful narrative: robots will replace surgeons with robust motors and genuinely surgical precision. Intuitive built a giant by selling Da Vinci robots. The problem is that, apparently, they may not be so good. Intuitive's robots were found potentially to worsen surgical outcomes! (see the first comparison of robotic and open surgery in prostate cancer). They may have other important problems. And the problem is: Nobody cares! If a patient goes to UCF and asks, “Hello, I want this surgery at Stanford because they have those robots. Do you have them too?” and UCSF tries to explain, based on extensive data, that its surgeons are better, it is already over. The patient goes to Stanford. The narrative is captivating, and the threat to health companies forces adoption by health systems. In my view, what Intuitive and Watson share is that they are story-driven marketing machines presenting existential threats despite technology failing to meet expectations.

Where are some other existential threats seen in health care? One major area is direct-to-consumer companies such as Hims, Nuryx, and Roman. Hims went from 0 to 100 million $ in its first year selling erectile dysfunction and hair loss pills (Viagra comes to your door with a phone call). Such growth has been seen nowhere else in health care. Their future is uncertain, but they clearly cause disruption and fear and attract attention. Incumbents dislike this, raising arguments from “pill mills” to “patient safety” to “it is a toy.” Have you heard this before?

In many ways, health care is not very different from other industries in resisting digitalization. How do we scale these approaches to solve serious, complex health care problems? What can we borrow from them for large-scale adoption and patient empowerment?

Inflation and Incrementalism

A third route to changing systems has been offering inflationary or incremental things. Inflationary things work very well in health care. Reducing costs is difficult because you take someone else's revenue. When inflationary, everyone earns money, so people adopt.

Inflationism dominates health care. To sell to large health companies, you must spoon-feed them. They cannot understand grand technological visions, so take a simple familiar problem and solve it understandably. Isolating simple but material problems is difficult, but this is the best entry point. Unlike other fields, people have struggled to move beyond their initial landing spot. Patient Ping is an example: it grew by taking an isolated familiar problem and solving it simply.

Another way to simplify your story is to accumulate evidence that your solution works. Virta Health did this through published research. It is difficult and time-consuming but a way to present incremental, accessible solutions to large companies. Their timelines are long, and venture capital return cycles of 7-8 years do not allow a 2-year randomized controlled trial. Omada Health's approach may be the best middle ground: commercializing prior Diabetes Prevention Plan research by making it technological.

What Happens Next in Digital Health?

I am tired of hearing about long sales cycles and watching promising companies with good products die. I want people to think systematically about bringing change to systems that need it. Investors must start taking business model risk —if you agree health care's underlying business model does not work, why invest in things fitting the current system rather than things capable of changing it? Founders should aim bigger, for more impactful companies. You need a vision and awareness of the material gap your startup could create in a problem capable of bankrupting society.

Insurers and providers need sandboxes where small companies can test and prove solutions quickly. Know that most will fail. Focus on product development, not testing perfect products. Use your expanded service capital to open paths for powerful solutions outside traditional business models. This capital need not be large: %1 of your patients. Ensure projects have the right internal leadership and maintain speed. Enable companies to test and iterate quickly with accountability.

Politicians must truly understand the secondary effects of incentive systems they created and speak with people in the valley about how regulatory change drives technological change. Reimbursement changes for digital health tools, data interoperability, artificial intelligence deployment, and patient-first care approaches can all create new business models. Digital health legislation may be one area where we most need to spark change.

My journey to catalyze health care change has taken the last 3 and a half years of my life, and I clearly see us on the brink of great things. We must stop thinking incrementally and focusing on what already exists. Let us focus on what could be. Positive changes are coming to health care, and we must bring them to life.