We spoke to a veteran of the digital health space (20+ years), and most relevant to our discussion was his role at Amwell back in 2018-2019 (>12 months removed) and prior experience with telehealth/telemedicine startups. We came out of this conversation more impressed with the role Amwell's modules can play in helping providers launch telemedicine into health systems, specialty groups, and across care areas.
- A widening array of telemed offerings will eventually be consolidated in what should be a robust M&A environment over the next 12-24 months. For now, our contact sees Amwell as leading and Teladoc catching up.
- Integration with an electronic health/medical records system – whether Cerner, Epic, or another - allows for data availability when and where it’s needed (e.g., physicians have the data before a visit with a patient, and then the data/notes are captured back into the medical record after). The process is more robust than we previously thought.
- While at very early stages, remote patient monitoring (RPM), artificial intelligence (AI), and natural language processing (NLP) are maturing rapidly and will extend the physician's reach from the exam room, through the smartphone, and ultimately to passive monitoring and chronic care systems that prompt patients and providers.
- The value is in the data analysis, not in the devices, which are "commodities." If utilization of telehealth settles at ~20%, growth from there will hinge on the integration of all the above (AI, NLP, integration, etc.).
What have been the main drivers of growth for telehealth/telemedicine? How have you seen the space evolve?
- Going back a decade, when texting was just picking up steam (outside of youth/personal use), there were some early projects that demonstrated how helpful telemedicine could be for initial triage of patients; however, there was no reimbursement, so there was no volume/incentive to scale.
- Early telemedicine was focused on urgent care – niche plays around different conditions became a focus in ~2018. The pandemic has driven it, but reimbursement was a big headwind. CMS has been leading that charge. Where CMS goes, the commercial payer market will soon follow.
- The reimbursement changes started before the pandemic. When Trump came into office, there was a push by the VA to drive telemedicine… promote it. Around then, Medicare provided supplementary benefit (2018-19), and it became a basic benefit in 2020 (6-7 reimbursement codes covering it). The caveat back then was that the patient had to see the doctor first, then telemed could be used for follow-up (those restrictions were problematic, and then site of care lifted). Parity is the goal (virtual same as in-person reimbursement). The changes must stay in place.
- Behavioral health gets attention and focus, but women's health, pediatrics, dermatology, chronic conditions (Livongo and diabetes highlighted it), and others are big opportunities. Orthopedic and musculoskeletal - musculoskeletal issues = a top-3 cost - low back, joint pain, etc. - incidence peaks in the 40s, rolls in 60s and 70s. 40% of the population has lower back pain… these issues are more costly than diabetes or cancer. There’s a huge opportunity for PT and niche apps like SWORD Health, Kaia, and Hinge Health, as well as other specialty care (Hims, Roman, etc.).
How much of the growth sticks?
- What I saw pre-COVID was downward cost pressure and highly competitive situations. It was all about cost as there was not a lot of utilization. I was working with some big players, and utilization was low (average actual utilization of ~1% of the member population). Payers saw telehealth as something with potential, but utilization wasn't there.
- COVID -> utilization spiked to 40% on average (even higher in some places). I equate telemedicine today to where AI was 3-4 years ago. It’s the same thing right now. When a vaccine comes out, and the pandemic fades, we’ll see a drop in utilization, but rates will be higher overall because people understand it, like the convenience of virtual care, and the technology is getting better (we can do more with virtual visits across specialties).
- My best guess is that utilization will settle and hover at around 20% and then grow from there.
- I see a lot of strategic partnerships, and new players focused on healthcare and remote monitoring/telehealth. Best Buy is a great example. Partnership with Amwell, TytoCare, GreatCall. Best Buy’s strategy for growth involves a lot of health care. A lot of the basics of a visit - vitals - can be done virtually w/ a kit at home.
- Another factor is RPM, which used to be “clunky.” Active vs. passive - having to key in data vs. passive capture via a vest or other wearable makes a big difference. AI-enabled = triggers for prophylactic response. More offerings = more utilization, focus on chronic care.
What makes any of the incumbents special? Start with Amwell…
- If you ask Teladoc what they do, the answer is “telemedicine provider, focused on service w/ a deep clinical bench (they have great MDs).” They bought InTouch, and Livongo sort of fits with that, so they focus on behavioral health, urgent care, now diabetes... Amwell bought Avizia, which had acquired Carena. They are a “tech company” that developed a telemedicine platform.
- Carena was straight telemed, Avizia inpatient w/ charts - decided to merge inpatient/outpatient offering, and Amwell bought the whole thing (smart play to offer a broader offering).
- AMWL is very different than TDOC and even SOC Telemed. It wants to be a hub, agnostic about who the provider works with. The provider groups all sit on a platform, and AMWL supports the ecosystem. This is relevant when talking to a health system that has a vision of incorporating its own providers onto the platform. In contrast, MDLive and TDOC want to use their own groups. Amwell has LiveHealthOnline, a wholly-owned subsidiary, but it’s separate; Amwell is focused on tech - I’m pretty bullish on the hub strategy because more systems want to use their own MD groups.
- The future is around strategic ecosystems of partners that can be put together, a holistic care model. “The digital front door strategy.” Patients will look for digital options, and the health system’s portal must be able to handle a lot virtually; Amwell says they can put whatever the patient is looking for on the platform (i.e., it can handle policy questions, if someone wants a visit, has a skin issue, etc. - it’s all ON the platform).
- 8 out of 10 times, a virtual process can resolve an issue - for the other 2x need brick-and-mortar. Virtual care = a physician extender, and Amwell’s view is use our MDs, yours, or partners' MDs, whatever.
- Epic and Cerner are scrambling because they need telehealth. Amwell w/ Cerner is a good, expanding relationship. Epic leveraging MyChart. The whole point is getting that data into the system prior or after a visit so it can be updated and leveraged. Amwell doesn’t care whether it connects with MyChart or another system; the system knows to pull the “problem list” out of Epic to present to the MD prior to the consult. It's all about interoperability. As a consult wraps up, the notes go back into EHR. Modules designed for Neurologists, other specialists, etc. all work the same way.
- SOC Telemed - inpatient focus, around specialists. It’s a little different, addressing the clinician shortage out there. This is another topic to discuss - it’s not about creating more MDs. It’s about stealing mindshare from existing and shift from B&M -> virtual because you can do more virtual in a day vs. in-person (at lower cost). SOC Telemed is tapping into that in the inpatient market - they go head-to-head w/ Amwell in some areas with the carts/modules.
- Amwell stumbled years ago because the platform wasn't supporting scalability, which allowed others an opportunity. It’s no secret that they took $100MM from Google and are moving to the cloud there.
- In a REPEAT of AI 3-5 years ago, the bigger player acquire and the smaller players but fade away (if not acquired).
- I think Amwell can scale faster w/ a partnership model vs. a service-heavy model.
Can you share your thoughts on Teladoc & Livongo? Would welcome your opinion(s) about the combination, risks or opportunities you see?
- Teladoc talks about giving providers an opportunity to use their own MDs - they do offer that now but didn't previously. I think Amwell is better positioned, but TDOC can catch up.
Does Teladoc need in-person? Do you know ONEM - 1Life - thoughts on their model?
- I don't think TDOC needs it per se, but clients want and need it so they have to answer that.
- Look at Teladoc’s revenue model pre-COVID vs. post-COVID. They haven’t been showing growth from new accounts - it’s adding services to existing accounts. Start with basic at 8 cents PMPM, and after 6 months add behavioral for another 2 cents PMPM. That’s not ideal to pivot toward using a health system’s own doctors.
- Amwell says what the system wants - tech enablement fee for each visit... getting a similar revenue either model, and TDOC can pivot.
- The ONEM model is interesting. They were “a little” ahead of everyone else in the move to hybrid. Their market is younger and employed. ONEM becomes the primary for employees for the $200 PMPY plus whatever.
Best role(s) for AI and NLP/ML in health care? Any companies stand out to him? Nuance, Google, other? Are all the devices commodities...TytoCare, Masimo?
- Commodities? Yes. These devices help facilitate convenience and the move to passive, but pulse ox, pressure cuff, etc. are plentiful. Bluetooth, great. With chronic care, wear a vest or other wearable, and all the "stuff" is in one device. For polychronic Medicare patients, it’s important to consolidate the data into a flow that 's intelligently filtered. The amount of data is enormous.
- AI is great at the front-end of a Telemed visit - Amazon, Google - they are leapfrogging telemedicine - i.e., "did you really need that?”
- 98point6 is very interesting. Text-based, highly AI-enabled/automated. “Real” AI is complex (considers the list of questions, decides path and outcome based on all the words used, images provided, etc. Automated symptom checkers are great. I’ve used 98point6, and I could tell it was a chatbot. I don’t think that I ever got through to a live body, but it diagnosed poison ivy (via an image) and offered to send a prescription to my pharmacy. I think the interaction was 100% automated.
- Sensely - Avatar-based chatbot for a video visit, voice recognition. It’s very cool.
- Amwell - Front end AI - chatbot, symptom checker.
- Livongo - NOT SURE it’s really AI. They’ve been working on it for a while, but I don’t think it’s like 98point6’s.
What else can you tell us about the end markets?
- Teladoc’s focus on self-insured via channel brokers that position it as a telemedicine solution has slowed. The health plans are still engaged/buying, but not at the rate health systems are. I don’t think Amwell has signed a new payer account in two years (could be wrong, but the point is that this end market is saturated). Amwell has been “killing it” with health systems.
- The risk is being shifted down to the provider level - value-based care, and it's all about outcomes because of the impact on reimbursement (+ or negative).
- Humana is a great example - they’ve shifted from a pure health plan -> more of an integrated player like Kaiser or UPMC. They could gobble up a virtual asset - I could see more M&A in behavioral health to start.
How much runway is there for Amwell with health systems once they are “in” one?
- This is where the expansion is. There’s a long runway - it’s a significant opportunity because of all the different conditions Amwell can offer modules for. At some point, there won’t be a choice - i.e., orthopedics will have to have a telemedicine offering or their patients will be stolen. COVID pushed it. Adoption by new specialists can be 5% or lower if forces, or 20-30% if handled properly. For the most part, if reimbursement parity remains, it should be higher.
- Regarding Amwell’s claim that 55% of clients are using 3+ modules by year three, that seems high but could be the case.