Skip to content

Is talking to software the next big step in healthcare delivery?

September 11, 2018

Provider to Patient Ratios by 2020

In my latest book, The New Mobile Age, we spend some time talking about the growing demand for care (we’re getting older and older people need more care) as compared to the flat supply of healthcare providers.  The outlook is bleak – we’re running out of young people to take care of old people. It is a global problem.

I can only think of one solution — to introduce more automation into the care delivery process. I’ve written about the need to adopt the use of technology to create one-to-many care delivery models in several recent blog posts looking at technology and the provider, voice technology, and connected health for our aging population.  It was also the subject of a recent TEDx Talk.

In this context, I’ve become very interested in chatbot technology and am wondering if it’s ready for prime time. Could we create a near-term future where your first interaction with the healthcare system is via an automated chatbot?  Are today’s chatbots up to the challenge?  Are we able to seamlessly escalate to a human being, if necessary, so you don’t feel trapped in chatbot hell?  Will individuals feel cared for and will we be able to spread our healthcare provider resources across larger populations of patients?

To achieve this, an individual’s conversation with a chatbot has to approach feeling human.  This is not a new concept.  In 1950, Alan Turing developed a test where a human being interacted with a computer and, in order to pass the test, the computer needed to seem human to the person.  Over the years, computers have gotten better and better. Alexa and Siri are ubiquitous now, but do they pass the test? Only in a very narrow way.

James Vlahos

James Vlahos is a journalist who writes about this topic and, in fact, created a ‘DadBot’ to immortalize his father’s life story before he passed away. You can read more about this in a very poignant story James wrote for Wired last year. He will be a keynote speaker at the Connected Health Conference on October 18 and I’m excited to hear what he has to say.  In February, he wrote another intriguing article for the magazine, about Amazon’s Alexa prize, which offered a significant sum of money to create a bot that could carry on a conversation for 20 minutes.  There was no winner, but I highly recommend reading the article, which illustrates the subtle ways our minds work and that, so far, no computer is able to think in such a nuanced, sophisticated way.

But, the functionality required in bots used for frontline healthcare interactions may not need to be as sophisticated as Amazon was shooting for. Several years ago, we collaborated with a computer scientist at Northeastern University (Tim Bickmore) to investigate whether a software bot could motivate people to be more physically active.  We called her ‘Karen the virtual coach’ and she had both visual (a cartoon-like persona) and auditory (an early speech-to-voice engine) components.  The paper is worth reading, but the bottom line is that we showed that folks who interacted with Karen three times per week were significantly more likely to achieve their exercise goals than those who did not. Interestingly enough, we found that participants either loved or hated Karen. At the end of the study, a few participants asked for her phone number as they contemplated inviting her on a date!

Our study was published in 2012, and in the last few years, a whole burgeoning industry of chatbots for health has sprung up.  It is still early going and, in my experience, none of these is quite ready for prime time.  Following is a brief review of several prominent companies working in this space.  I’m sure I missed some, so if your favorite health chatbot did not make the list, leave a comment or send me an email so I can be educated.

Conversa Health focuses on three areas: care management to deliver efficient chronic care management, decrease remissions and optimize pre- and post-surgical care; marketing/patient experience to increase patient acquisition, improve patient satisfaction, patient retention, and generate incremental patient visits; and improving patient responses.


XebraPro is a decision support tool for physicians, not to be confused with Xebra, which is a medical imaging software platform.  XebraPro, from Physician Cognition, has two versions, one for differential diagnosis and one called XebraED, an education tool. Unlike Conversa and some of the others, XebraPro is a tool to improve physician accuracy and efficiency.

Buoy, an online a symptom checker,  guides you through a series of questions and then recommends a course of action.  Their business model steers patients to a provider organization.  Using their site or app does not feel exactly like having a ‘chat’ but rather filling out a questionnaire.

Medumo focuses on care navigation and instructional support to help patients.  The primary delivery mechanism is text messaging or email, and the primary value proposition to the provider is getting patients to their appointments. Patients who don’t keep appointments generate a significant administrative burden and Medumo has set out to solve for that.  Mostly their communication is outbound to the patient, so it does not feel like a chat, per se.

Babylon Chatbot Interface

Babylon believes it is possible to put an accessible and affordable health service in the hands of every person on earth, using a combination of artificial intelligence and natural language processing.  They made instant headlines by winning a contract with the National Health Service of Great Britain. Their interface is very similar to Buoy, guiding an individual through a series of branched logic about their symptoms to a recommended course of action.

Lark combines artificial intelligence with behavior change design to create a scalable, personalized care management platform. Lark claims to be clinically validated to deliver positive health outcomes across each major chronic disease state, with its main virtual coaching platforms used for diabetes and hypertension.  The interface is a bit wooden in my experience, and there’s no attempt to disguise this to be anything but a dialogue with a software bot.  The bot asks questions and you choose various answers.  It seems overly simplistic to me.

Memora is an automated discharge chatbot that can also manage follow up paperwork, enabling a care team can focus on the patient instead of the details of discharge planning. Their virtual assistant is named Felix, and the idea is for Felix to have an identity.  The goal is threefold: to increase a care team’s capacity; rapidly identify high-risk patients, and deliver a world-class patient experience. It also streamlines time in front of a screen for the patient, promises 24/7 responsive team members, and an immediate start.

Tess by X2AI

X2AI seems like the most up-to-date version of AI of all of the technologies I have reviewed so far. With the goal of supplementing human therapists, researchers are instructed to teach empathy to artificially intelligent messaging tools. Their assistant is called Tess and she can schedule an appointment with a therapist, have a dialogue with a patient or connect the patient to one of the on-call therapists around the clock, for a small fee of $5 a month. Another useful feature is the nightly text encouragement from a therapist and suggestions that accompany each text. The goal of X2AI is to create a patient experience that includes interactions with Tess and some with people, with the software seamlessly triaging the interactions.

There is so much rapidly-moving innovation in this space — from companies with vastly different personas and consumer/patient interactivity — that we have to keep close tabs on it. From XebraPro, a diagnostic aid for the physician, to Buoy and Babylon, symptom checkers, to X2AI, a virtual mental health counselor.  They all share one thing in common, which is a backbone of artificial intelligence.

At some point, the first part of any healthcare journey will be interacting with software like this. In some cases, that’s all that will be required. In others, the interaction will move to a telemedicine solution with a provider and, in still other cases, the interaction will move to a face-to-face experience with a provider.

How far into the future is this?

Reimbursement Update: It’s raining new codes!

August 21, 2018

Reader alert: reimbursement codes are not exciting but are critical in advancing connected health… and, after all, we all want to advance connected health. Read on!

It is just shy of two years since the American Medical Association (AMA) asked me to co-chair their Digital Medicine Payment Advisory Group (DMPAG).  At the time, I was waking up to the idea that, for us to speed adoption of connected health in the provider community, we need more billing codes. This may sound obvious to all of you, but I was a late-comer to this party. My reasoning was that value-based reimbursement is growing and value-based contracts are very telehealth friendly, so why should I waste my time on bolstering the fee-for-service reimbursement system?

The insight I gained from this collaboration was beautifully articulated by the AMA’s CEO, Dr. James Madara, during an interview I did with him on my new podcast, Well/Connected (click here to listen to Jim talk about this directly). Jim said that no matter what reimbursement system you choose, providers need to have a way to quantify their work product and codes allow them to do that. Indeed this has helped me understand why adoption of remote monitoring for care of chronically ill patients has lagged. Until very recently, we had no mechanism for providers to document the level of work required to take care of patients in the context of remote monitoring.

2017 was the first year of DMPAG activity and, early on, we decided to move two new types of codes through the process, one for physician-to-physician online consultation and another set for remote monitoring.  I detailed this journey in a post in November 2017.

In that post, I had referenced the Lemony Snicket movie, “A Series of Unfortunate Events,” noting that Medicare reimbursement for remote monitoring was just the opposite: it was the culmination of aseries of fortunate events. Back in November, these new reimbursement codes had been moved forward to the point where they were sitting with The Centers for Medicare & Medicaid Services (CMS) committee that proposes a value for them (known as the RUC process). It’s old news now, but as of January 1, 2018, CMS also unbundled an old code – CPT 99091 – that allows a practitioner to bill for the evaluation of remote monitoring data.

I don’t have specific data on utilization of that code, but anecdotally, it has not been a big hit. It requires a doctor (or nurse practitioner) to review data for 30 minutes/month in order to submit the code.  In our experience, unless a patient is very sick and complex, no one provider devotes 30 minutes of time/month reviewing remote monitoring data.  This code was as step in the right direction but predictably not a big winner.

About a month ago, CMS came out with their guidance on what new codes they propose to include in the code structure starting in 2019.  Some of what they published was predictable, moving forward with the codes we submitted last fall, and some was unpredictable, with the addition of new codes. But as far as I can tell, all of it is positive!  CMS is now collecting comments on these proposed new codes and, once again in November, will release what is to be included in the 2019 code set.

So, I warned you.  I tried to think of a way to make this topic interesting, but there’s just no way around it, CPT codes are necessarily dry and technical. The exciting part, however, is the receptivity of CMS, under current leadership, to move connected health forward.

New and Exciting Proposed Reimbursement Codes!

Two of the new codes are different from CPT medical billing codes. So, without getting too deep into the weeds, following is a very brief explanation of these two new codes, known as HCPCS codes (Healthcare Common Procedure Coding System). If you want to learn more about the difference between CPT and HCPCS codes, follow this link.

New Proposed Code #1: Brief Communication Technology-based Service, e.g. Virtual Check-in (HCPCS code GVCI1)

This is to replace a code (99441) which covered a brief telephone check-in with the patient between office visits to determine if the patient needed to come in sooner or make another adjustment.  This is a new area for reimbursement and CMS is looking for lots of early stage guidance. It’s not clear what the fate of this code will be, but the fact that they proposed it is encouraging.

New Proposed Code #2: Remote Evaluation of Pre-Recorded Patient Information (HCPCS code GRAS1)

This is to cover remote evaluation of recorded video and/or images submitted by the patient (e.g., store and forward), including interpretation with verbal follow-up with the patient within 24 business hours.  This would seem to be a dream come true for my dermatologic colleagues (and others who evaluate images as part of their clinical decision making).  This is also early stage, but CMS suggested a value(0.8 RVU) for this code, leading me to believe they are serious about it.

Also of note is that the interprofessional Internet Consultation codes (CPT codes 994X6, 994X0, 99446, 99447, 99448 and 99449) were recently valued by the RUC, and CMS is seeking comment on some of the details.

Finally, there is good news for the home health industry, as CMS is proposing reimbursement of remote monitoring services in that sector as well.

Admittedly, all of this is a very superficial review, and for those of you interested in the details, start with the CMS announcement and work your way through it.

And even if you’re not interested in the fine print details of reimbursement, for telehealth adoption, the future looks quite bright, considering that, when CMS pays for services, it is common for other carriers to follow suit. It’s a reality I have come to appreciate – even in a value-based care delivery system, providers need to quantify their work and billing codes for connected health are a necessity.

And what does all this mean for connected health? In the next few years, as these reimbursement trends continue, we could finally begin to see the needed acceleration of telehealth adoption. And that is a very exciting proposition!




Aging and Technology: An Intimate Discussion with Thought Leaders

June 27, 2018

We all want to live a long and healthy life…

This was my opening statement at an event we co-hosted earlier this month.  We did so in collaboration with our friends at the Summit, after learning that they were bringing their yearly thought-provoking gathering on all things aging to Boston. When Ray Dorsey first brought me into the fold as a member of their advisory board, he said to me, “The ‘d’ is for disrupt, not digital,” which, in my opinion, best describes why this is a must-attend conference for anyone in or interested in the field.

So we wanted to do something collaborative with them; to welcome Ray, Anna Stevenson and company, and show our support in hopes that they’ll bring the Summit back to Boston in future years. For our part, we collaborated with Ray and his team, as well as the Personal Connected Health Alliance (PCHAlliance), to host a two-hour preconference event on June 5, the eve of the Summit.  We called it Aging and Technology: An Intimate Discussion with Thought Leaders. Included in the mix, besides yours truly, were friends Bruce Leff from Johns Hopkins, Charlotte Yeh from AARP, award-winning journalist Laura Landro and Ray himself.  Three of us gave brief talks (Bruce, Charlotte and me) followed by a panel discussion including Laura, moderated by Ray.

I enjoyed the event and gained many insights, so much so that I thought memorializing it in a post would enable others to learn too.

I was up first, and my talk covered themes from my new book,The New Mobile Age: How Technology will Extend the Healthspan and Optimize the Lifespan,as well as some insights I’ve already gained since we published it about six months ago.  I debuted my new stage buddy, the robotic cat (one of the Joy For All Companion Pets from Hasbro), which my daughters named Checkers. I covered many themes but, in an effort to be succinct:

  • We all want to live a long healthy life: ‘long’ as in longevity and we’ve done well there, adding 25 years to the lifespan in the past century; ‘healthy’ is where we have uneven results, that is the healthspan.
  • From a societal perspective, we need to better engage older individuals, providing them with opportunities to maintain a sense of purpose, social connections and increased physical activity. Technology can help with all of these.
  • For individuals, the decisions we make that lead to a shortened healthspan are a combination of data and emotions. Technology can help us with the data part of the equation.
  • Care-giving is in need of an overhaul, going from our current one-to-one model to a one-to-many model. Technology (robotics, artificial intelligence and other digital and emerging technologies) is critical.  I brought Checkers onstage to point out that, if we use technologies such as companion pets wisely, we can begin to attack this challenge.

Bruce Leff, a geriatrician and Director of the Center for Transformative Geriatric Research at Johns Hopkins,is a pioneer in bringing the hospital into the home. He has overseen research and implementation in this area for years and opened his presentation by telling a story about house calls when, as a medical student, he noticed how the relationship changes when a provider visits a patient in their home. You are a guest and behave differently and deliver more personalized care.

Bruce further asserted that the future of hospitals is specialized, acute care. (I agree, but we keep on building buildings. When will the tide turn?) Following that, he suggested that a supply chain for services in the home needs to be developed, as home-based care will become part of the US healthcare system, likely within the next 25 years, and hospitals will eventually turn into large ICUs.

Charlotte Yeh, Chief Medical Officer for AARP Services, is always inspiring.  Her message is that we must re-think what getting old really means. People over 50 are happier, better at problem solving, pattern recognition, empathy and have twice the success rate in entrepreneurship. Wisdom is a real thing. And if that wasn’t enough, she pointed out that the 50-plus community contributes more than $7.6 trillion dollars to the economy. Charlotte built a compelling case as to why we should stop thinking about older adults as a burden, and instead see them as an asset.

She also emphasized how important a sense of purpose, positive attitude and connectedness is, and provided some meaningful statistics:

  • Older individuals with a sense of purpose have 40% lower health care costs
  • Older adults who feel isolated cost Medicare $6.7 million additional dollars
  • And, a positive view of aging helps you live 7.5 years longer

She concluded with her rallying cry: Instead of aging in place we should be thriving in motion!

We also had a terrific panel discussion. Ray asked penetrating questions.  To kick off the conversation, Laura Landro, who had just published an important article in a special report on health in The Wall Street Journal, talked about trends in care delivery, noting that the hospital as we know it is becoming a thing of the past. Things like micro hospitals and ambulatory surgery are where it’s at today.

Charlotte added that there needs to be a shift from the medical model of care to a personal model of care — in healthcare healthis the ultimate outcome.  For the consumer, living wellis the ultimate outcome, but our current provider structure has failed to acknowledge the shift.

Ray asked the group to address those factors standing in the way of greater adoption of connected health technology. I made the first volley, stating that we haven’t done our best at thinking about what the individual wants and needs, or how to design technologies to appeal to the patient. For example, products to help the ‘sandwich generation’ — designed for adult children of aging parents — can be a turnoff for the parent or feel intrusive. Charlotte rightly pointed out that we also need to understand what a technology is solving for. For example, feelings of isolation are very different then feelings of loneliness and are solved via very different tools.

Responding to the question about how to close the intergenerational gap, Charlotte suggested we start in the workplace, hiring older adults, which has been shown to improve productivity. I noted there is also a digital divide, which creates design challenges. For example, people with two or more chronic conditions tend to use less technology, yet they could potentially benefit the most.

Ray posed a final question: If you could make one change to help aging Americans, what would you do?

Bruce suggested that we liberate payment from the doctor visit, which could go a long way to opening up new opportunities.

Charlotte responded by suggesting we change how we measure success. Rather than looking at data on mortality, morbidity, hospital readmissions and the like, she would measure life satisfaction.

If I had to identify one change to help aging Americans, it would be to accelerate value-based payments. In this way, we change what we measure and change what we pay.

And, finally, Laura urged that we work on patient engagement, and finding ways to figure out those who are in most need and then pay for services that will help them take better care of themselves.

We wrapped up the event with a networking reception and the hopes that Ray and his team will bring the Summit back to Boston again next year. We are also looking forward to continuing the discussion — Partners Connected Health, PCHAlliance and the Summit — as part of the 2018 Connected Health Conference taking place in October.

The importance of this topic can’t be overstated.  As we enter the era where our over-65 population outstrips our youth in numbers, we need to completely rethink the older demographic and how we can leverage this amazing asset as a society.


Privacy Trade-Offs: Where do we draw the line in healthcare?

May 10, 2018

Everybody has something they consider private.  I think of it not as an absolute but as a sliding scale.  We all have stories about (typically young) people who share everything online in a very open way, seeming to not care about potential ramifications. But, I suspect even they have information they’d rather not share.  There are others who eschew any online presence entirely because they feel all of their information should be private.  Importantly, there is no right or wrong here; it is a personal matter.

And, there’s the old adage, pay attention to what people do, not what they say (bear with me on what this has to do with privacy — and connected health).  How many times have you heard that?  The root of this pronouncement lies in the fact that our brains color perception in so many complex ways that we report information that is often at odds with reality. Logic would dictate then that accurate, independent measurement that reflects reality is a better source of truth than asking any individual for their opinion.  Hey, isn’t that what wearables, mobile phones and sensors do?  Yes, they do!  Thus, if I want an accurate, quantitative representation of your life so I can help guide you to a healthier future, I want access to all of that ‘digital dust’ that you leave behind every day without thinking about it… GPS data from your phone, mobile purchasing habits, step counts, number and frequency of outbound messages, etc.  If I could get access to all of that information and create a unique persona of you, I’d have a better chance of guiding you to a healthier state than if I just asked you for answers to a handful of questions. The challenge lies in how an individual defines what parts or pieces of that data should be considered his/her ‘private information.’

The recent Cambridge Analytica debacle was a wakeup call for me.  Until then, I had the attitude that if people were so concerned about privacy, they must have something to hide.  Now I see what having your private information lying around like ‘loose change’ can do.  It IS a real challenge.  What are we to do?  Rely on inadequate, self-reported data to make decisions that guide an individual to improve health or convince them to share data that they may consider private.

Walgreens Balanced Rewards App Dashboard

There are four examples I know of where companies are flirting with this dilemma in the marketplace.  United Healthcare, for a couple of years now, has been giving people up to $4/day if they meet certain activity goals – as measured by a wearable.  Walgreens, via its Balanced Rewards program, offers customers in-store savings if they meet certain activity targets. Oscar Insurance and Humana, via their partnership with Vitality, have similar programs.  Thus, a small subset of consumers is getting used to trading private information for a financial, health-related reward.

I wonder if the recent noise about Facebook and Cambridge Analytica is changing consumer ideas regarding these programs.  Will enrollment enthusiasm cool?

Then there is the inevitable conversation as to whether health data is really different.  There are numerous data breaches today, credit card hacks is just one example, yet people continue to shop online, bank online, book travel online. The risk, and the convenience, is worth it to most people.  The biggest fear that sets health data apart is the fear that an insurer will deny coverage based on these data.  The Affordable Care Act, by prohibiting plans from denying coverage of preexisting conditions, should have eased these fears.  Our current political uncertainty undoubtedly raises anxiety.

I am not aware of any insurer who is collecting wearables data and then using this information to increase an individual’s cost of care.  With these early experiments, insurers are focused on rewarding individuals for healthy behavior.  Could we see a future where companies funding the cost of care use creative benefit design to shift more costs to individuals demonstrating poor health behaviors?  I think so.  It would show up as a higher baseline premium, while maintaining rewards for those achieving healthy behaviors (i.e., the consumer showing poor choices would not feel punished).

In The Internet Of Healthy Things, we talked about a number of these issues, including a vivid description in chapter 1 of a future where I trade lots of personal information that enables a personal health coach to guide me to healthier behaviors, in exchange for a premium reduction.

There has been a lot of talk recently about Facebook’s business model.  Some have suggested they offer a subscription service which allows users complete privacy. The company has rejected this notion, probably because selling targeted ads is far more lucrative than any subscription model they can envision.  By way of analogy to the healthcare insurance world, a future where you remain largely anonymous to your insurer is likely to be very costly to you.  Although insurers are not selling ads, they will get similar value out of knowing about you so they can target programs to you that have the best chance of improving your health.

I firmly believe that it is critical — in support of lowering healthcare costs and improving health outcomes and provider productivity — that we begin to synthesize these passively generated health data and use them to drive individual health recommendations.  Data breaches like the Cambridge Analytica example are scary and will set back public opinion on this topic.

The industry is taking notice.  The recent OpenMedReady framework developed by Qualcomm, Philips and others is one example.  Maybe if we work together we can thread the needle and get this one right.

Do you have suggestions on how we can leverage personal health data to drive better outcomes but assuage anxieties around privacy invasion?

(Focused) Reflections from HIMSS 2018

April 30, 2018

Connected Health Experience – HIMSS

Most folks who attend the annual meeting of the Health Information Management Systems Society (HIMSS) these days will tell you that to make the most of it, you ‘need a strategy’.  On the one hand, just about everyone you would want to meet is there, but with 42000 attendees, you are unlikely just to run into anyone.  Increasingly, people hang out with their own tribe as there are many side meetings. In the last few years, I’m happy to report that all of the areas of health information technology that I am passionate about have become more prominent, so I can pretty much fill my time with connected health events and networking.  This year, I had the extra treat of being asked by my colleagues at the Personal Connected Health Alliance to Emcee their event, The Digital and Personal Connected Health Forum, on March 5th.

The staff of PCHAlliance did a fine job curating this event and all of the talks were excellent.  Herein are a few nuggets that I gained from attending that are still resonating with me weeks later.

David Asch – HIMSS 2018

The day opened with an inspiring keynote from David Asch of Penn Medicine’s Center for Health Care Innovation. David is one of the most prolific thought leaders in the space.  His speech did not disappoint.  My favorite part was when he talked about the need for rapid validation techniques for innovation as opposed to large-scale clinical trials.  The latter have their place but they are time-consuming and expensive.  Asch argues that at the beginning, one needs to employ techniques that allow for iterative development.  He used the analogy of A/B testing, which many of us are familiar with in the world of web design and called out 3 interesting validation techniques. Much more detail is available in his 2015 NEJM piece, but here is the Reader’s Digest version.  The first is called the vapor test.  The idea is to offer something in the open marketplace to see if there is consumer interest.  If a user clicks on the product, an explanation such as ‘this item is out of stock’ is presented.  In reality the manufacturer is simply trying to find out of there is demand.  The second is the ‘fake front end’, which tests out a change in work flow before implementing it.  The third is the ‘fake back end’ where rather than build sophisticated chat bots or other AI systems, you employ actual people when volume is low in order to learn about how the product will be used once it really is automated.

We deal with the challenge of rapid prototyping and clinical research differently (e.g., we employ matched comparison methodologies and the like) so I found David’s talk quite informative. 

The next talk was a tag team, Greg Weidner and Ann Somers-Hogg from Atrium Health.  They talked about a program using automation in service of hypertension control and showed that 80% of those patients could be brought under control without a human interface. They cautioned, however that as we introduce more and more automation into healthcare we need to design programs that feel as human as possible.  This is very consistent with my most recent TEDx Talk.

Rosemary Ventura from New York Presbyterian gave a very practical and interesting talk on n how they consolidated messaging apps on the inpatient service.

Withings products from Nokia Health

Nicolas Schmidt, from Nokia, came up with a new phrase that I really liked – “Grandma Approved”.  His whole presentation was on the journey that he and his colleagues took in designing the original Withings line of products and how they were designed with elegance and simplicity of use in mind.

David Higginson, from Phoenix Children’s spoke about their initiative to put an iPad in every hospital room. The story had multiple layers of insight, including their very patient, family centric approach, to all of the necessary IT integration requirements and the vendors they chose to do these.

Jody Hoffman gave a very detailed policy update, focusing on new reimbursement trends, licensure and interoperability.

Angela Higashi from Kaiser Permanente talked about one of my favorite connected health applications: wound care.  She is wound specialist and told the story of how they used imaging at the point of care to extend her expertise without moving her.  The program continues to expand.

Image courtesy of Qualcomm Life

Jim Mault is an industry veteran and a real pro when it comes to telling his story.  His was the first time I’ve heard about AI in action. So many times, when I talk to clinicians about patient-generated device data, they express worry that they will be asked to review long lists of normal readings. Qualcomm Life is now pioneering a software layer on the inpatient side that takes all of those streams of data and gives the clinician insights rather than just data.  Bravo. Other industries have done this, and we can’t do it fast enough in healthcare.  I can’t wait to see it move into outpatient applications!

Juhan Sonin, from MIT gave us a glimpse into all of the data streams that are being generated that we don’t even think about (mainly data streams from our mobile phones) and how we might use them to improve healthcare.

Josh Neumann, from Salesforce did a really nice job of helping the audience think about healthcare through a customer relationship management lens.

Lee Hartsell, from Duke, talked about crowdsourcing research.  He pointed out that only 10% of users used an app because, in that case, there was nothing compelling for them to promote continued use.  He made a strong case for bringing the community of patients into the design process.

Amy Schwartz from Battelle made an excellent case on why design thinking is critical for any patient-centered initiative.

Vik Bakhru gave us a perspective on the challenges of scaling digital health in the context of social determinants of health. We need to change reimbursement models – move more toward preventative and think about the big picture.  Big art, not big data!

Jill Carroll, from Mt. Sinai gave a thought-provoking commentary on incentives. They are helpful but not a panacea.  He talked about many behavioral economic principles including present bias, the power of habit, cue routine reward and how loss avoidance is more powerful than gains.

A terrific overview.

I left the day feeling educated and stimulated.

Value-based or Fee-For-Service Connected Health Reimbursement: Which Canoe Should We Put Our Feet In?

April 17, 2018

About 10 years ago, I and many others, started talking about how care delivery enabled by connected health should be an ideal strategy in the world of value-based (VB) reimbursement. To date, there have been just a few instances where this has come to pass. Most relevant is Kaiser Permanente, where > 50% of patient interactions are virtual.  Unfortunately, there are few other examples of organizations that have invested heavily in connected health and state publicly that it represents a strategy for success in a value-based world.

Image courtesy of National Telehealth Policy Resource Center

By contrast, in the past decade, there has been significant progress in payer reimbursement for telehealth as a service (fee-for-service [FFS] payments).  For example, 48 states now have Medicaid requirements for telehealth reimbursement (10 years ago it was about 25); 21 states have requirements for remote monitoring reimbursements; and 15 for store-and-forward telemedicine reimbursement.  Currently, 33 states have parity laws that mandate private payer reimbursement for telemedicine services.  This clearly reflects the growing mainstream adoption of connected health.  The ongoing work of the American Medical Association’s Digital Medicine Payment Advisory Group (DMPAG – which I have the privilege to co-chair) is also an important force in creating CPT codes which will enable even more fee-for-service payment activity.

For healthcare provider organizations experimenting with value-based payment scenarios (ACOs, bundles and various risk arrangements with local health plans), doing business in the last several years has been described as having “a foot in two canoes.”  The tools you use to optimize reimbursement in the two worlds (FFS and VB) are sometimes diametrically opposed.  As we look at the trends in connected health adoption, is it time to think in terms of FFS-driven business plans and declare which canoe we should put our feet in?

Before I disclose my opinion, it is worth looking at how value-based programs have been implemented, to see if there are any clues as to why connected health has not been the savior as I thought it would be.  I’ll try not to over-complicate this part, as there are multiple value-based models and they get pretty complicated pretty quick.

Medicare ACOs are based largely on a shared savings model.  For the population that Medicare defines as ours, providers get paid through the usual Medicare financial channels (i.e., fee-for-service billings), and try to manage various costs in such a way that we meet savings targets on a yearly basis.  Given that the basic transactional model hasn’t changed, it is probably not surprising that care delivery at the ground level has not changed that much.  If I am a doctor seeing Medicare patients, I still get paid for office visits and procedures.  It’s the job of the delivery system I work for to manage those costs and contain patient interactions so I don’t incur unnecessary costs by billing for services I perform.

With the exception of full capitation (after our experiments with this in the mid ‘90s, no provider has any stomach for this model), all value-based models are variations on this theme.  As a result, there is much hand waiving about providers getting paid for ‘outcomes and quality’ and, at the system, level we are.  But those individuals who are doing the work, day in and day out, are being paid for services rendered just like they used to.

There are many other nuances to talk about regarding value-based reimbursement strategies, but this is the fundamental reason that connected health has not become more of a critical component to success in these VB programs.  Whatever the payment model at the top of the food chain, if a provider can’t get paid for a service rendered, it is predictable that she won’t provide that service.

This realization hit me a year or two ago and I decided to put more personal energy into creating opportunities for ‘ground-level’ providers to adopt connected health. It was right after that that the AMA graciously asked if I would be willing to help with their DMPAG. It was a perfect fit for how my thinking was evolving.

I am not trying to make the case that value-based payments will go away. On the contrary, they are here to stay.  I am just as bullish on connected health as a strategy for success in the VB world as I was 10 years ago.  Our own examples of success at Partners HealthCare — such as using home monitoring to manage readmissions in patients with congestive heart failure, and to streamline management of patients with hypertension — are proof that connected health can provide value in these settings.

However, until we change how those providing care record that they provided a service and thus derive their income, we can’t expect providers to happily embrace connected health as a strategy, whatever the system-level incentives.

Another lens to view this through is the use of connected health in care delivery.  One of the fastest-growing segments is video-based, virtual visits.  This tool improves patient access but has minimal impact on efficiency of care delivery. Thus, the future of virtual visits, in the context of value-based programs, is not clear.  By  contrast, asynchronous telehealth (direct e-visits with patients and e-consults between providers) provides an opportunity for both efficiency and improved resource utilization.  As a dermatologist, if I am asked to review a short history and a couple of images generated by one of my primary care colleagues, in many cases, I can help that individual care for the patient without the patient seeing me.  The differential for what I get paid for the e-consult, versus an in-office visit, is significant and helps the organization with its top line resource management.

The case for remote monitoring is even stronger in this context.  One nurse in a call center can manage (by exception) 100 or so patients with congestive heart failure.  This is where efficiency really kicks in.

So, putting feet in the fee-for-service canoe makes sense for now. It is the only way to insure provider adoption.  However, we must keep in mind the big picture of how we fit into value-based arrangements.

What are your views?

Bring Your Own (Health Monitoring) Device: Progress and Challenges

March 12, 2018

As we see virtual visits go mainstream (witness the recent prime time TV ad from UnitedHealth Group during the Winter Olympics), adoption of remote patient monitoring is lumbering along, still in the land of early adopters.

There are several reasons for this lag (and for the corresponding growth in virtual visits).

  1. The unit of service is a visit. For now, doctors still make most of the decisions on how healthcare is delivered. They have a deeply engrained view of care delivery measured in units of visits.  For millennia, this is how we’ve been delivering care and capturing/codifying related work – in terms of visits.  Remote monitoring opens a world where the service offered isn’t a visit, but rather surveillance of a population with proactive, preventative care and management by exception.  Despite our commitments to the Patient-Centered Medical Home and value-based payment models, we have yet to get comfortable with the workflow around surveillance and management by exception.
  2. Payment is still a concern. There is progress (see my related post) on this with the unbundling of CPT code 99091 by the Centers for Medicare and Medicaid Services (CMS) and new codes expected for 2019. But uptake of 99091 has been slow because it allows for time spent on the collection and interpretation of patient generated health data at a minimum of 30 minutes of a physician’s review time, per 30 days.  It is hard to scale that in a busy practice.
  3. There are three aspects of the technology that need refining: cost, usability and EHR integration.  I intend to expand upon these items in this post.

At Partners Connected Health, we’ve been working on each of these issues, using a new ‘bring your own device’ (BYOD) infrastructure and a process that democratizes the deployment of remote patient monitoring devices. But let me back up for a moment.

For several years, we managed remote monitoring programmatically.  We had dedicated staff, a preferred solution and managed our own device inventory.  We reasoned that the best use case for remote monitoring was in the context of our value-based payer contracts and that the deliberate roll out of this care model would make the most sense.  The logic was good, but the costs of maintaining inventory were simply too high.

At one point, we did an analysis and showed that at the current rate of program effectiveness and cost, we’d need to enroll 10 times the number of eligible patients just to break even on our risk contract bonus payments.  When presented with stark data such as this, we asked ourselves, “Can we make the program 10X more effective or should we think about lowering costs”?  As we grappled with this reality, we saw other use cases for patient-generated device data emerging across our delivery system; ones that we hadn’t anticipated and that our program structure could not support.  For these reasons we felt we had to venture into the brave new world of BYOD.

After surveying the marketplace for options, we formed a partnership with Validic and changed our process in some interesting ways. This is my first progress report on our new approach.

The infrastructure we set up with Validic (and also using tools from Intersystems Healthshare platform) is called the Connected Health Integration Pathway (CHIP).  It makes it relatively easy for our patients to collaborate with their physicians around their device data.  The first step is that the doctor invites the patient to participate through a link in our patient portal.  The patient goes through a straightforward set of steps and links their consumer device account with their patient portal account (this is the step that Validic enables).  Through the Healthshare integration, the clinician is able to see patient device data in the context of their Epic record. They can communicate about the data via the patient portal or in the context of an office visit.

This approach gets us out of the inventory business, for starters.  It also enables clinicians around our system who have novel ideas about using patient-generated device data to easily set up their own programs.  Our role is more enabler/support as opposed to program oversight.  In this way, this new infrastructure should help us with the cost conundrum referred to above.

We have been doing this for a few months so we’re just beginning our learning.  Here are a few early insights.

  • The overlap between people who wish to share their data and those we wish to monitor is small.  This is really not a surprise, but it points out that we are still some ways away from true BYOD.  Most of the clinicians that are using CHIP are finding ways to distribute preferred devices as part of trials/research programs or other efforts that have budget to support device distribution. So, yes, we are out of the inventory business but we’re not really seeing BYOD.
  • Despite our efforts to simplify onboarding, the technology is still complex.  Roughly 50% of our earliest cohort of patients had trouble connecting or staying connected. While onboarding is straight forward, patients must use third-party device manufacturer apps on their mobile devices as conduits to move data from sensors to the cloud, and these apps are widely variable in their usability.  One glaring challenge is that our preferred blood pressure monitor — from an accuracy/clinical perspective — has a very challenging, non-intuitive app. Just our luck. We’re watching carefully to get a sense of whether the resources we had put into inventory management will simply re-emerge as needed tech support resources.  This part is particularly frustrating for me…We need to all work together to reach the goal of frictionless data capture or ‘wear and forget’ data uploads.
  • There are indeed many more use cases appearing that we had not thought of.  For example, some of our providers are showing lots of interest in using activity trackers to follow post-operative progress and as a proxy for general health improvement.

It is early days in our new BYOD world.  I know we’ll get there.  I can see a path to ubiquitous use of patient-generated device data through continued improvements in technology (especially ease of use), improved EHR integration and new reimbursement pathways.  It is an exciting time.

What’s your experience with BYOD?