Robbie Hughes, MEng is founder and CEO of Lumeon of Boston, MA.
Tell me about yourself and the company.
I’m an aerospace engineer. I’ve been a computer geek all my life. I was particularly interested in the problem of the computerization of industries versus the digitization of industries. That led me when I was still a student into the law, accountancy, and healthcare. All of those industries were activity-based in their reimbursement forms, and all I saw was using computers as expensive typewriters rather than the way I’d been brought up to use them.
Characterize me as a computer geek who was naive enough to think I could get rid of variability in healthcare delivery and stubborn enough to stick at it for now 15 years. But the thing that interests me, and the thing that got me into this in the first place, was, how do you deliver a common standard of care across a network? That’s the variability problem I’ve been going after all this time.
Lumeon is an agility layer that sits on top of the EHR. It helps providers personalize and operationalize a common standard of care across the enterprise. The people who buy that tend to be interested in doing something different, innovating around an operating model, innovating around a new way of delivering care. That lends itself in particular to risk-bearing entities who want to principally use automation to cut out costs and transform care. That’s a lens that we bring to it.
We started in Europe. We moved over to the US, and obviously there’s a huge amount of difference between those two environments. But there’s also a huge amount that is similar, so we’ve been lucky to be able to isolate a lot of what’s common between the two and bring it from one environment to another in a relatively interesting and different way.
Are providers interested in standardizing care when their patient population is a mix of fee-for-service and value-based care?
Absolutely yes, but we need to be really careful about the language. What we don’t do is standardized care. What we do do is standardize decision-making that results in the personalization of care. So the problem as I see it is not how to do the same thing to every patient, it’s how to apply a common standard or apply the same decisions in the same way to every patient. That results in the appropriate application of the right care to the right patient, which in a fee-for-service construct, very happily is usually reimbursed.
The effect we tend to see is not only in elimination of waste — which is good in both environments — but also an increase in throughput, which tends to increase reimbursement and revenue as well. We’re in the slightly strange position of being able to drive up revenue in a fee-for-service environment as well as cut costs in both the capitated or risk-sharing environment.
That’s the core of what we do. It’s basically ensuring that every case, every patient gets the right care, and because the fee-for-service model reimburses the right activities, generally speaking, you will find the reimbursement goes up.
Is it hard to get enough information from the EHR to allow you to provide the best recommendations for all patients?
It’s very difficult. The way we do it is what we think is a little bit easier. I’ve been doing this long enough to know that you don’t try to eat the whole elephant in one go. When we started out, we would go off and do 100-site enterprise deployments that would take two or three years to roll out. Whilst that was, let’s call it educational, I wouldn’t describe it necessarily as fun.
The approach we’ve taken instead is to try to think about what is a digestible version of that problem that can be applied quickly and that can deliver value quickly for the customer, so that you never actually need to solve the problem of, what is the entire universe of care for every patient and every possibility? I don’t believe today that’s a tractable problem. Instead, what we tend to focus on is identify processes where there are gaps or discontinuities or grit in the machine, if you like. Then, how do you apply automation to that to deliver that lift, that personalization, but also that control and predictability to ensure that you are operating at peak performance?
You’ll know that the typical areas that you’ll find this will be around care transitions. For example, that will be around surgery and obviously in population health, where you’re trying to get large populations to do specific things, but each one of those things need to be specific to the individual. Those are the kinds of areas that we tend to specialize and see the most benefit.
Analytics-powered population health was mostly an aspirational legacy software vendor’s overused marketing term a handful of years ago. Has the definition or the expectation around PHM changed?
This was one of the really curious things to me when I came to the US a few years back. I looked at population health management as a category, and I thought, that’s interesting. That should be the application of the appropriate care to the individual at population scale. That’s what we do.
But the reality of the market seemed to be somewhat limited to meeting your quality measure obligations, and most specifically, looking at population health as an analytics and insight problem rather than an action problem. Population health management as a noun rather than a verb.
For us, the analytics are interesting. They tell you where to point the machine, but the problem we’re interested in is, how do you actually operationalize that? How do you solve that last-mile problem so you can drive the engagement, drive the personalization? Anyone can do the analytics with enough horsepower, but actually driving real change in a health system so that you are appropriately intervening with the appropriate patients with the appropriate care at the appropriate time — that’s a hard and interesting problem. That’s the thing that gets us up every morning.
The pandemic has possibly set us back, where we’ve moved to video visits that may be disconnected from the the patient’s usual providers and interrupting their normal health maintenance activities. Has care coordination suffered, or has the pandemic done us a favor to show us what we need to change?
Telehealth is probably the most interesting version of this care coordination problem. Health systems have lurched towards swapping face-to-face visits with video visits, which is a fine and a reasonable thing to do. But what nobody’s really thought about, or at least nobody that I can see has really thought about, is the governance around this.
When is it appropriate to have a telehealth visit that is provided virtually rather than a visit that’s done face-to-face? When is it safe to do so? What are the benefits? What is the standard of care that might be reasonably considered in a remote or in a face-to-face environment, how are they different, and what do you need to do differently?
For me, the orchestration of virtual care and the safety netting of it through the use of a combination of remote patient monitoring, screening, or any number of the other myriad interventions that exist for us today is the ultimate care coordination problem. It isn’t just now a problem of, “this patient is due for their flu shot” or “this patient is overdue for their colonoscopy.” This is now a problem of, for this patient and their presentation, that the next thing that they need to do is share this information, because it’s missing in their medical record. That will then tell us whether they can have the bit after that in this form or the other, et cetera, et cetera.
This orchestration of the fragments of care delivery is going to get dialed up to 11 if we are serious about using… I’m going to use the term virtual care, because I believe that’s different from telehealth in a meaningful way. I think that’s what the consumer wants. The consumer wants something that looks like every other industry, but there is a safety and a governance aspect to the application of these types of interventions in our industry that has not yet been, shall we say, road tested in any meaningful way.
I’ll bet there’s going to be a ton of lawsuits, not just in the US, but globally, next year from patients who have been misdiagnosed, mistreated, or forgotten about because of this very problem. When the dust settles from all of COVID, I think this is going to be one of the more interesting problems for the industry to address.
Much of the value in a visit is simply asking the patient how they are doing and using their answer to guide the next steps. Are we overlooking the value of allowing the individual to electronically document and contribute their own sense of wellbeing, activity level, or concerns?
One of the overlooked aspects of automation is that it should, if done well, enable hyper-personalization. For me, automation is not, at least not in our industry, about doing the same thing for every patient. It’s about looking at the marginal cost of every single activity and trying to reduce that to zero so that you can implement as many different activities as you possibly can to build up the most robust picture and then use that to drive the appropriate intervention.
In your example, I would advocate that the face-to-face consultation could be augmented by tele-triage in advance, whether asynchronously or synchronously, to determine the best use of the face-to-face time that that physician or clinician will have with the patient. It’s a perfectly reasonable thing to do. But in the case that it’s not face-to-face, you could apply the same model, but you can also look at other things.
If the consultation is face-to-face, for example, perhaps the patient has a sweaty palm, and as they’re leaving the consultation, they shake the hand of the physician and say, “Oh, just one more thing, Doctor.” That’s a classic pattern that, in a face-to-face environment, a physician would tend to leverage to gain better insight. But in a remote consultation, they can see that the paint is peeling off the walls, that they don’t have a chair to sit on, that they have 13 cats on the sofa, and there are people shouting in the background. You can build up a picture of the patient that is — I don’t want to say more complete or less complete, but suddenly different. The cues and signals that you look for in these environments are going to be different.
Again, not to say that either of these is right or wrong, but the important thing to realize is the expectation and the baseline that we set for care delivery in the “old normal” is completely different to what we might anticipate in the “new normal,” and we need to adjust. We need to design our interventions appropriately, and we need to recognize that the patterns, cues, behaviors, checklists, or whatever that we had previously are no longer going to be as useful. That’s a huge, huge opportunity if it’s embraced.
Again, this is kind of why I got into this. The trick is, how do you bring it together? How do you orchestrate it with precision? Because there is such a thing as the objectively right care for a given patient. It’s just that in this industry we tend to apply a lot more subjectivity to that than perhaps I think we should.
Will hospitals and practices whose capacity is once again being challenged by the pandemic respond by using those technologies that were rushed into service in the spring – such as telehealth and contact-free check-in – or will we see another wave of innovation?
We first need to come to a common understanding about what the core problem that we’re solving, and I don’t know that the industry has necessarily done that yet. People have applied the solution at hand to the symptoms that they see, but there is another level of optimization that needs to take place to create the sustainability and to create from scalability of even those same solutions and those same interventions before we get to another round of innovation. There’s a lot that can and will be done, but we have a lot to fix on the ground first. I’m not convinced necessarily that there is a universal view in the industry about what “good” looks like.
I would say that there’s the reimbursement problem which needs to be addressed one way or another, and obviously that’s going to drive a lot of behavior. Consumer expectations are being set. I think there’s going to be a lot of conflicting opinions around the level of reimbursement anticipated because the standard of care will be different. I think that’s an entirely reasonable debate, but I would advocate for much more freedom in terms of how people think about reimbursement, particularly around service lines and particular outcomes.
A lot of simplification can happen that will create innovation. I see a lot of complexity being introduced in order to manage some of the risks and bridge to value transition. Whereas if you look to other industries like the cosmetic surgery industry, it’s well published that cosmetic surgery and the cosmetics industry more broadly has been publishing a fixed price for a long time. Costs have been driven down there in an environment that is broadly similar to many other surgical interventions in healthcare. If we can get to a place where there is predictable pricing for predictable care, that will unleash a huge amount of innovation, and we will see a lot of adoption of all kinds of both operating models and technology potentially to support them. But everything begins and ends with money, so I would advocate for that kind of approach. I think if we do that, we will see the kind of movement we all want to see.
What changes do you expect to see with the company over the next three to five years?
The core emphasis of the company is on the US market. The core things that matter to us are around being aligned with our customers. I got into this because we have a very firm belief that it is possible to both take costs out of care and to improve the quality of care being delivered, however you define quality. Every time we’ve done this, the quality comes alongside cost reduction. I’ve yet to see a single example, over my many years of doing this, where the cost has increased and the quality has gone up. It has always been that the cost has gone down and quality has gone up.
That’s our North Star. The one thing we do, the one thing that drives us, is, how do we improve the quality and the consistency of what our customers deliver? Nothing else matters, really. If we do that, then our customers will speak for us. If our customers speak for us, then we will have commercial success, and we will create the flywheel that everyone wants.
But the healthcare industry is not one single, homogenous market. It is extremely diverse, extremely amorphous in payment model, operating model, structure, patient population, et cetera. It would be naive to suggest that one approach will work for each different environment. The customer intimacy that comes from the analysis we do, from the deployment work we do, from that strive for quality, is what makes us different and is what allows us to adjust for that. But I wouldn’t necessarily say that it’s a straight line path to success.
Anyone who gets into this industry who is trying to do anything, let alone any of the problems we’ve decided to solve for, is going to be in it for the long haul. But it’s nothing more than singular focus on that one thing, driving for quality and taking out waste. I think if we continue to do that in the way that we’re doing, we’ll all be successful, no matter what happens in the broader market.
Do you have any final thoughts?
It’s a fascinating time to be in this industry, and it’s a privilege to be able to work with some of the people we do. If I was to go back and give my 23-year-old self some advice, it would be to pick an easier problem to solve than trying to get rid of the biggest problem in the biggest industry in a country 3,000 miles away from where you’re based. But it is an absolute privilege to be able to do what I do, and if it didn’t work, I wouldn’t still be doing it. I’m grateful for the opportunity to talk to you and hope to be doing it for many years more.
Comments Off on HIStalk Interviews Robbie Hughes, CEO, Lumeon
Great seeing Dave Furhmann get a shout-out! Since he helped grow Epic Care Everywhere from its early days, I'd bet…