Thesis Maps previously explored a16z’s vision for the future of bio and healthcare. Today, we chatted with JD Tyler, MD, CEO of JustAskEvie, whose company is one we’ve identified to watch within the healthcare “operating system” space.
Thesis Maps (TM): Could you tell us a little bit about your background, and what led you to start JustAskEvie?
JustAskEvie (JAE): My background is in medicine: I did a dual residency in internal medicine, as well as pediatrics. It was six years of training that's squished into four years, and I truly thought I would just be a hospitalist, which is a physician that only takes care of patients in the hospital. That was my intention once I graduated from training and I was happy doing that. I've always loved technology and am a bit of a tech nerd myself, but I'm much more passionate about healthcare.
I noticed that every single healthcare organization started adopting these electronic medical records (EMR). There was a push to move records into the care of the patients, which I was a fan of because I truly saw the big picture that the EMR is a great way to get objective discrete data and really produce more of an evidence-based medicine approach versus a typical anecdotal approach. I tell you this, because I knew just how important EMR was going to be, but the reality was, it was a bottleneck of my day. It was just additional work, so it was something added to my already full plate.
Just by a coincidence really, one of the hospitals I covered was right next door to a Cerner office, which is one of the larger EMR vendors, and that was probably the most common one that I used. At the time, I wondered if they hire physicians to help them make this work for physicians. And I actually went to work for Cerner with intention of being there for around six months, just part-time to help them understand how physicians work. I love technology and I thought I could speak their language a little bit, and that six months actually turned into almost seven and a half years. I was on the physician lead side developing software to help physicians work, not be additional work.
It was actually at Cerner that I first noticed that organizations were hesitant to take any sort of upgrades. The reason was that training is hard to do and support is expensive. When I did see organizations roll-out new technology or upgrades, they would struggle, but I was able to help them when I was on-site, as a peer, able to speak physician to physician. When I left Cerner, I held positions as Chief Medical Information Officer and Chief Medical Officer for two organizations onboarding new EMR systems, and I witnessed firsthand just how difficult and expensive it was to do onboarding and support.
From all of these experiences I saw a giant gap. There really wasn't any good onboarding, training, or ongoing support. You couldn't scale it, because the best option was peer-to-peer, but you really couldn't recruit physicians because they are already busy and it was very expensive to compensate those physicians to travel to a location and be helpful.
So I wondered how to connect a network of people who are struggling to learn the EMR to the network who has already learned it, who speaks their language? How can it be done real-time through technology and be cost effective and productive? So that's what led me to start JustAskEvie.
TM: Beyond training and hesitancy to invest in support, why do you believe existing players have not significantly addressed the shortcomings of the incumbent products as far as ease of use and virtual support goes? And how have hospital systems tried to solve for this on their own?
JAE: When it comes to Cerner, Epic, McKesson, Allscripts, etc., I think it never really occurred to them. They do have physicians that will visit on-site, but they aren’t elbow support people. They're more of a physician that helps guide on the software side as to what the organizations should take as upgrades and what would enhance patient care. Vendors do offer elbow support, but they're non-clinical.
Additionally, there isn't a single version of Cerner that is at every single client; same with Epic and Allscripts, and so on. They all have little different versions of the core software. It's not like installing Windows or MacOS, there are a lot of different versions that are installed. It would be very difficult for Cerner to be able to appropriately support in a way that’s cost effective, and so the responsibility somewhat fell on the hospital systems themselves.
Hospitals have generally tackled this by establishing a helpdesk, or dedicated lines that go directly to a clinical IT support person. It's usually not a physician because organizations struggle as it is to find enough physicians to hire and have them take care of patients. They'll typically have IT people that are super users, and they do a good job, but it’s just the little nuance of not speaking that clinical language.
I've talked to plenty of organizations who have said, “We’ve tried to create what you created. We just couldn't do it. We just couldn't figure it out. We didn't have enough resources or time to invest in this type of software.” And building that network of super users is the tough part: the software was one thing to create, but recruiting a lot of super users that are physicians, nurses, whatever other role, is tough to do as well.
I knew that JustAskEvie would make a difference in healthcare, and so I've been focusing on just executing and now spreading awareness.
TM: How’d you get the first couple customers? Was this something you just brought to them, or were they doing a formal RFP to evaluate potential solutions?
JAE: They definitely did not seek me out: we got the first two clients through grunt work and networking. They had already picked on-site at-the-elbow support from their EMR vendor. That was the thing, they weren't evaluating us as a virtual solution against anybody else, because at the time everything was on-site. Everybody was at the healthcare system, nobody was doing anything virtually.
TM: So, how is on-site support pricing structured, and to what degree is that a significant expense for hospitals?
JAE: It’s a very significant cost, which is why they either cut back on it or they try to get a better price. When you pay for on-site support you try to use as much as you can of your own internal team because you’re already paying them a salary. In order to get additional support, you're looking at anywhere between $85-125 per hour for non-clinical people to show up and provide support at-the-elbow. And that's not including that if they're not local, you have to pay for their travel, per diem, their lodging. It gets very expensive. A lot of places just do about two weeks of support, but in doing so, they will spend hundreds of thousands of dollars, depending on the size of the organization. If you're going to want a nurse or a physician, you can probably double it. You’re probably paying $150-250 an hour for a clinician to be able to do the same thing. It's just not cost-effective.
Most companies such as consulting firms that do this recommend a roughly 20:1 ratio of users to support people. So, if you have a health system of 2,000 people, then consulting services say you need about 100 people on-site to support those staff members. So you're paying a 100 people roughly $100 an hour, and that's a low figure because you have to include per diem and everything else.
This is an inefficient approach; those on-site aren’t even doing a good enough job because you can't physically be everywhere at the same time. On the other hand, if you're virtual, you can be in a lot more places. So, you can even stretch the 20:1 ratio, to maybe 40:1, because most of the time spent by the at-the-elbow person on-site is walking from floor-to-floor in the hospital, trying to find people who need help.
At JustAskEvie, we calculated the true cost of answering one question when you're at-the-elbow and that's about $50. This is based on how much the hospital likely paid for on-site support, including travel, per diem, lodging, etc. It's about $50 per question. And that’s just the first week, when the big bulk of the questions come in. If you look at week two, three, four, five, six, people still have questions, they still need help, but you're paying the same amount of money per hour. So the cost per question is going to actually be higher, closer to $100 and beyond.
My approach was, just buy a bunch of lifelines and that's what you pay for. You're paying for results. And that’s how we compensate physicians and nurses who answer the questions. If you're really good at your job and you can answer these questions virtually very efficiently, you can answer a lot of questions each hour and get paid per question.
TM: What KPIs do you focus on as a founder?
JAE: I would say that perhaps the biggest KPI is “time in chart”. Cerner actually measures a lot of stuff like that; it doesn't tell you exactly what's going on, but it gives an indication of how long they're trying to figure out things. I want that number to be as low as possible.
Are the providers actually learning it better? Do they feel more comfortable asking for help and getting tips and tricks? Is that time in chart going down? That would be one of the KPIs. We’re working to get a big enough dataset to say, “There's statistical significance that for my providers that get support, they spend less time in chart.”
Another one that is harder to prove out, and, it would have to be from a bigger health system and longer time, would be medical errors. A lot of medical errors occur within the EMR. Not on purpose, it's just because somebody probably didn't know how to use it. They probably picked the wrong thing or wrong order set, or they just missed something. A lot of vendors, they're trying to increase automation, which means you're using data to help improve patient outcomes. I'm all for that. The bad part of that is it becomes alert fatigue many times if it's not done well.
For me, I know JustAskEvie is a part of that equation. If everybody was really comfortable with the EMR and they knew how to read it and write their orders and communicate with it, then that should get rid of a lot of the medical errors and decrease malpractice issues. And that in itself, by making less medical errors, you're improving patient outcomes. And so that would be another KPI that I would want to focus on in the long run.
TM: Where do you see the business in a year from now? What are some of your biggest challenges?
JAE: A lot more clients, for two reasons. One, the pandemic proved out the virtual support model. Coming out of the pandemic, we’re expecting that we’re going to have some competitors, and that's going to make us better as well. We’ll have someone to compare ourselves to and improve on our experience. That makes it more challenging, but that's part of life. Two, the majority of the time that organizations seek out support is during IT projects: implementations, new installations, upgrades, etc. During this pandemic, all of those have taken a back seat in healthcare and rightfully so. Every healthcare system has been focusing on COVID patients. And now they're all focused on vaccine roll-outs so we can get back to some sort of sense of normality. A lot of those projects have been on hold, which means when things start calming down in 2021 or 2022, the floodgates will probably open and they will have all of these projects to do and they will need a lot of support for it.
Right now, it's just been bootstrapping, just sweat equity, our own money and time. You know, we've actually made money on this and, and it's been a great experience, and we've built the foundation of the software in order for it to scale very quickly, but what we'll need is marketing, sales, everything else it takes to run the business. How do we pay the Evies? How do we recruit the Evies and make sure that they're appropriately knowledgeable for the right organization? All that's going to take funds.
TM: As a part of scaling, how are you thinking about positioning the product for different types of healthcare systems?
JAE: We understand the challenges of a large healthcare system versus a small, critical access hospital. And we built the software to scale that way. We can easily roll it out to a critical access hospital same day. The critical access hospitals, those are not going away. There's so many rural communities here in the Heartland, in the Midwest, where those are absolutely needed. But, the problem is they're so understaffed and under-resourced, they actually need someone like us.
Larger healthcare systems, they probably already have staff that's pretty capable of doing support from 8am to 5pm. We can take the friction out of the request process and the support process, using our software, where we replace traditional helpdesk phones with an “easy button” and show you how to use it. And then we can cover 5pm to 8am and weekends, when on-site support staff isn’t there.
It's a different size market, but we know how to approach that versus a critical access hospital. I would not be shocked if in a year from now, if there's other companies providing remote support. I'm priding myself on making sure that we’re a higher quality of support; we may be a premium service a year from now compared to others, but we'll let the market decide.
Enjoyed this post?