This week I am talking to Abboud Chaballout, founder/CEO of Diagnoss, the medical coding “whisperer” company. Like many of my guests Abboud has an interesting journey, in his case through law school developing experience founding and building companies as an entrepreneur and working in the medical coding and billing world where he learnt about money flow in the healthcare system and especially in medical offices. Along the way he discovered the challenge of many medical offices and systems trying to manage their debt and revenue cycle which all served as a foundation to creating Diagnoss a company using AI insights to offer real time help to doctors.
As he points out there are two coding points in most parts of the healthcare system where a code is selected and the translation mechanism is driven primarily by reimbursement. Get a code, get reimbursed, but this fails the system and the everyone working in the system.
The selection of a code takes place with a coder and a physician in partnership. The physician picks codes, an activity they were not trained for and have no time to find those codes from a huge list that is poorly filtered and then passed on to a medical coder who is trained to pick codes but did not train in medical school and does not have the full context the clinicians to.
This reminds me a lot of the early days of medical transcription where transcriptionists were in a related role and as they learnt and added to their clinical knowledge the really experienced medical transcriptionists would be able to give the diagnoses on a medical transcription before it was diced by listening to the details of the dictation/note.
With the Diganoss system the physician receives real time insights to help the doctor pick codes as they progress through documenting the case in the EHR. The system uses metadata from multiple sources to guide a more customized list of options based on the individual physician supplemented by the context of patient and the clinical information. The system also uses the clinical documentation from the EHR to provide more filtering of choices offered and will remind the physician to include codes from all the available clinical information
Listen in to hear how they achieve insights that can be delivered to the physician at the point of care to provide better, faster more accurate billing and the exciting future that starts to offer additional suggestions such as tests that the physician would like want to order helping solve one of the ongoing challenges of administrative burdens that adds to clinic burnout
Listen live at 4:00 AM, 12:00 Noon or 8:00 PM ET, Monday through Friday for the next week at HealthcareNOW Radio. After that, you can listen on demand (See podcast information below.) Join the conversation on Twitter at #TheIncrementalist.
Listen along on HealthcareNowRadio or on SoundCloud
Raw Transcript
Nick van Terheyden
And today I’m delighted to be joined by Aboud Chaballout, he is the founder and CEO of diagnose a booth. Thanks for joining me.
Abboud Chaballout
Thanks for having me, Nick.
Nick van Terheyden
So as I always do, I think it’s important to get the context of individuals that come on my show. Tell me a little bit about your background, how you got here, what the sort of journey was to founding a new company in this space?
Abboud Chaballout
Absolutely. So I have launched a few companies in the healthcare, healthcare space. In the middle of law school, I actually took sabbatical to launch a company around the financing and litigation of outstanding b2b healthcare debt. And what that what that basically means is we were providing liquidity to providers by buying their unpaid bills. And these are bills unpaid, by insurance companies, to the providers, and then litigating them, and recovering on that debt. And through that process, I learned a lot about how money flows in our healthcare system. And I and more importantly, I learned a lot about what pain points providers face in their day to day operations of their businesses. And and the reason that this was this, this was so vivid was because these providers were were in distress selling their medical bills to begin with. So I saw a few common threads and a few a few common patterns that really gave me insights into what it means to operate a clinic or a hospital. At but but at the same time, I also was able to taste what it means to operate a medical business myself, because I ended up managing an urgent care in a pharmacy during that that same time period, and through that experience, really learned about how expensive and how difficult it is to run a well oiled and quality operation within a clinic. And this was just with just one urgent care. The idea was to launch throughout the state of California with with multiple units. But I decided that one was enough and and got out of that business as quickly as I stepped in. And I, at the same time also managed a medical billing company, and learned a lot about what it means to service medical providers from a revenue cycle perspective. And one takeaway from that experience was that the the, the incentives between providers and revenue cycle, vendors can often be misaligned. And, and essentially, the way I look at it is, is now I’ve seen healthcare from a number of lenses and a number of angles, and decided to take that learning and see if I can do my part and making the system a little bit better.
Nick van Terheyden
So I truly fascinating, but I’ve got to ask, how does a lawyer end up in that sort of initial startup, there must have been something that triggered this to get you to sort of jump out and do that.
It’s a good question and this, this goes into my let’s call it personal ambitions, Aye. Aye. Aye. Aye. I didn’t know this, or I couldn’t articulate this early on in my life, I always thought I was going to be a doctor, and then ended up going to the Middle East where I served as a Fulbright fellow under the State Department. And when I came back to the US decided that I no longer wanted to go to medical school. And I and I thought that I can have more of an impact and play a bigger role as a lawyer. So I went to, I went to law school. And while I was in law school, I realized that I just wanted to do more. And, and and the next, I guess, space where you can do more is, is in the entrepreneurial space, I saw an opportunity and jumped out of law school for a few years to jump in.
Nick van Terheyden
I it’s one of the things I love about this program is the opportunity to discover that because you know, the journeys that get us to these points are always important in the drivers. And ultimately the same driver that, you know, physicians and anybody in healthcare has, you know, helping solve those problems. But those different lenses are so important to deliver the value because, you know, if you just have one set of individuals, you can’t really solve for it. And, you know, you bring this incredible expertise from a legal standpoint, and obviously, all the billing experience, which, you know, incredibly important that, you know, most folks don’t focus on so very interested to hear. So clearly, you know, this fantastic background that really delivers insights that I think other people would not have. And now you’re focusing on something that I looked at and thought, wow, this is the medical coding whisperer, or medical coding whisperer, a company that is providing a solution to help solve that problem in all of these physician offices where they’re struggling with a system that is designed to work, but not for them is my sense of it. Tell us a little bit about that, if you wouldn’t?
Sure, lots to unpack there. When it comes to coding, since you bring up coding, when it comes to coding, the way I’ve always looked at it, and the way I always describe it is that it is a translation mechanism that our system depends on for a variety of reasons. The most important reason as far as providers are concerned in the us is that coding gets them to reimbursement. So you translate what you did as a provider for a patient and what is what is the element that the patient carries to the insurance company and then the insurance company will pay you some amount of money. That process as I saw it from all my experience is broken. Because the the way that that translation happens, it typically will happen with the participation of two parties, the provider, and either a medical coder or a biller, and they do this in concert with each other. Now, the reason that that’s broken is because in many situations, you have a provider, initially picking codes, but the provider was never trained in that translation exercise ever in medical school. And on top of that, they have no time. They literally have no time. The medical coder, on the other hand, is trained in that translation exercise. But but their problem is they never went to medical school, yet they’re expected to read these complex doctor’s notes and then do a proper translation. And then therein lies another point, which is you know, they depend on the notes, but the notes themselves are often not as complete as they could be, because the provider literally has no time. So there are there are many issues on that contribute to the error rates and coding which is around 30% that contribute to how it feels like it’s it’s administratively burdensome. And what what what I thought was that it would be very valuable and helpful if we helped providers on the front lines of this panel. Pick better codes. And where were when I started the company. My initial vision was, let’s let’s have the let’s go from doctor’s notes to codes to build, to build with one with just one click. But but we’ve we’ve shifted a little bit in our thinking, to to not be as naive. And to understand that coding is still complex and and that a machine is going to is going to play its role. But the role is not as simple as notes to Bill his notes to codes to bill. And and the reason for that is as I’ve gone through this journey of building an AI company, I’ve basically concluded that, that that AI plays very specific roles. But those roles are within very defined task sets. So so what I’m what I’m envisioning is, there will be AI solution for a provider, and then to help them select the initial codes. And then there will be an AI solution post provider to just validate and verify that these that now these groupings of codes are the right codes given the payer mix given, you know, various other parameters that will that that one can essentially rely on to go to bill. And, and, and I believe firmly that providing a provider with a tool that just helps them pick codes is extremely valuable in its own right, even though that code may not be the code that the the clinic or the hospital is going to rely on to build. And there’s a couple of reasons for that. One reason why it’s, it’s a huge value add, because today,
providers have no support in that code picking process. They have they at best, they can rely on two things, number one, their memory. And number two, they can rely on a search bar that the EHR provides for them, which is often buried somewhere. And even when they use it, they’ll search for a code, let’s just say, let’s just say they’re looking for mild intermittent asthma, they’ll search for asthma, they’ll type a St. And then they’ll get this long laundry list of codes, you know, revolving around asthma, and then they’re stuck scrolling, and reading and maybe going cross eyed in the process of going through a few dozen codes to pick the one that they want. So So this process is it takes time. There’s a lot of clicks involved with it. And it’s really not, it’s really not efficient. There’s no Smart Search associated with it. So so we just believe strongly that the there’s a gap in support here, in terms of in terms of the provider and we want to deliver, we want to deliver that support to the provider. And by the way coders today. They do have support. There are what they call computer assisted coding tools. Excuse me, yes, computer assisted coding tools for medical coders and billing teams. But these tools are not designed for providers. And if a provider was to use it, they it would be way too distracting.
Nick van Terheyden
So for those of you just joining, I’m Dr. Nick the incrementalist and today I’m talking to Abood shabba lute. He is the founder and CEO of diagnostics. We were just talking about the whole coding system and the challenge of these two groups that pick the physician in the first instance, who is struggling with that, and then the coders and they lack the sort of crossover training the supporting infrastructure. One of the things that really strikes me about this that’s kind of interesting is it reminds me a lot of the times, going back when medical transcription was a much bigger activities still exist. And one of the things that the medical transcriptionist that were gold dust were the ones that were able to you could challenge them, they could listen to listen to history and physical and they’d get to the diagnosis before the actual physician had dictated the diagnosis and they had learned so they were creating that insight. Same with physicians. I imagine in terms of the picking codes, there are probably some that are good. What you’re doing is bringing an overlay. So help us understand how that works and what it is you’re doing, because it sounds like, you know, fantastic resource that physicians are going to love for a number of really exciting reasons.
Yeah, absolutely. The, the way that we do it depends on the level of intelligence that we’re that we’re that we’re basically going for, or that’s requested. Although we’re an AI company, you know, we’re very, we very much look at more how more basic solutions can add value from the get go. So we’re one thing that is is interesting about what we’re doing is that we’re able to add value on this continuum of intelligence. And what that allows us to do is be able to add value from day number one, rather than, you know, take a few months to develop something as hyper intelligent for for for a customer to utilize. So what what is it that what is it that we do, on a very basic level, what our technology will do is look at the metadata surrounding a patient encounter, and provide a set of codes that that could be relevant. And, and here, I want to, I want to emphasize the word metadata. Um, so for example, we can give you your most frequently used codes as a provider, and that’s auto populated, you never have to go to a billing pick list and select favorites, or build that yourself, it’s just there. And then the codes are available to you. In order of frequency. And what you’ll have is a search bar that you can use from any screen within the EHR. So what what way to put it in contrast, today, in order to search for a code, you have to go to a very specific screen within the EHR, usually, it’s going to be on the assessment page of your note. And, and, and that’s where you’ll search for a diagnosis. With with our tool, you can be documenting the history of present illness, and then you can search for diagnoses as you’re going right right there on that page. And you might even be able to start picking codes, and even post them to the to the assessment page or to the billing screen right from the history of present illness page, just just as an example. So that that’s on that’s on the the basic level, on the more let’s call it on the more sophistic sophisticated intelligence side of the spectrum, where our technology will do is actually read your note in real time, and predict
codes targeted to that set of notes. And this has a variety of benefits. Number one, you’re not you’re essentially getting codes served to you on a silver platter as a provider, but also, you you are, it’s it’s more likely that you will start picking codes that you would have otherwise forgotten because because the machine is serving as the second pair of eyes. Or as this expert that’s sitting on your shoulders, reading your notes and whispering codes into your ears just you know, writing you Hey, you address this, you should consider you should consider this code. And and they’re there for you to just glance up and click on if you if you agree with those codes. Now in either implementation, our goal is is is one of one or both of two things, number one, to significantly reduce the number of clicks, it takes you to pick codes. And number two is to help you pick more accurate codes and a more comprehensive list of codes. And one one note here is and this was an IT this was an interesting discovery that I that I found through the journey of building diagnostic is for many providers, coding is not coding what they call it is document Meaning my diagnoses. But the way EHR is are built is is the way they have you document your diagnoses, they’re essentially having you pick a code, even though there’s no code on the screen, if that makes sense. So, in my opinion, most providers are coding, some of them just don’t know it.
Nick van Terheyden
So, as I think about that process are you’re also so you’re clearly taking the content that they’re entering putting in? Are you also taking content that exists on the screen? Are you able to take additional inputs that maybe they haven’t entered but already exist as part of the EHR?
Yes, yes, absolutely.
Nick van Terheyden
Yeah. So all of that context contributes on an ongoing basis, there’s this real time activity that sort of provides additional insights, one of the things that really sort of strikes me is that, you know, much as anything, this is an efficiency feels like it sort of addresses, you know, some of the challenges that we have the productivity challenges around, eh, hours, which is definitely one of the major contributors to burnout at this point.
That’s right. And, and, and that’s where we’re headed as a as a organization, now, we’re thinking bigger than than just coding. And what we ultimately want to do is help improve EHR productivity for providers. And, and what one way to think about that is, okay, we’re, we have our sidebar, and it’s integrated with your EHR, and we’re giving you as a provider codes, but we can also, but, but while we’re giving you codes, and as you interact with our tool, we can probably start giving you likely labs that you’re going to order and help you avoid going to the lab section of your EHR, you can just order the lab right there from our from our sidebar, or help bring on screen the most likely prescription that you’re going to write and again, help you save clicks, or, or you may have questions, certain questions that you’d like us to answer. And, and we can build the model. And and and serve the results of that model on screen for you. So for example, now that I’ve picked these grouping of codes, what’s the likelihood that my claim is going to be denied. So I can I can, we can essentially serve that serve that on screen for you, so that you can make certain adjustments right then and there, rather than wait until a coder or a biller looks at it, and then sends you a message and then force you in a situation where you now have to go backwards in your, in your workflow, to remember what happened in a particular encounter, just to fix some documentation, or to fix some aspect of your code selections. Um, so so. So really, there’s a very strong path where we can help providers just be more efficient. And the way we look at it is, as consumers, we have a lot of tools at our disposal, that helps us be more efficient. One example of this is, if you use Gmail, Gmail will predict, you know, that you write half a sentence, it’ll predict the other half of your sentence for you. And that, that, you know, let’s say that even works 50% of the time, it’s it’s super helpful. So so we want to bring this type of seamless productivity boosts to providers working in EHR, and we believe that we can make a difference in their day to day practice.
Nick van Terheyden
Wow, that’s really exciting. I mean, I think, you know, as a concept starts to give real value, and remove, as we always call it, Scott work, you know, the drudgery that exists around healthcare, the administrivia that just sort of frustrates people. And it feels like this is a real potential solution for clinicians all over that, you know, can potentially add value to what they do and allow them to get back to the patient, which is ultimately the goal. So, exciting stuff. Any last thoughts in the remaining minutes of where you see this going and what the opportunities are?
Yeah, absolutely. And I believe, I believe that if, if, if we I believe that if we provide providers with with these predictive enhancements for their EHR is not only will it make their lives better, as providers, but it will also it also stands to improve the provision of care. And I’ll give you just a very quick consumer example. I believe it was Pinterest or Instagram that did a study on the use of emojis, if they if if they successfully predict the next emoji that you would want to use in a post. And what they found is by doing that, they actually boosted emoji use. So imagine what that means for providers in in a healthcare setting. So one, one problem that I’ve heard from from clients, especially in the value based care world, is that some providers are not picking enough diagnoses to paint the right. Story, the right patient story in a value based care setting. Well, if we’re if we’re predicting codes for you, I mean, there’s a strong chance here that it will boost that that the comprehensiveness of a patient Nope. So I just wanted to leave you with that example.
Nick van Terheyden
Absolutely. Fantastic. All right. As usual, we’ve run out of time. I just remains for me to thank you for joining me on the show. A bird it’s been a great pleasure. Thank you, Nick.
