This week I am talking to Benjamin Zaniello, MD MPH, the Chief Medical Officer for PointClickCare, a healthcare data integration and management company. Ben is a practicing Infectious Disease doctor who has struggled with the challenge of finding useful information in the burgeoning Electronic Medical Record (EMR)
Ben is an infectious disease doctor, who has done his own spelunking into EMRs seeking critical patient data. For his specialty, this is already challenging as he looks for “bugs and drugs,” (what diseases have they had and what antimicrobials have they been on). Even with additional access to more data and more EMRs with more patient data, there is no easy way to look through them
We discuss the challenges of EMRs and how they have taken away the focus of doctors’ eyes and attention from eh patient to the computer screen as aptly captured in the JAMA article some years ago featuring the 7-year-old’s view of her consultation with the doctor
Ben quotes from The Rime of the Ancient Mariner by the English poet Samuel Taylor Coleridge
Water, water, everywhere,
And all the boards did shrink;
Water, water, everywhere,
Nor any drop to drink.
as a shorthand for the knowledge or information in our EMR but our inability to find useful information in these healthcare databases and systems
Listen in to hear our discussion on the challenge of finding the unknown unknowns and how we can use incremental steps to free the data and release insights and knowledge to be accessible to doctors and other clinical team members struggling to deliver the best possible care to the patient
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 welcome Ben Zahn yellow, he is the Chief Medical Officer for point. Click care. Ben, thanks for joining me today.
Ben Zaniello
Thanks excited to be here.
Nick van Terheyden
So if you would tell us a little bit about your history and how you got to this point, you’re an infectious disease doctor, which, you know, I’m going to say in the last three years was probably pretty relevant and important. And you were maybe a little bit busy with some of that. But you’ve done a whole range of other things. And it’s always helpful to get a bit of context about the individual that we’re talking to. So tell us tell us your story.
Ben Zaniello
Probably unusually, compared to maybe a lot of other Doc’s that spend time and technology I actually did technology. First, I graduated Stanford in the rational exuberance of the 90s. And went straight into technology decided that wasn’t entirely as socially aligned, I was essentially helping people either sell or create faster widgets, and went into medicine. And after Well, plus years of medical training, fellowship, etc. I’ve used healthcare tech as a good opportunity to align my technology background and my, at that point, current passion for medicine. Again, a lot of medicine is about doing things at scale. I think most physicians, and I’m certainly a practicing physician, I see patients every week. I love that individual relationship. But the opportunity to do it at scale is a really big part of healthcare tech. In other words, not what I can do for this patient, but what I can do for a population. So I went kind of back into technology while just seeing patients part time, a little over 10 years ago, and mostly have not looked back.
Nick van Terheyden
Yeah, so it’s interesting, I would say that I had a similar sort of experience, but I didn’t start in technology. Certainly in college, or you know, qualifications, my journey through that was through school, through the original sort of concepts of technology, I could see this as a clear pathway. But medicine was what I wanted to do. And then I just became frustrated with the fact that nobody was talking technology. In fact, I carried around a little portable device. And the biggest mistake I made was that one of my patients made a little holster for it to make it portable. And I should have patented the holster, forget the technology because I’d have made a lot of money with all these clip on things that people
Ben Zaniello
think I was gonna go in a totally different direction with that which I thought you’re gonna say your biggest mistake was walking around the hospital with a Oh, no,
Nick van Terheyden
I was very proud of that. No, I was, I was known as you know, that weird guy that had a brick clip to his side. And it was a two line LCD screen with an ABC keyboard and 16k of memory plug in. But I used it, it was very effective. It was fantastic. But that launched me on my career. So I see that blending of the two worlds. And, you know, I’m gonna take a little bit of ownership here, and I’m interested to hear your thoughts about it. I contributed to the morass of electronic medical records, I implemented some of the very early versions of this. And I think we brilliantly decided that we were going to take the paper record and create a digital version, which was of course the right thing to do, right.
Ben Zaniello
Yeah. And also, since we’re, we’re being upfront here, voted i in the sense of having a technology background, and then moving through the long and painful pre medical, medical school, residency and fellowship, I actually helped implement, in most cases epic at all of those institutions, because I was either a practicing physician or a med student and new technology and could explain it to others. So I got sucked into that now I would like to think that I was the low person on the totem pole and I was just doing what I was told as a brain medical students Hello, so I really actually more blame you. But I, I was I was part of the system. I was absolutely part of that.
Nick van Terheyden
I’m gonna take even more ownership. I just it was a horrendous mistake to actually automate what we had on paper and produce a digital version it it just it did not work. And of course, much like the the move in technology. Oh, we’ve got digital, you know, Pat, we’re going we’re getting rid of paper. How Hello, have you seen these printers the high volume, it was the complete reverse. And I think we did the same with electronic medical records, we didn’t really solve problems. And I know you’ve been knee deep in that. Tell us a little bit about your history through that process and where you see us in terms of the electronic medical record.
Ben Zaniello
Sure, I think it’s kind of widely known today that most providers physicians, particularly but also nurse and med techs spend more time staring at a screen than they do actually practicing care. There’s been many studies on this, and each has more depressing that the last that in a patient typical patient encounter, the doctor is trying to in that 20 minutes allocated patient time that probably started late, and also involves seven other screening questions now also includes having to capture all the data from the encounter to create a bill. And so they’re sitting at the screen, they’re not actually looking at their patient, very few people can type and create a articulate note, while staring at a patient’s. This is a huge bummer. Right? It’s been a huge bummer for providers, it’s an even bigger bummer for patients. I mean, again, the practice of care is not always so much a complicated diagnosis, or even a miracle treatment. It’s really understanding where your patients are coming from hearing their stories, and then often providing them the reassurance that it’s not necessarily a medical diagnosis for which they need 17 pills, it’s often a reassurance that what they’re going through mental, behavioral, health, physical or otherwise, will probably resolve on its own. But listening for the gotchas to ensure that your patient has a healthy happy life, and that you also retain your license. Again, this is all been under buying, because I’m spending the entire time staring at the screen. One shift we’re seeing, and again, a little bit about that new from new electronic medical records, is how can we automate some of these tasks? So a lot of the things, the screening questions, for example, that everybody is still presented with when they walk into an office, here’s a piece of paper, going through these check marks, okay, can we capture this in an automatic fashion on an iPad, of that automatically go into the medical chart, and anything that say, Yes, I am waking up in the middle of night coughing blood, that actually gets relayed to the providers so that they can have a specific interaction or about it. So there’s a step there. And then the other piece is just the workflow part, which is creating a streamlined system, where as the patient kind of moves through the encounter, the medical record is moving with you, it’s pulling up the right data, it’s putting it in the right place, all these things need to happen. Or else we’re gonna see what we are seeing today, which is really unhappy providers, Doc’s that, say, I went into this profession to help people. And really, I’m just a glorified typist.
Nick van Terheyden
Yeah, and that’s really, you know, such a cool problem, because you and I both went to medical school to see patients not to see technology, even though we’re both geeks, clearly, I’m just going to call you that because you sound like, Well, I’m just, there we are. And, you know, there’s value to all of this. I mean, I saw the value I carried around this ridiculous device that you’ve called me out for now. But I used it and I could see, but it wasn’t the focal point, it was not where I spent my or at least where I wanted to, and I didn’t have to in those days, to be clear, it was, you know, centered around paper. And, you know, we still, we had personal connections, we’ve removed all of that. And the other thing that’s happened, and I think, you know, you’ve certainly got some perspective on this is this massive increase? And I’m going to call it data, but I’m not sure that’s even giving it the best term here. It’s certainly not knowledge. It’s just data on patients and associated facts around their state. Perhaps we’re overwhelming the clinicians in that experience. How on earth are we going to fix this? I mean, it just, it’s getting worse. We’re getting more and more of this information.
Ben Zaniello
I just want to clarify something very quickly. However, there is probably a gradient to how much of a geek you are and people that wear stuff around their belts are certainly higher than me. I just, I just started
Nick van Terheyden
Oh, gone a second, calling me more geeky, just because I had a clip. Oh, please, not
Ben Zaniello
just me. Gotcha. So the data thing is really interesting to me. And it’s again, careful what you wish for. So A huge part of my career, let’s say in the last five or six years, has been focused on this buzzword of interoperability, right? It’s this idea that we’ve created these silo health systems that are not exchanging data with each other I practice in Seattle, Seattle has three big health systems, Virginia, Mason, University of Washington and Providence, they have hospitals, within a half mile of each other on first hill called pill Hill, patients would just move between the three, sometimes with bad intentions, but also sometimes with good intention that er is crowded, I’m not getting care there, I’m gonna go there, because I’m going to be seen faster, there was no exchange of data between those three hospitals. So my I was the CIO at one of those health systems. And my goal was, how do I make it so health systems are sharing data, not to mention the rest of the care continuum. So the post acute skilled nursing facility, you discharge a patient to the clinic that is going to catch a patient when you discharge him from the IDI, that we both have the intro problem of health systems, not sharing data, but then all these other entities on the care continuum, also need access and visibility into this data. So we spent a lot of time building pipes, right, working with the big EMR manufacturers, including the one I work for point, click care, but also epic, and Cerner and others to get the data out there. Well, this is a little bit of a careful what you wish for, or, you know, the quote, Water, water everywhere, Not a Drop to Drink, right, we’ve got all this data now. Now. Epic provides everybody talks about care everywhere. And I can click over a couple of tabs when I’m seeing one of my patients, and see everything that patient has done at every other health system on Epic, mostly limited to Epic, but a lot of health systems that are an epic, well, guess what? It has everything they’ve done in an epic system. So going through that data, the word I like to use the spelunk, because it’s like you’ve got a flashlight, and you are desperate to find the kind of gold or in that mind, the one stallion piece of data in that just morass of data that terror everywhere in these systems now have is virtually impossible. You I’m an infectious disease doc, like part of my job, when I am called as a consultant is to Splunk in the EMR to find out what are all the bugs that someone has had, what are all the drugs that they have been on, and encapsulate that and come up with a narrative. That’s my job. In er medicine, you have a short period of time to rapidly triage a patient. You can’t go over to a third party or portal healthcare information exchange, and look through all that data to find Well, first of all, is there something there? Right, you don’t even know the answer to that question. But if you go over is anything relevant to your care? And that’s a little bit the new problem. I mean, there’s actually been some great governmental work on this. We have test QA, we have information blocking roles, all these things to free the data. And if you and I talked five years ago, I would just keep repeating free the data free the data, let’s force cajole carrot stick everybody into fearing the data. Now I’m like, Oh, my gosh, we have so much data. And now people are throwing in what’s coming off their Apple Watch and all the all the Internet of Things. Right? Now we have too much data. It’s how do I actually make that data informative and actionable for patients but also the providers themselves?
Nick van Terheyden
So for those of you just joining, I’m Dr. Nick the incrementalist today, I’m talking to Ben zaniolo. He’s the chief medical officer for point click care. I was just taken aback for a second because right here on this show, for the first time, we had The Rime of the Ancient Mariner and included that was Samuel Taylor Coleridge, I’ll just tell you and, gosh, if you can cast your mind that far back into your English, I guess it was that we, we learned that the water water everywhere, nor any drop to drink. It’s extraordinary that that’s, it’s taken this long to come up in my podcasts. But Ben, you did it. So thank you for that. And I’m just gonna say that, you know, you you get the geek award for the spelunking in, in these EMRs to identify that information, because I think that’s exactly the task is that you’re attempting to do is to identify and get in there. And of course, the question is, is there any valuable information We don’t know. But we’ve got some new tools. And you know, I’m almost hesitant to say, Hey, I’ve got an answer to the technical problem that we created were some new technology. But let’s go there for a second and say, is there some scope to actually automate some of this and start to surface in a, I’m going to call it safe way because you hidden in there was something that I know as an ID, Doc, you, you’re always thinking about the zebras, right, the unusuals. And it’s the same in emergency medicine. Can we do it and still include the zebras? And you know, because if you surface all the data and say this is it, that’s what people look at, did we miss something important? How do we how do you think about that?
Ben Zaniello
Yeah, this is when I get to use whatever buzzword of the moment, which again, years ago, we would have liked probably talking about machine learning or neural networks that have, of course, what we’re talking about these days is artificial intelligence. Like any Doc, I little bit, throw my hands up and you know, with fear and say, What is AI going to do to my job? Right? It seems like it’s incredibly capable of writing my 16 year old college essays. I should say it’s application essays. He’s not a Doogie Howser. He still firmly in high school.
Nick van Terheyden
I was a Doogie Howser but go on.
Ben Zaniello
All right. All right. All right. Okay, just checking. Again, back to the holstered device. This doesn’t surprise anyone. But I I’m incredibly excited, what can I say I’m excited by the promise of AI, I am, I do have my concerns. I’m not worried that it’s going to take my job or it’s going to take over the hospital and shoot laser beam therapy. But I would like to find kind of the the use case that drives value. I think we’re already seeing some of these clinical use cases, I think radiologists, for example, are recognizing that a computer reading a radiograph is getting that much better each time. And at some point, they’re both going to be at a automate those million X rays, and also flagged the zebras. That is a little bit, I think the promise of AI in that kind of data data data every way or which is they will both look for the things that are most relevant. This person in the ER, they have had a trauma, I am going to I know that trauma and blood thinner are actionable items. In other words, I’m going to flag that at some point in this person’s medical record their last primary care visit, a year ago, it was flagged that they needed to have a new INR. Right, they needed to have their blood thinner level checked. That is actionable information in the IDI, that before we do surgery with this patient head before we do a procedure, before we do a lumbar puncture, we actually want to know if their blood is thinned and it needs to be reversed. So I think that type of thing, that sort of mass data being processed, and then the nuggets being pulled out is a huge opportunity. I think sensitivity around you know, this is the best clinical pathways for this person, then probably like everyone else, I throw the health equity card and say, What is this been trained on? Right? We have so many examples where medical clinical pathways have been trained on quote, the wrong model. The example I always use when I was trading in New York, about half to three quarters of my patients were African American. And I was being trained on the same hypertension study, everybody was the all at trial 50,000 plus people and you start with hydrochloric thiazides and then you go to an ACE inhibitor. And if those don’t work, then you can move on to well, even at that time, we are talking amongst ourselves, these do not primarily work on our black patients. And we almost always are wasting our time and reducing the opportunity to lower their hypertension. Because it turns out that huge trial, the model, the data set for which we were trained on didn’t have many black participants and as a result ignored the fact that Amlodipine and other antihypertensives were much better. So that’s just one narrow example. But I think we are already seeing it with some of these data models that AI is trained on that they’re not necessarily identifying the best thing for the patient in front of you. They’re identifying it based on the data set that they were trained on,
Nick van Terheyden
you know, extraordinarily important point, that whole issue of the equitable inclusion, certainly in the origin data set and, you know, raises another point, we’re not going to be able to get to it. But that data and those protocols for want of another term, are exactly what’s in use in the Denial System and the insurance industry that says no, no, you must do the following things, because that’s what we know works. And we’re forcing this in other ways, not even in the clinical. So even if you’re a smart physician that says we will know, I know that that doesn’t work, because I’ve had all this experience, you don’t even get an opportunity to do it. So but I, you know, put that to one side, I want to in in the balance of what we’ve got left. Just talk about something and you know, I’m not going to be political, but it originates political, which is, you know, Donald Rumsfeld and his known knowns known unknowns and unknown unknowns. How do we get to that point of? Sure. I mean, there’s never 100% surety, let’s be clear, but you know, conflict, let’s call it a confidence limit that says, I can take what surfaced up? Is this just a constantly incremental approach that says, we’ll eventually get there? And at some point, we’ll get the checkmark and say, Is it FDA approval? I don’t think so. But what is that point where we can have confidence in what’s being surfaced and presented to us?
Ben Zaniello
We have this pre existing research pathway in healthcare that essentially looks at medication, for example, and says, Is this the next great medication now, careful what you wish for, again, right, we’ve all seen how that pathway has been abused in the industry such that new medications come out, that turn out to not be any better than the old medication, they just are more profitable, they ended up being pushed into the system, I think there’s an opportunity to build a better but analogous model around technology. There’s been a discussion with FDA about regulating algorithms. I think that could be unnecessarily complicated. But knowing the great academic researchers across the country, I would love to turn them loose on looking at these algorithms, comparing them against conventional practice, seeing which is better looking for the holes in the models, and ensuring that these things actually work back to the kind of incremental approach, this doesn’t happen tomorrow. And I’m okay with that. I’m okay with avoiding the unintended consequences by taking our time with this and not rushing forward. Again, people are okay with a chat GVT written essay, I think they’re less comfortable with knowing that AI is coming up with their clinical play on that may or may not be in reviewed by a, you know, oncologists for their cancer. We’ll get there. But I think we do need to take a careful approach, particularly in healthcare. Listen, everybody thinks their industry is special and a snowflake. But healthcare is different. When it comes to applying technology. It’s not just about keeping the lights on. This is about people with lives. And, again, they may not realize this until they’re in the emergency room themselves, or until they have a sick loved one. But as soon as that happens, you start to approach healthcare, and innovation and technology very differently. I think that’s a good thing. But it does involve some amount of patience and careful proceedings versus a headlong. Let’s use AI for everything tomorrow.
Nick van Terheyden
To quote from The Life of Brian, you’re all different. Yes, we’re all different. So I think a great summary of the sort of process and how we can get through this, you know, and obviously, I’m a fan of the incremental approach, you know, finding the small steps and, you know, I’m with you, I think letting the researchers loose and opening this up, I don’t think shutting it down as the way that this genie is out the bottle. There’s plenty of opportunity, but it’s not entirely clear exactly how. And it brings me to the sort of closing point which is it’s not really artificial intelligence. It’s augmented The intelligence that still needs the application of the brilliant clinicians and physicians who bring the additional value. Unfortunately, as we do each and every week, we’ve run out of time. So just remains for me to thank you for joining me on the show. Ben, thanks for joining me.
Ben Zaniello
And thanks for having me. Super fun. Any opportunity to call someone a geek? A fellow geek. I’ll take
