How technology can streamline tedious health care processes [PODCAST]

Listen and followApple PodcastsSpotifyYouTubeMore apps · Browse all episodes

“Companies that embrace digital processes will reap the rewards, as shown by the Forrester survey taken in 2020 after the pandemic began. The survey found that digital document processes were helping organizations maintain their business resilience and even reap direct rewards of their digitization effort. By leveraging digital document processes, organizations can increase employee productivity and collaboration in a remote environment while continuing to serve customers. Nearly three-quarters of respondents said digital document processes supported their business continuity efforts.”

Bevey Miner is a health care executive.

She shares her story and discusses her KevinMD article, “NLP and AI technologies streamline tedious health care processes to enable faster care.”

Did you enjoy today’s episode?

Rate and review the show so more audiences can find The Podcast by KevinMD.

Subscribe on your favorite podcast app to get notified when a new episode comes out.

Click here to earn 1.0 AMA PRA Category 1 CME for this episode.

Also available in Category 1 CME bundles.

Powered by CMEfy – a seamless way for busy clinician learners to discover Internet Point-of-Care Learning opportunities that reward AMA PRA Category 1 Credit(s)™. Learn more at about.cmefy.com/cme-info

Do you know someone who might enjoy this episode? Share this episode with anyone who wants to hear health care stories filled with information, insight, and inspiration.

Hosted by Kevin Pho, MD, The Podcast by KevinMD shares the stories of the many who intersect with our health care system but are rarely heard from.

Transcript

Kevin Pho: Welcome to the show, where we share the stories of the many who intersect with our health care system but are rarely heard from. My name is Kevin Pho, founder and editor of KevinMD. Rate and review the show at KevinMD.com/rate. Subscribe at KevinMD.com/follow. Today on the show, we have Bevey Miner. She’s a health care executive, and her KevinMD article is titled “NLP and AI technologies streamline tedious health care processes to enable faster care.” Bevey, welcome to the show.

Bevey Miner: Thank you. I’m excited to be here and to talk about something I’m really passionate about.

Kevin Pho: OK. We’ll get to your article in a little bit. First off, briefly share your story and journey to where you are today.

Bevey Miner: So what’s interesting about our health care system is that we have so much unstructured data, first of all, which is plaguing us with the problems of a lot of data entry, a lot of human intervention, a lot of bureaucratic processes that can really be streamlined. At the end of the day, we may look at this as workers that are not necessarily working more efficiently, our administrators not working more efficiently. But at the end of the day, when you look at how efficient processes can streamline care and really make a difference for health outcomes and patient engagement and patient satisfaction with our system, it really becomes extremely meaningful when that can happen.

So we’re looking at this from an administrative and a clinical process in the use of AI and NLP. That’s what I’m really passionate about. We’ve all heard stories about NLP, AI in particular, in health care that can help diagnose, and those are plagued with some controversy right now. So where we are focused is really trying to streamline that administrative, bureaucratic process and remove some of the human intervention that we find bogs down the system today.

Kevin Pho: So it’s no surprise, and I’ve been talking about this for years now, that the intersection between health care technology and medicine has not always gone smoothly. You have the advent, of course, of the electronic medical record systems, and that hasn’t always been as smooth an integration. I’m a primary care internal medicine physician, so I certainly see that the attempt to sort out all that unstructured data has been met with more than a few obstacles. You talk about some of that, as well as potential solutions, in your KevinMD article, “NLP and AI technologies streamline tedious health care processes to enable faster care.” Now, for those who read your article, just walk my audience through it. Share the story of why you decided to write it.

Bevey Miner: So in looking at how bad this problem is, first, look around your offices. Look at the people that follow you and where the paper is. We’re still standing at fax machines, and we’re still sending paper faxes. Why? It’s easy. A physician will turn to an administrator and say, “Fax this,” and they stand at a fax machine. We did a survey of our administrators, and we have a lot of administrators that use our solutions at Consensus Cloud Solutions. We asked them what’s the most efficient way they find to communicate, and they said the fax machine. Over 60 percent of them, 61 percent, said they still use a fax machine today, and they envision still using it five years from now.

So we have to first get the system off of paper, because paper has all sorts of issues, and I can go all the way down that way. But once we get the physician, or we get the administrative processes, off of paper and we move them to something like a digital cloud fax, which is what we offer, it’s still an unstructured document. It still needs to be data entered. We have to find the patients, and it might be a PDF that’s attached to a patient’s chart. But as you know, and your listeners know, physicians still have to look at all of that locked-down data. We can go a step further. We can make that data meaningful, we can make it actionable, and we can put it in the hands of the physicians and the clinicians that are doing treatment, so that they can make a decision that has to be immediate, that has to be accurate, and that has to engage the patients.

We’ve all lived through some of those horror stories, even as patients, when we’re wondering, “Where’s my latest lab result? How come I didn’t get that?” A lot of patients don’t go out to their MyChart and watch everything. They’re waiting for their doctor to quarterback them the right way. That’s why I was passionate about writing this article, because I believe it’s time we started to look at some of these technologies that are in other industries and how we can deploy them. It’s about time that we started to stop this bureaucracy. We’ve got 60 percent burnout, and physicians say the reason for that, 60 of them responded, is bureaucracy. I worked at Allscripts and NextGen in the days of meaningful use, and it really started back then. A lot of physicians said, “Just give me my paper charts back so I can look at a patient when they’re in my exam room.” But I think we can get there. So that’s why I wanted to make sure that your readers and your listeners could hear that story.

Kevin Pho: So give us some common scenarios, let’s say in a typical primary care office, where you can really paint a picture. Tell us a case study or scenario of some of these technologies in action. What would that look like?

Bevey Miner: So think about administrators having to get in all of this data, all these documents, for document management. Let’s talk about a referral management process and how locked down some of that data is for referrals. The one instance I’ll give you is that a PCP needs to refer a particular patient over to an orthopod to get some sort of surgery done, and they’re going to do it in a way that all that data is unstructured. I mean, if it’s on paper, it’s horrific, but even take a digital cloud fax, it’s still unstructured. It goes over to a surgery center, it goes over to an orthopod’s practice, and it’s going to sit with someone that’s going to have to data enter it and look at scheduling.

If you can pull out the patient’s demographic information, now you’ve got all of the first name, last name, addresses, and everything else that has to be data entered. You pull that out as structured data. You pull out the diagnosis codes, the ICD-10 codes. You can pull out the clinical, the lab values. All of that now is completely structured data, and the way that looks in the cloud is that it is a lot of metadata that’s just in flight right now. When that metadata can be mapped into systems, the administrators can accept that patient from a referral standpoint almost immediately. They can add all of the lab values. All the information on the patient can be put into the admission system and put into the electronic health records, so that when the surgeon is seeing that, looking at it, looking at the course of treatment that needs to be done, it’s already structured in a way that they can make a decision fast.

If that process goes as a manual process today, it can take a week, it can take two weeks, for a patient that needs to get into surgery fast. If it’s done with NLP, AI, and structured data mapped into a system, you can shorten that to 24, 48 hours, and now you’ve accelerated care. The health outcomes are better. In this case, the patient’s bone isn’t setting wrong, and you didn’t have to have a human actually touch it. So we reduce errors, we reduce burnout, we can redeploy some of those resources that would just be doing data entry for patient engagement, for patient consult, for patient communication. The technology exists. We just have to deploy it and adopt it.

Kevin Pho: So one of the things that you mentioned was how it could reduce burnout. Give us an example of one of your clients where your solution reduced the burnout of the clinicians.

Bevey Miner: So let me give you an example that’s clinician and payer. That, in and of itself, right there, payer and provider, there’s burnout for that whole process, and there’s never been a love fest between payers and providers today. But one of the problems in our system is prior authorizations. As patients, we’ve all sat and said, “How come I can’t get my medication?” You go to the pharmacy: “Oh, it had to be authorized.” It wasn’t authorized, you didn’t go through a step therapy, all sorts of reasons. Or, even more complicated, a particular procedure, an MRI, or maybe it’s chemotherapy treatment. It doesn’t necessarily make any difference what it is.

Our prior authorization process started out with noble intentions on behalf of payers. They sat back thinking, “Look, we have clinical data that can show what’s the best treatment, what’s the best course of action.” But today, in that prior authorization process, the physician sends that to the health plan, waiting for an authorization. The patient is sitting on the edge of their chair going, “I’ve got to get in for a test. I have to get in for an MRI.” And there are hundreds and hundreds of people in certain health plan workloads that are manually looking at these prior authorizations. First, they have to find the patients within the member databases. Then they have to look at all the data for medical necessity, all sorts of information.

When you can make that very structured, a physician could send in for a prior authorization, and in flight, like I mentioned before, all that metadata is now mapping directly to that health plan’s portal, or their system, or their workflow for an authorization. You can even add robotic process automation, which I didn’t mention in our article. That’s something new we’re introducing. It says all that structured data has now gotten into our member portal. We understand what the physician is asking for. That’s all structured. If this, if this, if this, and if this, let’s just say it’s four different criteria, that’s authorized. That can all be done automatically too.

So it is a problem, and I’ve recently done some research on it. It harms patients significantly if they can’t get in for the treatment they need, treatment that can help save their lives or make sure that the medical outcome is a better result. And physicians are the middle person, knowing that they wanted to provide this kind of treatment, which is why they’re asking the health plan to approve it. The back-and-forth, back-and-forth administrative process is horrific. They’re calling, their administrators are calling to get the authorization, the patient’s calling the physician, and they’re going, “I’m trying.” It’s something that we’ve just got to fix. There are studies, and some famous cases, that note how bad that problem is. Some health plans are even saying, “Look, we have some docs that are so great at this that if they send in for an authorization, we automatically approve it.” So that almost means the movement is, “Why did we even do this?” I think that’s a pretty good example, because it creates tons and tons of burnout. I don’t care what specialty you’re in or what type of clinician you are, it’s a problem.

Kevin Pho: We’re talking to Bevey Miner. She’s a health care executive. Her KevinMD article is titled “NLP and AI technologies streamline tedious health care processes to enable faster care.” Bevey, you mentioned earlier that there are some controversies regarding natural language processing and AI. What are some examples of those?

Bevey Miner: Well, not to necessarily throw any technology under the bus, but very, very early on, we tried to just model artificial intelligence into the health care system and said, “Look, we’ve got something like IBM Watson.” It won Jeopardy, which is really great. It’s an absolutely smart brain. It’s got some fabulous machine learning. We can do some amazing things when Watson can look at symptoms. They can look at the SOAP notes, they can look at a lot of different observations on a particular patient, and also all the lab results on the tests, and make a decision on a diagnosis via an artificial intelligence brain. There are famous cases, and we all know that IBM Watson tried to do this at MD Anderson for cancer diagnosis, and it didn’t work.

Our artificial intelligence is machine learning. It’s repeated patterns, and it’s fast. It can learn really, really fast, and it can correct itself and make sure that inaccuracies are improved upon. In fact, in the e-prior authorization process I was mentioning before, we’re even looking at handwriting recognition on “urgent” or “stat,” because docs are always famous for doing that. We’re nuancing that, and this is a physician’s handwriting we’re nuancing, so our accuracy level is getting better and better all the time. It was released well before its time, and today we’re working on different ways that we can look at artificial intelligence for smarter diagnosis.

It doesn’t remove the physician, and it never should remove the physician. There’s a science and there’s an art to the way a physician will diagnose and determine a treatment plan. No way should artificial intelligence and machine learning replace that. And so there’s been some fear about how patients can be treated using an AI technology. There’s fear that Amazon could do it. There’s fear that Google’s got this thing going on. I don’t think the diagnosis piece is one that we’ve figured out just yet, or even the treatment piece.

So for us, when we look at the use of NLP and AI, we don’t want to replace the physicians for that really critical piece, the reason they went to medical school to begin with. We want to get rid of all of the bureaucratic, administrative processes and the clinical processes that bog down treatment, and then let the physicians. There are a lot of different things you can do with population health, comparative effectiveness, retrospective and prospective kinds of research, and there are all sorts of peer-reviewed articles that come out to help physicians determine the right course of treatment or how to diagnose. So it’s not meant to be removed, but for us, it’s really about the other pieces that we think can give physicians less burnout and less stress and put them back in that seat of quarterbacking the treatment the best way.

Kevin Pho: And my final question: What are some of your take-home messages that you want to leave with the KevinMD audience?

Bevey Miner: I really want your audience to start to embrace these technologies in a way that can really help them do what they want to do, and that’s to help patients, treat patients, and diagnose patients. In that administrative and clinical process, it’s about taking that first step. It’s about not being afraid of the technology, because it has to actually be completely proven, because these are new technologies that the system in health care is extremely, and I don’t want to say extremely, but they’re on the outside going, “We’re not quite sure about how to use this.” These technologies, done well in partnership with a company like ours, can really have some meaningful outcomes.

So it’s about time that the health care industry said, “Oh, let’s dive in. Let’s start to test some of this stuff. Let’s look at the proof, and let’s look at some case studies and use cases,” and started moving, because we have burnout everywhere in this system. We’ve got shortages of clinicians right now, and we have shortages of administrative staff, and that’s hurting patients. We’ve got a way that we can solve this, so embrace it and start to test it.

Kevin Pho: Bevey, thank you so much for sharing your time and insight. Thanks again for being on the show.

Bevey Miner: I appreciate it.

Listen and followApple PodcastsSpotifyYouTubeMore apps · Browse all episodes