The Pediatric Lounge, Where Pediatric Physicians Come to Share Their Stories and Success

104 Data Is Everybody's Business and AI Uses in Medicine

Dr. George Rogu, MD, MBA and Dr. Herb Bravo Season 2 Episode 32

Since 1994, Dr. Barb Wixom's research has explored how organizations generate business value from data assets. Her methods include large-scale surveys, meta-analyses, lab experiments, and in-depth case studies; five cases have been placed in the Society for Information Management Paper Awards competition. Barb is a leading academic scholar, publishing in such journals as Information Systems Research, MIT Sloan Management Review, MIS Quarterly, and MIS Quarterly Executive. She regularly presents her work to academic and business audiences around the world.

Before MIT CISR, Barb was a tenured faculty member at the University of Virginia (UVA), where she twice earned the UVA All-University Teaching Award (2002, 2010), which recognizes teaching excellence in professors. 2017, she was awarded the Teradata University Network Hugh J. Watson Award for contributions to the data and analytics academic community. Most recently, she won the 2021 Association for Information Systems AIS Outreach Practice Publication Award for her data monetization research.

Barb authored her new book Data is Everybody's Business (MIT Press, September 2023) to inspire workers across organizations to monetize data. She actively works to encourage women, young people, and underrepresented populations to learn about data and pursue data-related careers.

Dr. Perry Kaneriya, MD, MBA, is a Harvard-trained Neuroradiologist with 18 years of clinical experience in medical imaging. MBA from Darden School of Business with Distinguished Performances in Operations Management, Healthcare Innovation, Finance, Valuations, Accounting, Economics, Entrepreneurship and Creative Design Thinking. Able to leverage extensive real-world experience and unique MD/MBA skill set to develop and execute innovative medical strategies to improve healthcare quality metrics by advancing cost-effective technology-based solutions and data-driven innovations.

We are committed to delivering exceptional medical solutions aligned with corporate goals and company mission—passion for lifelong learning driven by genuine curiosity about emerging trends in healthcare and innovations in medical imaging.

This episode is made possible by a generous sponsorship from  Physician Computer Company.

Support the show

🔴 Subscribe for more Doctor stories like this: 🎧Apple and give us a 5-star review.
Read more on the TPL Website

SOCIAL Media Pages ===============================
📝 - Substack
🎧 - PODCAST
👥 - FACEBOOK
🐦 - TWITTER
📸 - Instagram
➡️ - Linkedin

The Pediatric Lounge - A Podcast taking you behind the door of the Physician's Lounge to get a deeper insight into what docs are talking about today, from the clinically profound to the wonderfully routine...and everything in between.

The conversations are not intended as medical advice, and the opinions expressed are solely those of the host and guest.



Data is Everybody's Business 

Physician Computer Company. PCC empowers independent pediatricians, streamlining daily operations and improving financial stability. A trusted pediatric partner for 40 years, we offer award-winning support, personalized training, seamless data transitions, and practice analytics. With inclusive pricing, a lively peer community, and a free annual user conference, you can focus on what matters the most: your patients. Explore more at PCC.com. That is P C C dot com. 

[00:00:39] Dr. Bravo: Good morning. I have two dear friends of mine, Dr. Barb and Dr. Perry. I've known Dr. Barb for a long time, even though she's moved away. Barbara, thank you so much for joining us today. 

[00:00:52] Dr. Barbara Wixom: It's a pleasure to be here. So much fun 

[00:00:54] Dr. Bravo: ahead. Thank you. And Perry, how are you 

[00:00:58] Dr. Perry Kaneriya: today? I'm doing well, Herb.

[00:00:59] [00:01:00] I'm really looking forward to this. Thank you for your invitation to be here. I've been eagerly anticipating this whole discussion about a topic that interests me quite a bit. 

[00:01:08] Dr. Bravo: Barb, if you don't mind sharing, how did we meet? It's been a long time ago, but I'll tell you. 

[00:01:13] Dr. Barbara Wixom: you, we met because you're really great at data monetization, but not, I'm sure, in the way a lot of listeners are thinking about that term at the moment.

[00:01:23] When my daughter, Yeah. Was four years old. We came to your office for care. And the first thing that jumped out at me was that we were electronically inputting our information. We needed to fill out the patient forms. And back in 2003 and 2004, that was radical. That was different from the everyday experience.

[00:01:45] And then when we would subsequently, sadly, we would have to come back a lot Because you helped care for our children when they were young, we saw you quite often. When we came back, we would have more and more surprises, delightful surprises in our interactions with your [00:02:00] office because we wouldn't have to do what I usually would have to do in manual ways.

[00:02:05] Fast forward to today in my research; I call that data wrapping, and again, we'll get to that later. But it's using data to delight your customers and your case patients to distinguish you. One of the reasons we came back to your office repeatedly, first and foremost, is because e you're a great doctor and person, but also because of your experience in your office, which adds a lot of value.

[00:02:34] Dr. Bravo: That's great to know. How is Chris doing? Your husband? 

[00:02:36] Dr. Barbara Wixom: He's doing great. He's doing great. Yes. Because of your work, we discovered early on that my youngest daughter had Lyme disease, and we could overcome that. And through her illness, we realized that my husband had it too; instead of having arthritis, he had Lyme's.

[00:02:54] And again, he was able to overcome that. Yeah, we're grateful for our experience with you for many reasons. 

[00:02:59] Dr. Bravo: I [00:03:00] always enjoyed having either of you in the office. I got my little MBA from both of them. We would 

[00:03:06] Dr. Barbara Wixom: have a lot of, we would 

[00:03:07] Dr. Bravo: have a lot of business. You guys taught me so much about business.

[00:03:11] Seeing you and the girls was so much fun. It was a pleasant time in my life. Now, Perry, how did we meet? So 

[00:03:19] Dr. Perry Kaneriya: her, our paths unexpectedly crossed earlier this year. You and I were both at a social gathering hosted by a mutual friend of ours who works as a radiologist in my former private practice here in Reston, and then you and I got to talking about, of course, when you have a bunch of doctors together, we always talk about medicine and healthcare, but then that led to a conversation about yeah.

[00:03:39] Non-clinical opportunities for physicians out there led to a conversation about AI and natural language processing. You and I both discovered that we had a shared interest in learning more about AI  potential use cases for AI in shaping the future of healthcare. 

[00:03:57] Dr. Bravo: I'll say his name. Arun is a great cook. [00:04:00] 

[00:04:00] Dr. Perry Kaneriya is a cook and a great mixologist. You never 

[00:04:02] Dr. Bravo has a bad meal in his backyard. It is a joy to be there. Barbara, could you catch me up? You moved from UVA, where you teach the IT master's degree here in Reston, and went to MIT to concentrate on research. So, what have you been doing in the last couple of years? I did. 

[00:04:24] Dr. Barbara Wixom: Yeah, I was at UVA, my alma mater, so Wahoo Wah. I was there for 15 years as a faculty member and went to MIT on a sabbatical to the Sloan School of Management. They have a research center where All that's done in that center is research, but it's an exciting twist.

[00:04:42] It's a nonprofit for one. So, all of the research that we produce is free to the world to help leaders succeed in tech. That's our mission: to help people succeed. I love that mission; it is very altruistic, which is essential to me. And specifically, my research is to help people succeed, help [00:05:00] leaders succeed in data.

[00:05:01] And so, for one, I hope the listeners will go to cisr. MIT. Edu and register again. All the resources, including healthcare examples and such, are there. But when I went up on sabbatical and I started researching, what I also found interesting was the opportunity to work very closely with people in the trenches.

[00:05:20] People in practice and academia in ivory towers, and what we know is trapped in academic journals and not released to need that knowledge. And so what our group does is, with practice, we still do academic research that goes off to journal land. But at the same time, we get market insights immediately through our website deliverables to help.

[00:05:46] And so long story short. It's been such a joy to have that opportunity that I stayed. I still, of course, am very close with UVA, but now, full-time, I just conduct research on data and such. 

[00:05:59] Dr. Bravo: Oh, wonderful. [00:06:00] Perry, you work a few hats, right? 

[00:06:04] Dr. Perry Kaneriya: I do. I still practice clinical radiology many hours per week. But a few years ago, I said to use my medical degree and radiology skills in a different context. And so I returned to get my executive MBA from the University of Virginia, Darden. And ever since then, I've been doing some; I continue to do clinical work, but I've been doing consulting work on the side, working with various tech companies and startups that are working in the AI space to figure out ways to streamline operations, improve operational efficiency in a process that's inherently very inefficient.

[00:06:39] There's a lot of work out there being done by companies. And to really appreciate why this is such a there's so much potential in this field. You have to step back and understand the world of medical imaging as it is right now. Radiology volumes are going through the roof for a variety of reasons. The aging population with complex diseases increases reliance on [00:07:00] medical imaging, and the number of images per study has increased.

[00:07:04] When I started training Many years ago, a standard CAT scan would have less than 100 images, sometimes 200. But now, CAT scans and MRIs routinely take several thousands of photos. And so you have this massive amount of volume that grows every single year. And that growth in the book of work is outpacing the increase in the labor force, right?

[00:07:24] There need to be more radiologists entering the labor market to keep up with that tremendous demand. On top of that, radiology workflow inherently has a lot of inefficiencies. So you have this kind of perfect storm. All this is happening in the backdrop of all this new revolutionary technology in AI.

[00:07:42] Many companies are focusing on how we can use existing technology to make What is inherently an inefficient process more efficient. So, I've collaborated with teams working on multiple projects to figure out ways to empower radiologists to increase their efficiency and [00:08:00] ultimately enhance patient care.

[00:08:01] I have been in the trenches for many years. I know where those pain points are. I experienced them every single shift, and they need doctors like us on board to help them understand; okay, when I read a scan, this works well. But this has only worked well for a few years. And so these are the pain points we need to fix because they're leading to these bottlenecks. How can we use AI to mitigate these inefficiencies? 

[00:08:25] Dr. Bravo: You have an advantage because radiology went all digital. About 10 years ago, right? The transition has been 20 years, but analog films still need to be made.

[00:08:37] Dr. Perry Kaneriya: I haven't seen an analog film in, I'd say, 15, 20 years. I

[00:08:41] Dr. Bravo: I don't know if Barb knows about this, but, back in the day, you take an x-ray, And then they take the plate, and they go into a dark room, and it's smell, it's reeked of chemicals, and then you waited, and then boop, this thing would come out, and then you put it up, and you read it.

[00:08:59] That's how [00:09:00] I don't know if you were trained like that, but I was trained like that. Yeah. 

[00:09:03] Dr. Perry Kaneriya: I remember the first half of my residency was we used hard copy films, and packs had just come online right in the middle of my living. And that was around 2000, 2001, when we finally started getting digital images.

[00:09:16] That was transformative in terms of not having to wait for the film to get developed and dry, hang it on the alternator, and then bring it down. Plus, those jackets were heavy, too. I don't know, Herb, if you remember how serious a complete patient radiology is. 

[00:09:28] Dr. Bravo: folder was. Yeah. Or trying to find it, they have carousels with motors.

[00:09:34] Dr. Barbara Wixom: Yeah. You 

[00:09:34] know, and this is a critical point because, in the research up at MIT-CIRS, we talk about the difference between digitizing and digital. And often, you can't go on to these sexy new opportunities that involve AI, for instance, in digital strategies. If you're not first.

[00:09:53] digitized. And that's one of the reasons why the healthcare field, in general, has taken some [00:10:00] time to move forward with data monetization. For one, what you just described is a movement to being digitized to simply have data to work with. It's the same thing with electronic medical records, right? We had to move as a field to electronic medical records to even have the data digitized so that we could work with it. So that's a crucial distinction. And now that's one of the reasons why there's so many opportunities today: we're getting over that hump of digitizing.

[00:10:30] Dr. Bravo: So radiologists are way ahead of the rest of medicine; the electronic medical record has been, for the most part, a failure in that process. There may be hope that it will, with AI and LP, we will get there, but we're still doing our best to digitize all of this. And it's been an uphill battle.

[00:10:50] Yes. 

[00:10:51] Dr. Barbara Wixom: I was gonna repeat the one thing, just to some simple concepts for people to keep in mind is in the book, we talk about the idea of [00:11:00] going from data to insight to action, and then you can get to value creation and realization. When we speak about AI, we're at understanding. But you must complete the data.

[00:11:09] It goes data, insight, action, data. So we have to remember that again, it's great to be so excited about AI, but you have to have that piece ahead of time. And so the more we can connect that digitizing and the data to the insight and the AI, the more it will drive interest and commitment to things like electronic medical records.

[00:11:32] Yes. 

[00:11:33] Dr. Bravo: Barb, what is the goal of data monetization? What are we trying to achieve when you talk about that in your book, which is a phenomenal book? Thank you. 

[00:11:42] Dr. Barbara Wixom: Thank you. It's a special effort for me. And this is about 30 years of my research in a book. But for data monetization, you must remember that I'm a business professor.

[00:11:53] And when I was in DC for many years, directing the master's program, I would have all kinds of organizations, [00:12:00] including noncommercial ones, think of government nonprofits, defense contractors, as well as commercial companies, medical, different entities that have powerful mission statements around health and such.

[00:12:12] But as a business professor, my constant messaging is, I don't care what kind of organization you are. If you want to sustain operations over time, You are a business. Now, the way your revenues come in may be different. Maybe you're coming, getting your payments from grants or maybe from donors or whatever it is, but you're still an organization where you have to remain economically viable.

[00:12:38] So it's essential for all organizations, including healthcare entities, to think about data monetization, which is purely we have data, and that's our input. The organization's output has got to include some kind of financial returns. Otherwise, again, it could be a more sustainable effort.

[00:12:58] And so what's nice about them [00:13:00] appreciating this about data monetization is okay. We have to produce financial returns, among other benefits. From our data assets, what should we be focusing on? So, data monetization helps us focus on what matters to our organization and what we care about. We can get into many opportunities, but we have to think of it like a business in the medical field.

[00:13:24] So Perry 

[00:13:24] Dr. Bravo alluded to that first step in radiology. The digitizing of films cuts a lot of Human capital out of the equation and inefficiencies, allowing the radiologist to be faster at reading movies with higher throughput, which makes the patient happier because the less time you're waiting for the radiologist to read your ankle x-ray in the ER, the faster you get out of the ER.

[00:13:53] That makes the patient happier, but it also allows you to diagnose critical things faster [00:14:00] and start treating. So that was a big win. So we've been, we saw that there. I'm still determining if we're there in pediatrics yet, and I think we're still far behind. 

[00:14:10] Dr. Barbara Wixom: Perfect. If we wanted to sit and come up with use cases in the medical field, we could come up with quickly, just the three of us doing a quick brainstorming, probably hundreds in each space.

[00:14:22] The opportunities are pretty tremendous. And just one quick example: one of the facets of health care I've studied now for nearly 30 years is the medical spend space. It started with a case study I did on Owens and Minor in the late 90s, in their idea of having information about how much hospitals were paying for medical supplies and making that transparent.

[00:14:51] And as they digitized the distribution channel, the supply chain, basically from manufacturers to hospitals. [00:15:00] illuminating what used to be very opaque. Suddenly, there was this realization that, oh, my goodness, in this hospital, we're paying 700 for an orthopedic screw over here. And for that same screw in this side of the organization, we're paying 1700.

[00:15:16] It's crazy. And so just making transparent some of our core processes in health care. It is Amazing how we can remediate inefficiencies, for instance, poor work practices. 

[00:15:31] Dr. Bravo: The first step is measuring, and that's sometimes easier said than done. So, we have a mastermind group, Tableau, to measure.

[00:15:42] We're only at the financial performance of pediatric practices, but to be able to, from one CRM, look at the different KPIs. The practice owner must know I must pull four different [00:16:00] data sets and put them in a visual. That's what you describe in your book. That's the first step of data, right? It's just bringing all that data together and putting it so people can read it. 

[00:16:17] Dr. Barbara Wixom: And actually, I like to distinguish data from data assets. Data is everywhere. It's like water. Even the water is everywhere. I'm not going to go outside to the lake and take a drink of it.

[00:16:27] I will go to the water sources I know are curated. Drink safely. Same thing with data. There's data all over the place. It could be more impressive. My opinion. What's remarkable is that data is purposely created into assets that are teased up for use. So, in your example, Herb, if you have and we would want to have a very focused objective, what kind of outcome do we want to create from this data?

[00:16:54] And then how do we create a data asset that can [00:17:00] recurrently be used to generate that output. And so that's what many organizations in our research focus on today: what assets matter and starting with one or two. A quick example: we have a retailer with convenience stores all across South America.

[00:17:18] A lot of them. And the data asset that they started with is called dead net profit. And all it is, it's a big deal. It's skew-level profitability. For any product that they sell across their convenience stores, and once they have that, then all of a sudden you have the store operations able to set the appropriate store hours per store, you have sales and marking able to put together the proper advertising, you have supply chain able to negotiate the correct prices.

[00:17:47] So it's the same thing for pediatricians. What are those initially the core questions that need to be answered because of the levers that you need to [00:18:00] manage, and then how do we create a data asset that recurrently can help us monitor and be used for that? And then we can move on from there.

[00:18:08] Dr. Bravo: Yes. Having talked to you and Chris for a long time, I don't think any business survives without revenue. So, when you start looking at data, you'd have to look at income first. Quality is essential to maximize the margins, but you need revenue to be the most talented radiologist or pediatrician in the world. If there's no money for rent and paying staff, you don't get to see patients, and you don't get to use skills. 

[00:18:39] Dr. Barbara Wixom: Exactly. And it doesn't have to come in through pricing, charges, or even reimbursement. It could come in, for instance, from a grant or some kind of donor stream. That still is money in. So you have to think thoughtfully about how you can create capital for use for [00:19:00] whatever your operation is.

[00:19:01] Dr. Bravo: As I read your book, and when you create data assets... And it's beautifully segmented into the different things that happen in the data management journey. The first step is to have an investment you can use; you have someone who can present it to the person who makes the decisions.

[00:19:23] The easy money is improving efficiencies to improve your margins. And that's where the first step in internal medicine pediatrics family practice is. Perry's work on that from a different end, using natural language processing to build data models so that radiologists can summarize all past studies. And be more efficient as they read more films. How is that journey going for you, Perry? 

[00:19:59] Dr. Perry Kaneriya: Barbara, first of all, I [00:20:00] love that term data asset. And for the sake of consistency, I'm making sure I'm using the term correctly in the context of medical imaging. Is it inaccurate for us to call data-driven AI software a data asset?

[00:20:13] Dr. Barbara Wixom: So the data running it. Yes, exactly right. So, if you think about all the different sources potentially being drawn to fuel your actions, the data sources would be the data. And then after the processing happens and after it's in a state to feed whatever the application is that you're moving with, that's the asset state.

[00:20:36] The idea is to use the term data liquidity for data assets, which has resonated. Data liquidity is the ease of use and even the ease of reuse because, ideally, you establish a data asset. This particular purpose. And then, if acceptable, since you've put a lot of work into creating that asset, it could be used for other purposes in health care.

[00:20:59] Again, as long [00:21:00] as acceptable data use is a crucial capability because that has to do with the appropriate use, the legal use, compliant use, and ethical use of that data. But if all of the checks are checked off, in terms of that use, we can reuse and recombine those data assets across the organization.

[00:21:19] Dr. Perry Kaneriya: Herb, your question was, how do we use AI and data monetization in radiology? Is that 

[00:21:23] Dr. Bravo: correct? I was more interested in how you're starting to explore natural language processing and models. To improve efficiency. To improve efficiency. 

[00:21:34] Dr. Perry Kaneriya: As discussed earlier, much of our work is divided into radiologists. We have two components. We have a visual part of our work and then the text-based component. The optical element is self-explanatory. We look at MRI scans, review CT scans, et cetera. The text component is we review prior radiology reports, if available, review Prior medical records before interpreting the current scan, and then generate a [00:22:00] text output in the form of a radiology report.

[00:22:02] So, for both the visual and the text-based parts, tedious components can be easily outsourced to AI, for example, on the visual side. When we see a nodule or a tumor, we must provide measurements for the surgeon or the radiation oncologist.

[00:22:19] Is that hard to do? Not really, but it's very tedious, requires. I was curious when, before this discussion, I said, let me see exactly how many mouse clicks are needed to measure one tumor. And I was surprised because I just did it. Mindlessly, it's like second nature at this point, 16 to 18 click drag sequences to measure one tumor in three dimensions.

[00:22:39] So, imagine how long it takes for five tumors in somebody's brain, their lungs, the liver, et cetera? It's tremendously inefficient, so that can easily be outsourced to AI. But now, back to your question about NLP. So, the text component of our job also has similar tedious, mundane, labor-intensive things that [00:23:00] can be outsourced to NLP and AI software.

[00:23:04] For example, basic things like spellcheck. Editing, correcting mistakes, all the right, left discrepancies sometimes radiologists by error, say in the body of the report, they'll say right renal cyst, but then in the impression, in conclusion, it says right ovarian cyst. There's confusion; the patient's a male, but why does the radiologist say there's a cyst in the ovary?

[00:23:24] All that stuff takes time to ensure those things stay visible from the text or the system's cracks. It's pretty easy. It's just labor-intensive and very tedious. We use NLP in applications now to do all that and take care of all that tedium, but they're also more complex tasks that one radiologist can't do well at the level they'd want to do because time is an issue, right?

[00:23:48] And that goes back to a summary of text sometimes depending on the kind of case I'm looking at and how complex the patient is, how many body parts are being imaged, how many prior exams have been done on that [00:24:00] patient. I'll spend, You know, upwards of 10, 12 minutes, sometimes reading the last reports before even viewing the first image on today's scan because I want to get an accurate picture of what all the previous radiologists have said before me because, as radiologists, it's embarrassing when you don't comment on something.

[00:24:20] One of your colleagues noted, last month or six months ago, because then invariably, if you don't address that suspicious finding that somebody mentioned back, two years ago on a cat scan, you'll get a call from the doctor or the patient saying, Hey, by the way, you looked at a scan the other day. I still don't think you mentioned that nodule in the patient's lung on the scan from two years ago.

[00:24:42] Much of that happens because we need more time to sift through all those text reports. And find that one sentence where the radiologist two years ago said, Oh, there's a suspicious, two-millimeter nodule in the right lung. So having NLP-based tools and large language models to go through that large [00:25:00] volume of text and spit out an excellent, accurate summary consistently would be revolutionary in terms of helping streamline our workflow.

[00:25:09] It takes the emphasis away from spending all that time just reading through text. And many of these companies want to maximize eyes on screen time, right? They want to keep radiologists' eyes focused on the images because that's ultimately the most important thing. We have to make sure we detect all the abnormalities.

[00:25:28] We need to see all abnormalities to keep our eyes on the screen as much as possible. We need to have all NLP.

[00:25:35] Dr. Bravo: In your book, Barb, you would describe this as the most advanced form of data science. 

[00:25:42] Dr. Barbara Wixom: So for data monetization, this would be called improving where we're using our data to make work better, cheaper, faster. And so clearly, when it comes to reading, in the radiology world, making these processes better, [00:26:00] more affordable, more quickly, is a classic case of what we call improving with data.

[00:26:04] If you think about returns and financial returns to really, feedback to our missions and what we intend to improve represents about 51 percent of the returns we get from our data. And so you think about it, we want to ensure that our doctors do high-value-added work.

[00:26:23] And if we can have our systems help when we have very complex tasks and when we have functions that have a lot of nonvalue add components, we can offset that and really work with the data, that's a big win and a lot of financial benefits, which we can funnel back into other parts of our organizational businesses.

[00:26:43] Dr. Bravo: You walk the reader in your book through the data value creation process. Yeah. Yeah. Let's talk about that a bit because I carry ADD. To me, there's no purpose in doing anything if movement is not involved. 

[00:26:57] Dr. Barbara Wixom: So it's not about the data. First, of [00:27:00] all, it's not about the data. We need to take data, though. Data has to be our source. So, we must remember that we need data. So, returning to before, we must start with more than deriving insight. What are we deriving insight from? We start with data, then we have to move to understanding, so we have some kind of anomaly we've just detected through the radiology process or reading these reading scripts and such, and then moving to action.

[00:27:26] So what will we do because of what we know? But then, it's about more than just taking action. We want to make sure that there's value creation that's happening. Better health. We want to more quickly get to patients whatever they need for remediation. And then, ultimately, we need value realization, which is mission fulfillment, but really some kind of financial returns from what's going on.

[00:27:54] We have less medical costs because we've detected something sooner. We have less medical expenses. And so [00:28:00] there's less. Again, cost remediations that we have to incur and such. So, moving from data, insight, action, value creation, and value realization, that whole process must happen for data monetization.

[00:28:15] And so that's why we want to pick. You can like our focus. We want to have more than just having data everywhere and lying around. We want to put our data to work through that entire process. We also want to continue at insight, for instance, okay, great. We just learned that a patient has a high probability of some kind of severe condition.

[00:28:36] We continue, and we want to take action to create And then realize value again. So that's the whole process we like to think through as we consider different ways of putting our data to work. 

[00:28:50] Dr. Bravo: So in the action component, if I read your book correctly, there are two ways. One is through coaching, right? That's the best way of getting people to do things, holding them [00:29:00] accountable, meeting next week, and seeing if you met your goals. That's essential management. Even better is when you can automate. That action. 

[00:29:09] Dr. Barbara Wixom: Yeah. Yeah. The more you can ensure that the entire data insight, action, value creation, and realization, the more you can provide it all happens, and then you can move many times.

[00:29:23] Let's just take, for instance, a healthcare example. If we have, and this is an actual case we did in a hospital where there was, we would take the electronic medical record data, and we're able to predict. So, we had an insight into the likelihood of a patient fall. So we're in a hospital setting, know which hospital patients have a high probability of falling.

[00:29:48] Okay. This is all great. So what? If we stop there, we know some people have a high likelihood low. Okay. Then, we have to take some action, which would be setting up a context [00:30:00] within the hospital where there are ways to first detect when a person with a high likelihood of falling is getting out of bed.

[00:30:08] And then how do we make sure that we find the right staff to respond? And how quickly do they have to respond? And you can just imagine going on and on, on all of the change management and the process changes that have to happen in that hospital setting to put that insight. Of here's a patient with a high likelihood of falling, who's about to get out of bed and to actually remediate an action.

[00:30:33] And then, of course, the value creation. If we move all the way through, we want to get to the point of making sure our patient doesn't fall. So we have an increase in patient health of those who are rest Engaged in our hospital and then value realization, hopefully, would be things like lower fall rates and lower hospital costs because we're getting people out [00:31:00] because we're not keeping them longer since they just fell, things like lower fall rates and.

[00:31:02] You can imagine reimbursement. So we want to be thinking all the way through to value realization, which is a benefit for the patient and the hospital itself. 

[00:31:14] PCC empowers independent pediatricians. Streamlining daily operations and improving financial stability. A trusted pediatric partner for 40-plus years, we offer award-winning support, personalized training, seamless data transitions, and practice analytics. With inclusive pricing, a lively peer community, and a free annual user conference, you can focus on what matters the most: your patients. Explore more at PCC. Com, PCC. com. 

[00:31:52] Dr. Bravo: In medicine, it is not just the money. The patient didn't fall and didn't end up with a hip replacement in a nursing home for eight [00:32:00] weeks.

[00:32:00] Dr. Bravo (2): It's a life-changing quality of life for that patient. So a lot of what we do is not more than just money. It's the return on investment is not just dollars and cents. It's quality of life or, in some cases, you stay alive. 

[00:32:13] Dr. Barbara Wixom: And that's your first mission, and keep that mission, please, especially as a Patient. Yes. That patient, right? That being said, we need to sustain. And so if we have a poorly run setting that saves patients' lives, yet isn't feeding, goes out of business next year, then that's not going to make that happen. Doesn't help anybody. Yes. Exactly right. And so we need to, as we save patient lives in this particular instance, also make sure that we're doing it sustainably, meaning attracting more funding.

[00:32:46] We may be pleasing donors of a hospital so well. So we're getting more donations coming in. That's also a source of revenue, right? It doesn't have to be the price of care, for instance. And so [00:33:00] this is how we need to think in terms of the long-term viability of the settings we're establishing.

[00:33:06] Dr. Bravo: And Perry, in radiology, you're pretty far advanced, right? The AI models can now read complex MRIs and CT scans. Thanks. And bring up to the top the ones with critical abnormalities that the radiologist needs to review before the normal. 

[00:33:25] Dr. Perry Kaneriya: ones. Yeah, absolutely. That's an excellent example of how data and its technology can add value in nonmonetary ways, for example, for the patient.

[00:33:33] So what you describe right there is a situation is an AI application, which pre-screens through a whole list of scans that are sitting on my cue, for example, or any radiologist's cue and says it recognizes something that may be life-threatening. Critical finding that needs to be brought to the doctor's attention as soon as possible, and it will automatically bump that to the top of the cues and kind of flag say, Hey, Perry, you may want to look at this next head CT as soon as [00:34:00] possible, because there is a subarachnoid hemorrhage on it.

[00:34:02] The neurosurgeon probably wants to know about that sooner rather than later because the current workflow we read is in order of when the cases land on our queue. And I may have 25 points in my queue. And that 23rd case may be the one you. That's a critical finding that needs to be read next. If we don't have any assistance in terms of A. I. To help bump that case up in priority, that patient could just sit there for an hour and a half without a diagnosis. Yeah, absolutely. There's that. That's value created for both the radiologist and the patient. With a relatively simple solution, case prioritization is just an example.

[00:34:37] For intracranial hemorrhage they have similar applications for pneumothorax, aortic dissection, and pulmonary embolism. So, the list goes on in terms of use cases for that simple thing: case prioritization. How many points can you bump to the top of the radiologist's work list to save lives?

[00:34:56] Dr. Bravo: That would be an excellent example of presenting the data. [00:35:00] The machine prompts the action and creates significant value, Barb? That's what we do; that's the holy grail there, right? That's what we want the system to do. The person, the human, is working off their capacity, and the machines are trying to add value to it and decrease repetitive stress. An injury where the clicks and flipping through pages and reading the normals first until you get to the last one, which is grossly abnormal. Then, in your book, you go from there; there are three steps to data monetization. The second step is wrapping. Yeah. So yes, go ahead. 

[00:35:48] Dr. Barbara Wixom: I was going to say, so you can set, although we can come up with a million different ways to use data in the healthcare field, conceptually, there are three different ways to convert data into money.

[00:35:59] In a [00:36:00] positive way. The one is what we just talked about. We can improve work. We can make work better, cheaper, faster. The second thing we can do is create value-added features and experiences. Think about how we're using data analytics to improve our offering. So, in this case, instead of making work cheaper, better, and faster, we are encouraging whoever we're serving to pay more and stay longer.

[00:36:27] We want to acquire new patients if that's who our customer is. And so, Herb, you are great at data wrapping. And I really believe that. If you think about as a patient, for me with my daughters when we came to your office back in the day, and if I had to look across different pediatrician experiences, the question is there a data analytics feature or experience that distinguishes you specifically that delighted me more than others.

[00:36:59] [00:37:00] And your use of data at the time Absolutely distinguished you because I would come in and instead of having to fill out the same darn form by hand every single time, which still drives me crazy when I have to do it. You had us go up to a computer, and we would put in our name, and all of our information would already be there.

[00:37:20] And all I had to do was confirm it. Now. Fast forward to today, that's what's out there. But back then, it was not. And so that distinguished your office. And so, as a consumer, if you will, of pediatric services, I had to choose which doctor to see. Hands down, I went to you because the experience delighted me and made me feel more comfortable.

[00:37:45] And I just felt like you knew what was necessary for the care of my family because of the way you were using data and sharing that back. So that's what we call data wrapping. It's how can you use data analytics? In whatever [00:38:00] offering you're providing, in your case, it's pediatric services, how to use it to again help a patient or help a customer acquire, better use, or better create value from your services.

[00:38:15] Dr. Bravo: Thank you. So, I am very data-driven, but for me, it started more. 

[00:38:22] For me, it started more by looking at the customer journey. Yes. Yes. And I figured that in the neighborhood that we live in, almost everybody's mom and mom or dad or mom and dad work, and there's nothing more frustrating than being put on hold for 30 minutes when you got other stuff to do.

[00:38:44] The daycare is you gotta take your daughter to the doctor because she spiked a fever. And so I said, I can order a plane ticket at two in the morning to go anywhere in the world, and I don't have to touch a human. [00:39:00] And that, to me, is a much more complex transaction. How can it be that people can't just say, I got to see this doctor at six o'clock today, after I pick up Jenny from the daycare.

[00:39:13] I can't do that online without talking to a human. I got my appointment. I'm in. I'm done. And my anxiety level is down. I know that the problem is going to get solved not at two in the afternoon, but it'll get solved at six. That's great. And that was my idea to eliminate barriers for parents to be able to soothe their anxiety.

[00:39:37] And know that they could be taken care of during that same day data wrapping, 

[00:39:43] Dr. Barbara Wixom: I understand. You really will because that helps, so as a customer, as a parent with a child in your practice, by setting up that capability that lets me better acquire your services. So how can data analytics allow us to develop, use, or add value [00:40:00] from your offering, which helped me get in to see you for my children?

[00:40:05] So that was a huge value add. So, that would be a great example of data wrapping. 

[00:40:09] Dr. Bravo: It's very frustrating because of HIPAA and all the regulations; finding a practice allowing you to schedule your appointment takes work. The technology is all there, but they don't want to; they are just the liability.

[00:40:26] What if the person with a headache subdural hemorrhage picks a CT scan for six weeks from today, and by the time they get here, they're dead. You know what? That same person might never make the call, either. They're going to be quiet. But there are other uses, I think, for data wrapping, for example, in radiology.

[00:40:49] The only thing that radiologists do that's preventive care is mammograms. And so if you can remind the patient to come in for their [00:41:00] mammogram, make it easy so that they don't have to, there's very little friction. In addition, be ready to do the mammogram and, if something's abnormal, cue him right into an ultrasound and the biopsy that same day.

[00:41:16] If possible, that so much decreases a woman's anxiety when she has dense breast tissue, and we're not sure to come back in two weeks for the ultrasound. Dear God, if someone had told me I had cancer, I had to wait two weeks; I'd be dead from the scotch I was drinking before I got the second exam; it generated a lot of anxiety.

[00:41:39] And what holds us back there. Please let me know if I'm wrong, Perry, but it's one's the HIPAA; it's just that everyone is concerned about the HIPAA and the lawyers suing us if we open. 

[00:41:50] Dr. Perry Kaneriya: that door. No, Herb, you're absolutely right. The technology is there. It's been proven in so many other industries.

[00:41:56] It's no different in healthcare. Still, there are so many unique [00:42:00] hurdles and obstacles that we all have to deal with because of HIPAA, because of security legitimately, but that I think is one of the key reasons why a lot of these innovations haven't really gained as much traction in medicine, which is really unfortunate, right?

[00:42:13] We were in the business of taking care of people and saving lives and extending life, but yeah, All this red tape just really stifles the rapid adoption of stuff that's been adopted by other industries years ago. 

[00:42:26] Dr. Barbara Wixom: That's frustrating. One of the hopes with. My book is called Data is Everybody's Business.

[00:42:32] And that title was purposeful. It was to make anybody feel like the concepts of data monetization or the conversion of data into returns, that it's a topic. Then, anyone can absorb and apply. And so the more we start level setting and understanding this concept, which is excellent and reasonable.

[00:42:55] It's a neutral term. It's just it's a thing that we all have to become educated [00:43:00] about because of the proliferation of data. But even in the legal world and in the government world, we need everybody to understand this to make some changes. For instance, legislation and some of our regulatory requirements so that we can create the benefits that we should from these advancements.

[00:43:20] Dr. Bravo: Absolutely. And then on that day and then, on the section or data wrapping, you talk about the four A's, what are the four A's, which is a fundamental concept for anybody in business, but more and trying to monetize data. 

[00:43:33] Dr. Barbara Wixom: When digital started becoming a thing when you started having all of these digital companies, and when I see digital companies as web-only types of organizations.

[00:43:45] Traditional organizations, a hospital, a practice. I thought, Oh, we can develop digital features, too. We have a website. Let's like offer stuff. And the problem is that not everything we do in technology [00:44:00] necessarily results in value creation, much less value realization, and what the four A's are like a checklist to help us think about.

[00:44:11] What we create digitally, whether reports, alerts, features, websites, or whatever it is, will result in value creation and realization. So, for instance, the example I always like to give is back in the early days when banks' first kind of foray into the wrapping space was to provide people their spend breakouts.

[00:44:36] You'd get a pie chart. You may have had this with your bank, but first, you get a pie chart that says how much you spent. And so, at first, I was like, Oh my gosh, this is so exciting. I can see a pie chart of my spending. And then you start thinking, you're okay, so what do I do with this?

[00:44:49] Like what is and the whole point of data wrapping is you want. Whoever you're caring for, whether a patient or a customer, you want them to create [00:45:00] feel value. They need to realize value for them to spend more, pay more, and stay with you longer. You have to be incentivized in some way.

[00:45:09] If you think about the four A's, you think about a feature like a pie chart of your spending. Does that help the person anticipate? Okay. Not really. It's like a backward, like a backward on what you spent. Does it help a banking customer adapt? I could look at my breakout and adjust my spending going forward because I'm uncomfortable with how much I'm spending on groceries.

[00:45:35] Act, does it inspire action? Again, if I'm just looking at a breakout, my spend doesn't necessarily have that. Anyway, the bottom line is what you are coming up with. Data wraps feature experiences for your practice. You want to use these forays as a checklist to really understand.

[00:45:52] Is this really something that will be added value for whoever I'm serving? And so for you going back to your practice herb and [00:46:00] things like going in and pre-filling out. Forms that I usually would have to do on my own. That would anticipate what I'm there for and what I need to complete.

[00:46:16] It is adaptive because you're adapting the content to me as this is my information, not somebody else's; you can see how the four A's work, right? And so it's an excellent little checklist. This has been empirically validated, showing that the higher you rate the four A's, the more useful and engaging a feature it will be for whoever you're serving.

[00:46:41] Our research has gone on to say that will inspire a lift in selling more, paying more new acquisition, retention, and customer metrics. 

[00:46:55] Dr. Bravo: This is similar to the three A's, right? Affability, ability,[00:47:00] and availability. And not only should you judge data monetization, but even your management style by these four A's.

[00:47:11] That's a lot of A's. But you should, right? You should anticipate what headwinds are coming to your business, adapt, Take action, and assess if the movement achieved the desired result. And if you're constantly thinking of your business that way, it'll be more profitable than the other businesses in town.

[00:47:33] When Perry's working as a consultant with these companies, is this helping the radiologists anticipate, adapt, take action, and get better results from a UX experience? That builds value for his customers, which means you should get better margins for the company.

[00:47:57] And that was, I really liked those terms that you put in [00:48:00] there. I will wrap it up with a bit of how you break up stuff because I liked it. So you said there are five data management capabilities. Number one is data management. And to me, that's just like the foundation. It's 

[00:48:15] Dr. Barbara Wixom: what we were talking about before, Perry, in terms of, there's one thing to get into insight and using AI, but you have to have the data to feed that.

[00:48:24] If the data is terrible, you're at a standstill. You're only getting very far if you have the correct data underlying your actions with AI. 

[00:48:33] Dr. Bravo: And then there's the data platform. And what is the data platform in that chart? 

[00:48:39] Dr. Barbara Wixom: Data platform is you have to have the data on a technology that makes it cost-efficient to serve that data.

[00:48:49] Not only inside, but outside of the firm. Think about things like patient portals today, where you have to potentially release to your patients and such. And so, a platform is essential. Think about [00:49:00] Potentially running analysis on an Excel spreadsheet on a single computer in a doctor's office.

[00:49:06] That's really tough to scale or to do anything with if an insight is produced. And so today, having platforms that we can tap into systems that's really important for us to, at scale, start creating value with the data that we're analyzing. 

[00:49:26] Dr. Bravo: So I had to laugh when I was reading your book because you say the Vicentine spreadsheets and pivot tables,

[00:49:33] You're not alone, like what 95 percent of physicians are using to figure out where their business is going and who they need to reach out to bring back to improve the quality of patient care. And that's just so far back. And then the third thing is that the data science, which Perry was alluding to, where in radiology, they're actually now using [00:50:00] an I to improve not only the quality, the margins, but the radiologist is in a sense.

[00:50:08] They're not bogged down with. Repetitive stuff that doesn't bring value, a machine can do that, and they can concentrate on the higher thinking, the higher order stuff where all the value lies. Yeah. 

[00:50:23] Dr. Perry Kaneriya: And I think that's enormous value, Herb? Everyone's talking about physician burnout, people leaving the professions because they just can't. Could you do all this tedium anymore?

[00:50:30] If you can tell a radiologist, look, we'll get you through your cases faster. So you can get back to your family earlier in the day. Once you get home, you will do something other than charting or documenting during family time. Talk about tremendous value right there.

[00:50:43] And you improve patient, you improve physician morale satisfaction. That's huge. 

[00:50:48] Dr. Bravo: Yeah. Yeah. From the clinician's point of view, is it suitable? One of the most significant drawbacks to EHRs is that it has separated me from the patient; there are so many clicks [00:51:00], and your nose is stuck to the screen, and you don't have a conversation with the families anymore.

[00:51:06] If you do, then your burden would stand by spending an hour or two filling in charts. And the patient doesn't benefit from that. The doctor doesn't benefit from it. Over 50 percent of what's in EHRs today is garbage. It's just templated clicks somewhere in the little burb. I actually wrote Johnny's got an ear infection in the left ear.

[00:51:31] It does not look severe. He's not toxic. I want to see him back in weeks. All the other fillers. If there was a way that we could use what you're doing in radiology, but this gets in the weeds, so where the EHR was designed, there's so much structured data, and some of the narrative fields can't be accessed easily that it's hard to do that data review by a machine.

[00:51:57] But if we could think of it more as if [00:52:00] I was a lawyer and I'm writing in my yellow pad what. Barbara's telling me about the company's problem, taking notes, and then the AI takes my data and what's in the file and then helps me write a letter to Barb about what I'm recommending that she does.

[00:52:21] That, that's where we need to move to, but we're far from that in the clinical space. And that yellow pad remains there for me to look at the note next time because that's what I'm thinking. The all-formatted message with dear Dr. Wixom: I hope you're doing well. And please call us back.

[00:52:40] Our law firm's here to serve you. That's fluff. I don't need to read that every time she comes in, right? I need to remember what we talked about and what her strategy was. But we're far from that in the HR. And then this was also genius you, you spoke to about it in your book, which is every [00:53:00] organization needs to be customer focused.

[00:53:02] At the end of the day, we're here to understand the customer's journey and take out the pain points because the more you like me, the more money I make. And even if I don't care about the money, the more kids I get to make feel better. But if I need help understanding your journey. And I just want to do it the way I like it, then you won't allow me to take care of your kids.

[00:53:30] And, but I loved it that you put that in the book about data. Because people need to remember, at the end of the day, what do we want to do with all this data? If we want to do our jobs better. And my job is to take care of kids. So I want to see more kids and do it better. And make the parents happy.

[00:53:51] I want to make the kids happy if the parents are satisfied with the results. I call it a win. And [00:54:00] then five was very interesting for us. It's very complicated, but acceptable use of data is essential. Yeah, for us, it's very all the laws and all the stuff makes it very difficult.

[00:54:10] And any closing thoughts, Barry? What do you think about this conversation? 

[00:54:15] Dr. Perry Kaneriya: No, I think it's as we were talking; I was just trying to think of examples of wrapping in radiology. And I realize we're covering without even realizing we're wrapping things. For example, Barbara, you're talking in your book, and in the data are podcasts about increased credit card usage.

[00:54:32] When the customers realize that the value they're getting is protection from fraud, right? And so that got me thinking about cases in radiology. So, for example, radiologists often recommend a follow-up scan in three months or six months because we see something a little fishy. We're not.

[00:54:48] convinced that it needs to be biopsied or removed surgically, the safest thing to do. And we don't feel we don't feel comfortable dismissing it entirely. Let's retake a look in six months and return for another repeat scan. [00:55:00] But studies have shown that many of those patients don't return for various reasons. Many of those patients have yet to come back for multiple follow-ups.

[00:55:05] Either the doctor didn't get the report, the patient was unaware, or the patient was aware. They tried to call scheduling but got tired of waiting, the patient was unaware, or the patient was conscious and tried to call schedule but got tired of waiting 45 minutes to schedule the exam. Regardless, a significant percentage of those patients have yet to come back for follow-up, and it's problematic for many reasons.

[00:55:20] From a business standpoint, those are lost revenues. But more importantly, it's suboptimal patient care, and potentially, you're losing an opportunity to mitigate costs by catching an abnormality earlier in its progression than waiting until the patient has advanced metastatic lung disease, which requires millions of dollars of care, surgery, radiation, chemotherapy, et cetera.

[00:55:41] But so what they've done, some companies out there are working on pouring through the data and figuring and just go back to natural language processing and large language models using that technology to go through hundreds, thousands of patient reports and pull out the ones that need the follow-up.

[00:55:59] And then [00:56:00] contact and streamline the process automated so that those patients get text messages emails saying, Hey, by the way, your director recommended follow up in 3 months. It's been 2 months. You still need to schedule. You can click on this link, and we'll make sure to prepare you right away. So I thought the parallels to wrap.

[00:56:15] Yeah, that's a wrap, right? That's a good wrap. 

[00:56:18] Dr. Barbara Wixom: It anticipates, it acts. That's 

[00:56:20] Dr. Perry Kaneriya: a great wrap. And so your credit card customers feel that level of protection from fraud. These patients need to feel that level of security. Okay, if I stay with this health system or keep going to this radiology practice for my scans, everything will be clear.

[00:56:33] If I forget, somebody's got my back,  remind me. So I said, wow, that's it. That's a classic wrap. Yes. 

[00:56:40] Dr. Barbara Wixom: That's a great example. That's a great example. 

[00:56:42] Dr. Bravo: What are your closing thoughts, Barb? You have a phenomenal mind, and you're looking at the industry for a whole other level, right?

[00:56:51] And it's such a diverse industry. Please let me know what your thoughts are. I've 

[00:56:55] Dr. Barbara Wixom: I've been looking at this long. I, back in the, I remember working with the health and human [00:57:00] services back in the day when they had data palooza. This was back, oh gosh, I don't even know when that started years and years ago.

[00:57:05] My big message is that data shouldn't be complicated. And that whatever type of organization you're a part of, you just need to think about your most significant pain. Regarding improving, what's your most significant pain? Is it associated with reimbursement?

[00:57:24] Is it your cost structure? Is it how you're acquiring patients? What is that? Pick a focus, think about how data can, and then move from data to insight to action. Can value be created in that particular space really make a big difference, or for wrapping, if you think about who you're serving, how can you become more useful and engaging to whoever you're serving to help them better, either acquire, use, or create value from your services?

[00:57:56] And just starting there with one use case example in [00:58:00] can move you forward in your practice and your organization. And then suddenly, you've built, you've learned. And that's a lot of this is organizational learning. It's, we've learned. And now we can build on that.

[00:58:14] Now we can showcase what we've done to others to start building motivation for people to get on board and help move forward. So we must create that snowball and then push it down the hill, right? So we are making a difference in healthcare because I know I, for one.

[00:58:31] I want the healthcare space to succeed. Could you please help us? Anyway, our staff can help. Anyway, data can help. I want to be a part of 

[00:58:38] Dr. Bravo: that. So, I hear you saying that you don't have to be IBM, Apple, or Google to start using data to delight your customers or be more efficient. And a small project is just fine.

[00:58:56] Dr. Barbara Wixom: It's actually encouraged. It will [00:59:00] demystify; I hope it will demystify what moving from data insight action to value is all about. It is empowering to have some transparency about what this means so that more and more people and more and more leaders in medicine and healthcare can, Again, move us in the field forward.

[00:59:23] Dr. Bravo: Great. And neither of you thinks that we will all be out of jobs in five years because of chatbot. We're going to get busier. We're all going to get 

[00:59:30] Dr. Barbara Wixom: busier. 

[00:59:31] Dr. Perry Kaneriya: I think we're fine. 

[00:59:33] Dr. Bravo: I think we're fine. We will need more computers, it's been a phenomenal hour, both of you. I greatly admire what both of you do, and I always like having conversations with you guys.

[00:59:43] So, anytime you want to return and talk about something, I appreciate the opportunity to have another conversation. It was 

[00:59:50] Dr. Perry Kaneriya: a pleasure. I enjoyed being here. 

[00:59:52] Dr. Bravo: I'd love it. In the meantime, we'll put Dr. Wixom's book on the show notes, and I highly encourage you to read it [01:00:00] because it's very well written with a lot of experience, but it really takes your fear away.

[01:00:07] Approaching data is another tool that can improve your business processes. So many people are afraid of the data that losing that fear is the first step and is a crucial first step to improving your organization. I hope you have a wonderful day. Let me know as well.

People on this episode

Podcasts we love

Check out these other fine podcasts recommended by us, not an algorithm.