Transcript#
This transcript was generated automatically and may contain errors.
Hey there, welcome to the Posit Data Science Hangout. I'm Libby Herron, and this is a recording of our weekly community call that happens every Thursday at 12pm US Eastern Time. If you are not joining us live, you miss out on the amazing chat that's going on. So find the link in the description where you can add our call to your calendar and come hang out with the most supportive, friendly, and funny data community you'll ever experience.
Can't wait to see you there. I am joined today by our featured leader, Hansel Palencia, Manager of Data Science at DaVita International, and Hansel, it would be great if you could just give us a little introduction about you. Yeah, thanks Libby. It's great to see everybody today virtually. It'd be really cool to do something like this, you know, in person.
That'd be really, really fun. But anyways, yeah, so I'm Hansel Palencia, like Libby said, Manager of Data Science at DaVita International. DaVita as a whole is a Fortune 500 company. I work on the international side of the business, as the name would suggest. So this is 14 countries outside of the US, namely Brazil, Chile, Ecuador, Colombia, Panama, the UK, which is where I'm located right now in London, Germany, Portugal, Poland, Saudi Arabia, China, Japan, Singapore, and Malaysia. So those are our 14 countries outside the United States. I know them all by heart, unfortunately, because I've been to most of them many times. And, you know, as we were saying earlier, flying is not my favorite thing in the world. But you know, it's great to work in an international company and to participate in delivering high quality, patient-centered kidney care, which is what we do.
So we are a dialysis care company. First and foremost, we do offer services related to chronic kidney disease, which is kind of the more beginning part, if you will, of the kidney spectrum or kidney disease spectrum. And we also do some things around transplantation, and like in-center dialysis, and a couple of other things, as well as palliative care in a few countries. But really, our bread and butter is kind of in-center hemodialysis is what we really focus on. And my role today as manager of data science is being responsible for all of our clinical data workflows across all of our countries, reporting up to our chief medical officer, who's ultimately responsible for all of our patient outcomes, patient experience, safety, clinical care, and quality in all of our countries. And honestly, I can genuinely say with, you know, full-heartedly that I love my job. And it is so, so great to be able to, yeah, just be able to serve people and to try to be the best that we can be, because really what DaVita is all about is trying to be the best dialysis provider that we can.
And yeah, we use data to do that. And, you know, hopefully today we can talk a little bit about how I kind of, how I started in data science, kind of how we grew data science within the international organization, and then how we've used data to impact the lives of thousands and thousands of people around the world. So I think that'll be a really fun conversation.
It will. And I always also ask what everybody likes to do for fun. So do you have any hobbies? Um, sounds really bad. I, once I had kids, I kind of lost hobbies because my kids are still really young. So I have, yeah, they're my hobby. So I have, I have a five-year-old and a two-year-old and, you know, I work a lot being very transparent, but, you know, the free time that I, that I have, I spend with my, my kids and my wife. And I do, I do love spending time with my family and, you know, throwing my kids around in the garden. The other day, maybe this is my bad parenting, but there, they were playing football in the garden. Yeah. The real football, not the American thing. And it wasn't really playing football. It was my son running around and me trying to peg him as hard as I could with the, with the football. And that was, that was a lot of fun. We quite enjoyed, we can quite enjoy doing that together. So we, we like to, we like to play rough and especially my daughter, she's a, she's a, she's a beast.
So it's a good time. Lots of people in the chat here who are agreeing with you that like, yep, my kids are also small and they're all that I do. Yeah. Megan says my 3.5-year-old is feral. Yeah. They, that age is, is amazing. It's a lot of fun, but it is a challenge. Yeah. But it's a good time.
Understanding dialysis and kidney care
Amazing. Well, let's give a little bit of context building for everybody. First of all, I wanted to make sure that I understand what dialysis is. So our kidneys filter our blood and if our kidneys can't filter our blood, we need a way to do that. And I think that that's what dialysis is. Am I right? Yeah, totally right. So I don't know how, I don't know how deep into like the detail or technical bits of dialysis you want to get into, but essentially I think you're totally right. Our kidneys, our kidneys do a ton of stuff for us. So like, like you just said, they filter our blood. So we have a bunch of toxins in our body that are created when we're consuming food or water or whatever. So we have like these toxins, essentially just, you know, stuff that we can't absorb into our, into our systems. And our kidney processes all of that information and expunges, expunges that through, through urine, right? Through waste. But our kidneys actually do a ton of other stuff as well. So they manage our electrolyte balance.
They manage our bone density along with a few other organs in our body. They have a major part of our cardiovascular health as well. And a few other additional things on top of that. So what my, my boss, who is a trained nephrologist loves to tell people is that nephrologists are the best doctors because they look at the patient holistically, because you kind of have to, right? You can't just look at the kidney. You have to look at the kidney plus the heart plus, you know, the what's the system that creates hormones, that system, the lymphatic system, the endocrine system, the endocrine, the adrenal system. Yes, all that, all that good stuff with the hormones, right? So they have to look at everything, right? Because all of that impacts the health of your kidneys, right? We only have two. So you lose one. Hey, ho, you know, you have another one. But if you lose both, you're in trouble because your kidneys are not like your liver, they don't regenerate, right? So once you lose them, they're gone. And there's multiple tests that you can do, usually a blood and urine test, which is called your your EGFR. And I'm going to butcher this, but it's your estimated glomular filtration rate, which is essentially a measure of how well your kidneys are working. And that's what you do when you have like, maybe you're getting diagnosed for early stage chronic kidney disease for the very first time, you can go and do that, that blood test. And that urine test that'll tell you how well your kidneys are working. And then yeah, if you if you are found to have early stage kidney disease, they'll start hooking you up with a nephrologist start giving you medications to manage things like high blood pressure, which you probably have, if you have early stage kidney disease, things like diabetes, which you probably have, if you have early stage kidney disease, maybe hook you up with those GLP ones that everybody's talking about these days, those those MPICs, right? You want to manage your weight to manage your high blood pressure, your diabetes, so on and so forth. So it's very holistic, trying to treat the whole patient.
And we are we take the same approach at DaVita, you know, trying to treat the whole patient, not just get them through a session. And dialysis, the challenging part of it is, it hasn't changed since like the 1980s, or like the 1970s, if you or maybe even later, actually. So a regular hemodialysis patient, you'll sit in a chair, usually three times a week for four hours per session. And I'm sure that I'm sure many people here probably know somebody with kidney disease. It's it's one of the most common diseases in the world. In fact, I think last year was elevated to the number five killer by WHO, by the World Health Organization. So it's like a major priority in terms of global health. It's, I'll call it a quote, unquote, an epidemic. I don't know if it's actually classified as such, but it's quote, unquote, an epidemic.
It's, I'll call it a quote, unquote, an epidemic. I don't know if it's actually classified as such, but it's quote, unquote, an epidemic.
And it's really a first world problem. Well, maybe not anymore. But it is a first world problem, right? Because we're living longer, we're eating worse foods, we're not taking care of our bodies. And that's resulting in people reaching that stage where, you know, the we do so much damage to our kidneys, where they eventually die, we need dialysis. But yeah, lots of different flavors of dialysis. But yes, that's a long answer to your question, a rabbit hole question.
Data science at DaVita International
Well, I think it would really help us to get some context around the way that data science fits into DaVita International. And so if you could help us out with maybe an example of a type of problem that you and your team face in data science, the types of data that you see, like, are you looking at patient data? Or is this clinical trial data? Like, if you could help us give a little bit of a clearer picture on that?
Yeah, 100%. So I'm aware that a lot of people here probably in the US. So maybe this is this is going to be interesting, from a like a data privacy perspective, or a kind of just a data infrastructure perspective. So we, from an international level here in London, we don't have access to any medical record system in any of our countries directly. And that's very purposeful. So we have things like GDPR. And in Europe, we have LDLP in Brazil. And we have very similar legislation in like Chile, Ecuador, we have kind of very similar kind of Asian legislations around data privacy in places like China, as you would expect, of course, as well as in some of our other countries like Malaysia and Singapore. And so we actually have very strict rules on what type of data we can actually view.
So when I first started at DaVita five years ago, I was kind of coming in as like the very first data scientist in the door at the international business. And that was extremely challenging, because we didn't really have access to a lot of data. And we had to kind of go through all of these data privacy hurdles and all this like, try to figure out what we're actually able to look at what we're actually not able to look at. And so we ended up getting kind of an anonymized patient data framework figured out some some working on our data privacy policies internally. And that kind of unlocked the first set of doors for us to start answering the questions of like, well, you know, what are our lab outcomes for patients? So like, what's our percentage of patients with a hemoglobin value of 10 to 12 milligrams per deciliter, who are on erythroprotein stimulant agents, so ESEs, essentially. And so like, just like very simple questions like this, you know, five, six, seven years ago, we would have had to go to the countries and ask them for that information.
There wasn't really a lot of governance around how they calculated these types of metrics, you kind of just ask the questions and hope they would kind of understand what you were saying, you had a language barrier there, there wasn't a lot of technical experience. So you probably had a clinician doing this type of work very messily in Excel. And so we started trying to clean all of this up and streamline this, add some governance, add some like really simple, like bread and butter, right? Like dictionaries, KPI dictionaries, data dictionaries, some very basic governance. And that was really what DaVita International's quote unquote, data science was for like a year and a half when I first started, really just trying to get us to a point where we actually know what's going on. Not that we didn't know what was going on before, obviously, patients were being treated, patients were well taken care of. But we didn't really have like the data in front of us to say like, this is actually what's happening in Clinic X, this month at this time, right? So it was that challenge. And so we evolved that framework over the past four years or so, into something a bit more actionable.
So a question we might ask today is, how can we, what can we change in a clinic that would then impact a very specific metric that is session by session tracked for a patient to improve the quality of life? So they can have better say, like pain management, for example, which is like a very much more specific and I guess, a difficult question to answer then, you know, what is our hemoglobin, right? Because that's just something that comes from a lab, send some bloods, you get it back, you see the value, right? So now we're trying to look at things much more holistically. And so we've added a lot of toolage, a lot of systems specifically to do that. So we added like a new global quality management system, which tracks things like safety events. And there's some really other cool bits around like nursing quality and things like this. We implemented clinical data repositories to standardize our clinical data. And now instead of tracking a single patient in a month, we're tracking their outcomes over every single session, which is a big leap for us. The next step would be kind of machine integration. And then we can get like the actual blood pressure readouts from the machine and things like this, which we don't have today because of a variety of reasons, and so on and so forth. And so, you know, the past few years, we've built like a hospitalization model. So just like some really simple, like bread and butter machine learning, you know, can we predict the probability of somebody being hospitalized in the next month based on their outcomes this month, or over the last three months, we've built some kind of rule based safety flags. So if we have reporting of patient safety events, let's say we had five patients within a week with fever and chills. Is there a potential of a multi-patient outbreak related to catheter infections, for example? Can we combine our hand hygiene audit scores with those infection rates, or with our mortality rates, or with our hospitalization rate to try to understand, you know, are there any risky places, right? Is there any risk profiles that we can generate for our clinics to try to understand like, maybe we need to do some additional training here, maybe we need to refresh the policy there, so on and so forth. So trying to get to that next level of less reactive, more proactive, and that's really what it looks like today. And we have
some additional stuff as well, like some stuff that will be coming in the future around like AI and using agents and things like this that we hope to get implemented at some point this year, which is really exciting. But honestly, I feel like those things that like AI, it's like more administrative value. There's obviously a lot of guardrails today around using AI in clinical practice specifically. So it's much more administrative. But yeah, like at the end of the day, we're really just trying to be as proactive as possible, trying to detect like patient safety issues before they happen, or trying to see where we could potentially have issues with, you know, the nursing quality that we're providing, or trying to, you know, prevent death, prevent hospitalization wherever possible. So yeah, this is like, this is how we use data science in our day to day.
Well, thank you. It sounds like you guys have been on a quite a journey from you joining as the first data science person to now. And I heard you mention that you, you were like, this was disparate stuff happening in different different places, often by clinicians often using Excel. Rachel had a question that was in the Slido that is along these lines, and I would like to know too, Rachel, would you like to ask?
Sure. Hansel, I remember it was maybe a few months ago when you and I last chatted about this, but I think you were telling me about the way you had to convince the team to let you use Shiny over Excel. I don't know if I'm remembering it right. But I think you're like, you could let me do it this way. Or I could do it this better way and save 30 hours.
Yeah, so it was very much like on a case by case basis, right? So they were, they were sending all this stuff in Excel. And it was like, okay, well, we can, I can build you like an Excel dashboard, right? And it's going to be very static, it's not gonna be very flexible. Or I can build you this really nice, Shiny application, I can automate a bunch of stuff for like cron, like just on local machine as much as possible. And I can kind of show you the value of this. And so the context of this, to give a bit more insight was around mortality tracking. So we were coming, so this was end of 2021. And we were kind of going through our second COVID wave in the US.
And so my my boss needed, like a very easy way to look at our mortality data versus so who is dying just like from natural causes versus who's dying to COVID. And I'm sure that that definition of who died due to COVID is very, honestly, very political, which is quite, quite sad. But, you know, that was, you know, I think the definition at the time was anybody who was who had been diagnosed with COVID within two weeks of death or something like this, which isn't isn't perfect, but it was the best that we had at the time, right. And so we had that definition, we had some nice stats that were coming like from our finance and ops teams. And so I started collecting that information, and I presented it. And of course, as a lifelong R user, I used R.
And I produced some really nice graphics and kind of was like, hey, look, like, here's a this really nice plots and like some SPC charts and statistical process control charts, which is using like the NHS, the National Health Service in the UK a lot, and kind of just showed like kind of what the tool could do. And then I said, you know, like, we could build a really nice dashboard, using something called Shiny, it won't cost you anything, I can do it all locally, let me just show you kind of what it looks like. And so I built it, everybody was really excited about it. And then I said, you know what, I could do even more. If we invested like 30,000. And at the time, it was I think it was RStudio team, because it was still RStudio or something like this. And so that's how I kind of got our initial our initial funding for tools. And so we started like very, I think we had like, it was like two dev licenses, one for one for usage, one for like testing and stuff like this. And we had like, like 20 or 50 user licenses.
And that's how it started. So I built some stuff that added some value. And then we built like a medical outcomes dashboard, we built a clinical care dashboard and a safety dashboard. And that was like our main kind of that we saw there's only those four things that was our main dashboarding suite, probably two and a half years. Just because as soon as we built that, it added so much transparency to how we had how we were looking at our medical outcomes, that the focus then turned to two other things specifically around like improving our ETLs like throughout our entire organization, right, so like going to the country. So then, so then I started to travel, which was going back to the beginning of our conversation, both good and bad.
So like, that I had to, you know, go to Brazil, and like, why can't you guys give me this data set that I need. And this is actually a really embarrassing and funny conversation. So I, this is actually happened, I went to Brazil, and we had like this template of data that we were asking all the countries to provide. And Brazil kept saying, Oh, no, we can't do it, we can't do it, can't do it. And then we went and I sat down with the team. And I was like, guys, why can't you guys provide this information? And they're like, Oh, no, we can, it's in this report. And they just pulled it out. And I was like, I had to fly all the way from Brazil, or from the UK to Brazil, and sit down next to you just to get this piece of information. And it was like, it was just miscommunication and stuff like this. And it was extremely frustrating. But it taught me the importance of like teams talking to each other, and really just like, how essential it is that everybody stays on the same page with with with digital transformation, right? Because sometimes, for executives, like, they don't even know what information they have. And they'll just assume like, Oh, yeah, this seems really complicated, right? But when you actually dig into the details, it's actually not.
Where inversely, sometimes they'll say, Oh, that's really easy. And then it'll be like the hardest thing. Like, it'll be like getting water from a stone, literally. So I don't I don't know what it is. Sometimes it's like we tend to, we tend to underestimate really easily or overestimate really easily how things will go, which is I think is probably a good analogy for life.
But yeah, so like, you know, that transformation, like you said, happened over time that that investment happened over time. And today, like, we don't have a massive budget for our stuff, frankly. But we get what we need, because we provide, we provide value to the organization, right. And that's through a variety of ways. In the beginning, it was very much a clinical lens. But today, it's it's across, it's across lanes. So we provide a lot of value to like our finance and our operations teams, we do a lot of really cool exploratory research on outcomes that are that are very at the forefront. My boss recently spoke at the World Conference on Nephrology in Yokohama, Japan, on some of the research that we're doing around like new types of treatment for dialysis care, which show like 20% 30% reduction in mortality and hospitalizations for patients. And yeah, like we've shown massive value in terms of just our clinical outcome.
So over the past five or six years with this, it's not just data work, it's data work plus people going out and actioning, of course, right. But we've been able to reduce underlying mortality by 27% since 2021. And we've been able to reduce underlying hospitalization rates by I think, around 10% since 2021. So, so really, really big improvements and kind of just like keeping people alive, which is what we're really after.
So over the past five or six years with this, it's not just data work, it's data work plus people going out and actioning, of course, right. But we've been able to reduce underlying mortality by 27% since 2021. And we've been able to reduce underlying hospitalization rates by I think, around 10% since 2021.
And, but then obviously, the step further from that is how can we make that life worth living, right? Because these people are with end stage kidney disease, it's like really, really tough. And so a big, a big new piece of our framework that we introduced last year was stuff around like patient experience. So this is your patient reported outcome measures and your patient reported experience measures. So things like, you know, how do you feel about pain? You know, do you feel like you have a good relationship with your physician? Do you feel like you understand your the treatment you're receiving? Do you feel like you understand? You know, do you feel like you receive education on your on your on your disease? So these types of things that sometimes like, frankly, are overlooked. These are really, especially right now, are really, really big points for us, because we want to make sure that we're not just keeping people alive, but that we're giving them a good quality of life.
And so, and so yeah, there's like massive amounts of programs that we're running. Yeah, to trying to trying to support that work. So like art therapy, exercise during treatment, you know, bringing like animal therapy, like Brazil brings dogs, like these beautiful golden labs into into clinics to like play with patients and stuff. So, so really trying to think about different ways that we can make that, you know, bingo during sessions, because they're there for a long time, right? Really trying to think about how we can how we can care for our patients in a in a better way, essentially. Yeah, well, you're there for four hours, I think it's the quality of the experience is a big deal.
From pre-med to data science
Well, I wanted to ask you a personal question, because I'm sure that your your hopeful background when you were looking into school was pre-med, does this sort of satisfy some of your, your pre-med desires? And could you just talk a little bit about how you got into data science from that pathway?
Yeah. You know, like I said, at the beginning, I love what I do. And hopefully that's that's coming across like, I've been told I'm a passionate person, which I very wholeheartedly believe as a Hispanic, you know, this is like, this is our people, right. But, you know, I really do love what I do. And growing up being a physician was always the plan. And I told Libby the story, I'll share it, I'll share it now. So I lived in Brazil for two years, I did two years of school before that. And then I came home after after Brazil. And I did, I finished my last two years at my four, at my kind of four year university. And I remember like that first week of school, I met with an advisor at the university, and I kind of sat down and I was really excited, I had all my classes planned out, I had, I was starting, you know, my classes in, like, I was taking like a biology class and a chemistry class and a couple other things. And I was super excited just to get on with life and kind of do what I had always planned on doing.
And I sat down in that chair, and the advisor looked at me, she said, you know, Hansel, I just, I really just don't think you're going to make it to medical school. And I was like, what do you mean? And she said, well, yeah, you know, like your grades, you know, before, you know, those in your associate's degree, you had like a B here in biology, and really, you need like an A plus to, you know, and it's like biology 101 class, you made like a like a C. And like, I think it was, I think I took like chem two in my associate's degree, I made like a C, because I just didn't care at the time, for some very silly reason. And she's like, yeah, these grades just not very competitive. I just don't know if you're this is like the path for you. And that kind of like destroyed my confidence.
And I was additional to all of my, to all of my classes, I was also taking an intro to stats class, which was required for the for the premed degree. And I went to my it was like, immediately after the advisor session, I went to that stats class. And there was a, the professor's name is Jay Hathaway. Really, really amazing, amazing, amazing guy does really great work with the with the Bill and Melinda Gates Foundation. If you're located in eastern Idaho, BYU Idaho is always looking for like companies to do projects with and they run like a data science society. And I was, you know, the whole really, really great program over there. And he was starting the program at the time. It was like a very new thing. I think he was just starting his career as a professor, coming over from INL, so Idaho National Laboratory.
And he said, after class that day, he came up to me, he said, Hey, Hansel, you know, like, it seems like you kind of like like this, you know, I was asking a lot of questions, and I was really engaged, which I think most people aren't when you hear like stats, right. But I found it really interesting. And he said, you know, we opened this new degree called data science. And I was like, I have literally no idea what this is, like my parents are dancers. Like my goal was like, I hate math, or I didn't hate it. But I was like, I never focused very well in it, because it was never it was never the goal. But I was always like decent at it. And so I was like, Okay, that's really interesting. I was like, well, you know, I don't have like a plan anymore, right? Because I've just been told that I suck, apparently. So why don't I, you know, why don't I just switch my degree? So he said, Why don't you come by my office hours at the end of today? And we can have a chat about it. So he Yeah, I went to his office. I was at the end of the day, I sat down in his office. And he said, Okay, well, give me your laptop. So I gave him my laptop, and he's installed R and Python onto my machine. And he kind of showed me some stuff like built a really cool graph. And I was like, all right, I'm sold. And so that very same day, I switched my major from pre med to data science with a minor in stats. And I never looked back. So that was like, I had literally zero knowledge of data science, I was like, zero knowledge of statistics, this was never the plan whatsoever.
And so that very same day, I switched my major from pre med to data science with a minor in stats. And I never looked back.
But I found something that I was that I really enjoyed. And I was, and I was fairly good at I honestly won't say that, like, I mean, I wouldn't dare to call myself a statistician, like, I'm very comfortable with statistics. I was trained by statisticians, but I wouldn't dare call myself one.
Being very passionate is one of your skill sets here.
implementing a new medical record system in one of our countries. And that is extremely challenging and extremely time consuming. And so I'm offering a lot of technical expertise on not necessarily the technology itself, but more around how do we look at kidney care? How do we make sure that the EMR itself has all the functionalities that are required to give good kidney care and stuff like this? So you might not call that data science. It probably isn't, to be honest, but that's kind of what my role has evolved into on top of everything else. So some days I'll be building like a new data package for our internal data scientists. So we have, like I was mentioning earlier, this quality management system. We're pulling data via an API. We extract all of that data. We build internal functions they can then use to query that information directly from this third-party database. Or we'll build specific data sets that are able to be used to calculate all of our internal KPIs based on our data dictionaries, which we like much more than them doing themselves because obviously then there's an additional governance layer there, right? Because we are using our stuff rather than just kind of doing something on their own. We might do like building applications, building dashboards. We might be building machine learning models. We might be doing data governance. I've done some very simple data migration when we're doing another project related to an EMR switch in one of our countries, medical records system switch in one of our countries. So my job varies. Some days I'll be a data engineer. Some days I'm very much a manager. Some days I'm like a project manager. It varies significantly.
Kind of a jack-of-all-trades thing, master of none. And I think just like, I know that answers the question, but just my personal thoughts on this. I think the way that data science has evolved today is that we've become very specialized. And so you have like one person that does one thing. And I think that really works in large organizations as a whole. At DaVita International, we're extremely lean. And so that doesn't work for us. And I think that it's much more common for that to be the norm than to be the exception. I think it's much more normal for you to be a bit more hands-on in very different parts, especially like if you're working in a small business, you're going to be with hands in a ton of different pots. And honestly, the value that a data scientist can bring to a small organization, I think is much larger than going and working for FANG, for example. The impact you're going to have in a FANG organization is going to be so small in the grand scheme of things. Whereas if you go and work for that mom-and-pop shop around the corner, just the amount of value they can bring into those people's lives, I think is enormous. And I think that gets overlooked a lot in our space. We tend to try to push for the big startups or the FANG or this and that. But there's tons of work that I think we can just create ourselves if we approach small business and say, hey, look, this is, again, going back to the demo, this is what I can do for you. Pay me a salary. It doesn't have to be a FANG salary, right? Pay me a salary, and I will add value to your organization. And just almost go and find jobs yourself. And I've actually done that once as an undergrad, just as a part-time thing while I was in school. But I was like, look, this is what I'm learning in class. This is the value that I could add to the business. Maybe you have data around this. Maybe you have data around that. And they got me a few meetings, and I eventually got an internship doing some stuff with a local business in Eastern Idaho. And it's literally just that initiative, that productivity to go and say, there's a lot of opportunity maybe in places that we aren't thinking about. And you don't need really expensive kit. You don't need a really expensive tech sack to do really meaningful work, I think, is the takeaway from that. You can do really meaningful work with some really simple stuff that adds value to someone. And you can get paid for it.
And honestly, the value that a data scientist can bring to a small organization, I think is much larger than going and working for FANG, for example. The impact you're going to have in a FANG organization is going to be so small in the grand scheme of things.
Amazing. I am putting a little description in the chat about what FANG is, because we said it a lot, but we didn't describe it. Oh, sorry. FANG no longer really exists, because Facebook is now meta, but it's like Facebook, Apple, Amazon, Netflix, Google. So big companies with really, really big budgets who were really scooping data scientists up by the handful in 2021, and then everything changed. Well, we have so many questions in the Slido, and I would love it if Russ could ask his live, Russ Conti. There were several comments about this, replies about this in Slido as well. Everyone making sure that they had lots of water. All right. Take it away, Russ.
Kidney health: what the data shows
Okay. Can you hear me? Yes. Yes. I had kidney stones several times when I was younger, and I was told it was the same level of pain as giving birth, which obviously I've never done, because they had to give me the same level of pain meds. I eventually learned that the solution was drinking a lot more water. I have not had kidney stones since. What does the data show predicts healthy kidneys?
Well, this is a loaded question. This is a very loaded question. I have to be careful not to give medical advice because I'm not a physician. This is not PSA. This is not medical advice. We totally understand. But the data does show that exercise and water are the two biggest things that you can do to prevent kidney disease. Exercise, because you're managing things like hypertension, so high blood pressure. You're managing things like obesity, because that's a major cause of hypertension and diabetes. Literally, it's diet, exercise, and water, so fluid intake. Those three things are pretty key. But obviously, you can get kidney disease from other things. Again, I'm not a physician, so I'm going to butcher this a little bit, so bear with me. But there's cases in Latin America where it's environmentally caused kidney disease, because the way that we eat and the way that the environment impacts people through pesticides, very rural communities and things like this are causing kidney disease today, which is really interesting. Obviously, massive amounts of toxins to your body can cause kidney disease, so make sure you stay away from those, I guess. But yeah, diet and exercise are going to be your major influencers if you have kidney disease or not.
Fantastic. Everybody in the chat is grabbing more water right now, thinking about how they can get exercise into their day later. Well, we had another question that was from Notabecco. It has an asterisk next to it, just like Natasha's did, so I will ask it. And it is, how do you measure subjective metrics like degree of pain for the pain management protocols that you mentioned? That's a really great question.
Measuring patient-reported outcomes
So there's a really great question. So we have standardized surveys that are internationally accredited. Maybe not accredited is a good word, but they're used internationally. These, like I mentioned before, like PREMS and PROM, so patient-reported outcome measures and patient-reported experience measures. And I'm sure that somebody that's really techy and great can link those surveys in the chat, because they should be openly available on the open internet. But there is like specific kidney-related PREMS and PROM surveys that are offered internationally, especially in the UK. The UK is a leader in this. It's one of the reasons why we action it, right? Because it's like a very UK thing. And so then we're obviously sharing that experience and knowledge with all of our other countries that maybe don't do as much around this.
And really, I guess that reaches a good point. Not every country is the same. The quality and standard of care in every country is different. But really what we're trying to do at the international level is strategy and governance, right? So what is one country doing really, really well that we can then take to other countries to enhance the level of care and quality of care and experience the patient's receiving in those other countries? So the UK is really good at PROMS and PREMS. We saw that and said, hey, PROMS and PREMS, is it something they do in Colombia? Why don't we implement PROMS and PREMS surveys for all of our patients in Colombia every year so we have a view of things like pain? Sleep is another really big one. So like quality of sleep. Specifically, it's pain, but it's like access site pain. It's like where you're actually getting hooked up for dialysis. You need a special type of vascular access that connects directly to your blood system, to your bloodstream. So like that specific site pain, education around that, it's a really big part of PROMS and PREMS. Things like, yeah, education from your physician, how well do you feel like you can talk to them, and kind of those types of things. And that's on a like heart scale, like most qualitative things are. So your bread and butter, one to five, like heart scale, five being the worst, one being the best. And then, yeah, the hope is one day I would love to get like some nice open text information from all of our patients of like, how do you feel today? And then they can tell us exactly how they feel, and then we can, we can do, they can tell us, oh, the dialysis chair really sucks, or the snacks are terrible. Maybe we can, maybe we can do something about that.
Amazing. Well, we've had, we've had lots of things that I want to just point out from the chat to let you know, there are lots of people who were identifying with your story of either being pre-med or wanting to do something completely different and ending up in stats, sort of like, oops, now I'm in stats. And I had this, I had the same story where I was a finance major and was looking at the finance degree going, wow, I hate this. I hate all of it. And I was in my stats class and I just loved stats. I was in business stats. I loved my stats professor. And I, yeah, after maybe like a week, I was like, yep, changed my major. This is what I'm doing for the rest of my life now. And these are my people. And then there was one suggestion that I really liked from Connor. Oh, I lost it because the chat moved so fast, but he said that his wife recommends making one good decision between your bad decisions. So like drink a glass of water between your mugs of coffee. I really like that. If you can focus on adding something in between your, your things you're already doing. Well, we had another question about some sort of predictive stuff and I figured we'd do that and then we'll hop into, we'll hop into maybe like a career oriented question for you. And Zach had asked, does a history of kidney disease also potentially predict other health conditions later in life?
Kidney disease risk and predictive factors
Yes, it totally does. It's usually the other way around. So it's usually a history of cardiovascular disease or a history of diabetes. This makes you more indicative of, of being at risk of kidney disease. There is something called, I'm going to say tangry and I could be, I could be wrong. I think it might be tangry or kangry. Tangry is KFRE. So kidney, what is it? Kidney failure risk equation, I believe. Yeah. Tangry is KFRE. And this is essentially a measure of how likely you are to get kidney disease based on a very simple metric. So it's like, do you have any type of cardiovascular disease? Do you have diabetes? What's your age, what's your gender? And then something else. I think there's a, there's a short version and there's a long version. Now, if you are going right now to go and look up your risk of kidney disease, PSA, it is not, it is not a perfect system, right? It's, it's just a, it's just kind of a recommendation thing. There's a ton of stuff that obviously influence your risk and some of it's genetics. And, but a lot of it's lifestyle, frankly, it's, it's much more likely lifestyle than it is, than it is. Well, obviously if you're like 100% predisposed to kidney disease, then that sucks. But it's, it's much more likely to be lifestyle than anything else. That's good to know. That means that we have some control over it. Yeah. Well, a hundred percent. We, we have, we, we have tons of control. Like, and again, if you're, if you're 90 years old and you end up with kidney disease at 90, I think you've done really well. Right. Yeah. Now, if you're, if you're 40 years old and you have kidney disease, I think you're, you probably need to make some serious changes. Hansel, how can you say that on my 40th birthday? Oh, sorry, Libby. Hopefully. Yeah. I'm sure you don't have kidney disease. Fingers crossed. No.
R packages worth knowing
Okay. Well we have officially 10 minutes left. I want to see if we can get two questions in here because I saw one from Tim, Tim Golden, that I'd love him to ask live. And then, and then one from me. Tim, are you available to ask your question live? I know your camera was blurring you out, but you're back. I see you clearly now. I don't know why, I guess I got more interesting. Yeah. My question is, do you, are there any R packages out there that you feel like people are sleeping on and that deserve more press? Well, it's a great question. And what I was not prepared for whatsoever.
I don't know about sleeping on, but I love me some good patchwork. I think patchwork is excellent. So patchwork is really good. Oh, I love patchwork. Patchwork is amazing. I think there's some really great functionality out of the M365 Office R package. So we've done, because like people in business love PowerPoints, I don't get it, but we've done some really cool stuff around automating PowerPoints with M365 Office R package, which I think a lot of people have never touched it probably. And some really cool workflows with that.
I think that those are perfectly wonderful suggestions. And just as a sort of like, oh great, Isabella, she's got the Microsoft 365 R in there already. And just as a reminder, we're having all of these conversations about medical stuff. If this is something you're interested in, even as a pivot, or if you're a student, I cannot recommend enough the R Medicine Conference, which is coming up in that first like week of May. And it's always so good. Even if you
are not in medicine, it's a fantastic place to be. I'm not in medicine. I attend it and love it and get so much out of it. So highly recommend. Andy, I see you have your hand up. It's been a while since anybody raised their hand. I want to make sure that it's correctly raised.
Yes, it is. Hi, I'm with Oregon Health Authority. And just on a quick plug in that R in Medicine Conference, David Robinson, author of Voices in the Code, is one of our keynote speakers this year. And that book is about transparency in algorithms for the kidney transplant system. So if you're interested in this very topic, register for the R in Medicine Conference. I just dropped the link in the chat. Thank you so much, Andy. Andy, we haven't seen you in a while. I haven't seen you in a while. I've been around. I've been busy. Okay. Well, public health, you know. With that, with our seven minutes left, let's ask Hansel our favorite question, which is, is there a piece of career advice that you find particularly meaningful or something that you wish you had known going into your data science journey? Maybe something you wish you had known before you'd started at DaVita as the sole new data science person at DaVita International.
Career advice: embrace the journey
Yeah. Embrace the journey. It sounds so, so cheesy, but really just embrace the journey. Like, love what you're doing. I think one of the reasons that I feel like I've been successful at what I do is because I really believe in what I'm doing. And like, with that, if you aren't loving the journey, because for whatever reason, you know, it's okay to, it's okay to change. Like, you're not quitting. You know, I worked, I worked in the National Health Servants coming out of my, my master's degree for like almost a year. And although it was really great work, and it felt very meaningful, it wasn't impactful enough for me. It was, it was very slow. It was very bureaucratic, as you would expect of a government agency. And it just wasn't, it just wasn't enough for me. I didn't feel like I was doing enough good. And so I changed jobs because I was like, well, if I'm doing, I know I can do, I know I can do more. I know I can do really amazing work. And it's just, it's not cutting it for me. So like, love the journey. If it's not doing it for you, you know, make the change, make the leap, you know, and I think if you're getting ready to do that, or want to do that, you know, start, start networking, you know, and I go back to the analogy of like the planting the tree, best time to plant a tree was 10 years ago. Second best time is now. Right now. Yeah. So like, you know, just go and go and start building those connections, building that work, you'll never know when you need it.
You know, I've leveraged a lot of the network that I've built over my, I'll say short career, because I'm young, over my short career. And that's been an amazing experience and has allowed like a lot of friendships to develop, right? It feels sometimes I've heard it said before, that like networking can feel kind of dirty. Because you know, you're like, you feel like kind of just getting to know somebody to use them in the future. But yeah, you know, but you know, I think I think the genuine friendship can come out from that. And you can you can really develop great relationships and friendships from from building a community. And I think part of that is like stuff like this, right? Like data science Hangout, you know, you guys know, I'm sure people recognize other people on the call. And then when there's something that comes up, they know they know who to reach out to, right? So Oh, absolutely. Especially if you get on the discord. I mean, we're all I would trust all of you with my life. I love you all so much. And I, I think Hansel, it's a great example.
Also, how you got your job at NHS, I think, because you we talked previously about the sort of golden opportunity that you have while you're in school to be networking and get people to say yes to you because you're like, Hey, I'm a student. I would just love to hear about what you do. It's a little bit easier. I want to talk a little bit about how you did that.
Networking as a student
Yeah, no, I think I think just like, like what you're saying, as soon as you say, Hey, I'm a student, like doors open, like magic. Yeah, I don't think I've never had I've never said no to a student approaching me and saying, Hey, I just want to have like five minutes of your time, I'll always have time for a student, because obviously, I was I was once a student, I understand what that's like. Similarly, I'm sure many people on this call understand, understand what that's like, is like, you know, being a student trying to find a job is very, very challenging. But even outside of that, right, just asking people's time, it's, it's hard to do. But honestly, I think everybody tends to be really kind, we probably work it up a bit more. But yes, the story. So when I started my master's degree at the University of Exeter, here in the UK, obviously, coming from the US, I had just gotten married, my wife was pregnant.
And I was starting this MSC, and I had like a set amount of savings. And I was like, Alright, I have like, x number of dollars or x number of pounds, I have to pay for my degree. And then we have like this much to live, we're in, we're in university housing. And otherwise, we have like no money, we have literally no money. And we were on a very tight budget was middle of COVID. And yeah, I remember it was like this, this really convenient space between the first and the second wave of COVID, where I had like one or two months of in person classes at at the university. And so I got to meet all of my MSC cohort, which is really great. Because we've like been isolated for the past, like, six months or something like this. It was like kind of the first time like, you know, fresh air. So we we went to that first, that first class. And we all introduced ourselves and some guy in that class was like, Oh, yeah, I'm a manager of a data team in the NHS here in here in Devon, which is like the, the county or the quote, unquote, the state in the UK, where we're based in. And I remember, I was like, Oh, that's, you know, the manager, he probably hire somebody, right? I was like, I remember after the class, I went up to him and said, Hey, you know, I'm Hansel. Yeah, I'm Hansel. It's really great to meet you. Just so you know, my wife is pregnant, I have no money, and I'm looking for a job. And in the UK, it is very unusual for students to work. It's it's kind of frowned upon almost because the focus is quote, unquote, school, and you should be focused on school. But obviously coming as you know, this idealistic American who knows nothing about Europe, I didn't really care. And so I kind of pushed that was like, yeah, you know, like, you know, if you something opens up, let me know. And he's like, Yeah, yeah, sure, whatever. And so over time, I got to know him a little bit better.
But every time I saw him, I said, Hey, don't forget, you know, spot opens up your team, you know, let me know. And he's like, Okay, okay, okay. So come about like a month or two before my son is born, I get a message from him saying, Hey, Hansel, just let you know, we're hiring for an analyst this year, you should apply. And so I applied went through the process, and I got the job, literally, I think it was like two weeks before my son was born. And then I went on paternity leave. And we have a month off for paternity leave in the UK, which was lovely. And it's full paid. And so I got like a full month with my kid, with like really nice benefits. And obviously, you don't pay for anything, because it's, you know, social health. And so we had like, no debt from like pregnancy or childbirth. And it was just like amazing, amazing experience. And then at the end of that, I had this really great job that I could just dive right into. And so that was, that was an amazing experience. And literally, just because I was like, I was really annoying. And I think introduce yourself to people.
Second best time is now. Right now.
Well, top of the hour, so I have to be annoying and say, we have to be done, which is awful, because I, as I have said before, I would talk to Hansel for hours and hours, because the energy is fantastic. This has been so fun. Thank you so much for joining us. I hope you had a good time.
Thanks, everybody. Feel free to connect with me on LinkedIn or whatever. Happy to have a conversation anytime. Wonderful. Yep. I think he's the only Hansel Valencia on LinkedIn. I'm the only living here on LinkedIn. So you can find us both pretty easily. Thank you so much for the birthday wishes, everybody. If you want to save the chat, there's three dots in the top right corner of your chat, probably if you are in the Zoom app, you can save that. And then we will see you next week at both the Data Science Lab on Tuesday and the Data Science Hangout. I think that the lab next week is going to be so fun. We have Garrick Aiden-Bouie talking about agent skills, what's the difference between a skill and a tool, all kinds of fun stuff. Garrick is also just a treasure. So come join us for that. Thank you so much, everybody. Have a wonderful afternoon, evening. It's Hansel's evening. I'll let you all go. Bye-bye, everybody.