In the pursuit of excellence in glycemic management, success is not only about achieving short-term goals but also about implementing data-driven strategies for continuous improvement. Ballad Healthcare’s Vice President of Clinical Transformation & Outcomes Optimization, Preetham Talari, MD, CSSBB, MBA, CPE, explores the power of data analytics in optimizing patient outcomes and transforming healthcare practices.
Get the Slides!
TRANSCRIPT
Jordan Messler: We're going to have Dr. Talari really share the story of how using metrics to drive change, that's an important part of our suite of solutions that we provide to our sites is really giving them that data to drive change. You can't improve what you can't measure, and we want to show sites the value they're getting from those reduced hypoglycemia events.
The final piece, though, is the center of this story. So, I mentioned all the parts of the solution in addition to Glucommander, but we really have the team, the multidisciplinary support that works with your team to improve glycemic management, the people and the processes to drive that technology. And you've heard from the last couple of days, our customer support team, our customer success team, our clinical practice team, our delivery team, all working in cohort together internally here at Glytec.
And our customer success team, a clinical practice team working with your sites as well to drive improvement. When you do this, when our sites are able to do this, they're finding, they're optimizing patient safety, having that reduction in costs through improvement, a reduction in preventable hypoglycemia, translating and length of stay reductions, consistency of care, standardizing care, supporting clinicians, and overcoming clinical inertia, driving change in the hospital.
And for our sites, that's translated into real results. A study a few years ago from one of our sites at Kaweah, and you heard from Emma earlier today at Kaweah Delta, they showed 9 million dollars after implementing Glucommander whole house. This 99.8 percent reduction refers to severe hypoglycemia reduction less than 40 when Grady Memorial Health System implemented IV Glucommander, and they showed subsequent studies showing sustained results in severe hypoglycemia reduction, and sites have seen 90 to 99 percent reductions in severe hypo with Glucommander IV.
We've seen reductions in time to target and managing DKA patients and reductions in preventable severe hypoglycemic adverse drug events and compared to other eGlycemic Management System software. So, really quick review of Glucommander and why we're here, but I'm really excited to present Dr. Talari.
He's the Vice President of Clinical Transformation and Outcomes Optimization at Ballad Health. I've had the joy of getting to know him over the past year. He's a fellow hospitalist prior to Ballad Health, he was at the University of Kentucky for over 12 years. He joined Ballad Health in the last year.
He's been instrumental in driving change there, and I'm excited to hear from him on how they're using metrics to continue to drive change at Ballad Health. Dr. Talari, thanks for joining us.
Preetham Talari: Thank you, Dr. Messler. Again, thank you for that introduction. I do agree with you that Glytec customer support teams, delivery teams, you know, they're all fantastic.
My experience has been that for the last one year I've been working with your team. So I totally agree with that. Again, good afternoon, everyone. Thanks to Dr. Messler for inviting me to this talk. We're going to basically, you know, Dr. Messler did the exciting part of the talk because I'm going to talk about data and measures, which is not so exciting, but I'll try to make it as exciting as possible by also sharing the Ballad experience with Glucommander and some of the metrics that we're going to talk about.
So, uh, as Dr. Messler introduced, I'm a internal medicine physician. I'm a hospitalist and my current title is Vice President for Clinical Transformation. Basically passionate about, in addition to the one on one care that we provide for our patients, I'm also interested in, you know, how do we improve systems and processes and really, uh, passionate about quality and process improvement.
And that's how I'm involved in this endeavor to improve glycemic control across our Ballad Health. Talking about Ballad Health, it is a 21 hospital system covering areas in about 29 counties, I think, in northeast Tennessee, southwest Virginia, and parts of Kentucky, and North Carolina, not only it has all these hospitals on in all these different counties, but has huge outpatient presence, and then also multiple rehab places.
So true integrated delivery health system in this part of the world. So, what we're going to cover mainly, in the next 30 minutes is, as I said, we're going to mainly talk about, you know, data. And how do we take the data and then come with different measures and metrics? What are the different types of measures and metrics that we need to think through?
And then talk about Ballad's experience with Glucommander and then how we are measuring success at Ballad Health.
So, when we talk about data, I'm sure you’ve all had these questions, right? You know, where, where do I start? Where do I get this data? Am I looking at the right data? And when is it that we are having too much data?
Right now we are overloaded with data. So how do I get enough data that is meaningful? And then when I get that data, how do I share with the right stakeholders? How do I share with the front lines, with the middle level, with managers or facility CMOs and CNOs so that they are held accountable to the outcomes that we are looking at.
And then how do we measure success in general? So these are all the questions that we think about whenever we talk about data, so we're going to address some of them in the next bit. So, we just talked about how much data that we have out there, right? In this time and age of big data we are overloaded with data.
And oftentimes it's difficult to take out the signal from that big noise. So it's really important for us to extract the data that is meaningful. And that actually tells a story and that's actionable because, you know, people are overloaded with data that is not meaningful, that is not actionable.
And that doesn't tell a good story. Eventually they'll ignore it. I ignore it when I get that big data that can't really interpret, that is not visual friendly. So, gotta be careful, really, to make the signal out of that data, which is meaningful, actionable, and tells a meaningful story. So, as we're talking about data, let's talk about metrics, right?
When we take this data and then come up with metrics to measure ourselves, we can talk about process metrics, outcome metrics, and balancing metrics. I'm sure most of you have heard this terminology before, but process metrics is exactly what it says, measuring processes, right? When we lay out that process map, when we want to measure different aspects of the processes that we're dealing with, that's what we call process metrics.
And then outcome metrics are the outcomes that we are willing to achieve based on the processes that we have established. As an example, as we're talking about glycemic control, the blood glucose timeliness, the reliability of meal time triad, we can say that these are all the processes that we want to measure, but then also the result of these processes is the outcomes such as hypo and hyperglycemia and other measures that we look at.
And, you know, it can be like a chicken and egg situation with the process and outcome measure metric sometimes, right? It's, it's quite natural for people to be confused with. Okay, what, what is the process measure? What is an outcome measure? Right? We talked about blood glucose timeliness, right? Is it a process measure, an outcome measure?
Because in the previous slide, we talked about blood glucose timeliness being a process measure, but one can easily make an argument that that's an outcome measure for different processes that are upstream to that. So it's quite natural to make that case. So It is a chicken and egg situation, whether it's a process metric or an outcome metrics, depends on the context that we are talking.
And, a process metric can become an outcome metric in a different clinical situation based on the relevance to that particular initiative. So the process metrics and outcome metrics have, you know, their own goals, right? These process metrics can help us find those quick wins, right? Because it takes time to have those, the needle moved on the outcome metric.
So we can look at process metrics to find those quick wins. And then also indicate these process metrics helps us understand how successful we're going to be with the outcome metrics and not only they help with our success, but then also support our improvement efforts over and, you know, over the long term.
So, this is an illustration of, again, just, for those of you who are new to this terminology, of process and outcome metrics. You know, if that's a process map, we measure different processes as process metrics and then outcome metrics at the end. As an example, if you take an example of catheter associated urinary tract infections, I'm sure in all of your health systems, this is something that you all want to improve.
And it's very much relevant to glycemic control because as we all know, appropriate glycemic control reduces infections like CAUTIs and CLABSIs and other infections in the hospital. So if we want to improve catheter associated infections in the hospital, you know, that's an outcome, right?
So that's, that's an outcome measure. You know, the rate of CAUTIs can be considered an outcome measure, but then what are the upstream things that we do to decrease the rate of catheter associated urinary tract infections? It can be decreasing the number of Foley days or decreasing the number of hours Foley catheter is in place or measuring other processes that result in the outcome of decreased rate of catheter associated infection.
So in that spirit, the mean hours Foley catheter is in place is a process measure that we can measure short term and then ensure that measure needle is being moved before we can see a real change in the outcome measure. So as we're talking about process measures and outcome measures.
We also need to be really careful in thinking about balancing measures, right? When we are trying to improve our processes, when we are really passionate about it, it's quite natural to have some unintended consequences, and these unintended consequences are the balancing measures that we need to have.
For example, in this case, in an effort to decrease catheter associated urinary tract infections. In an effort to take out Foleys, there is an unintended consequence of replacing Foley catheters when these are not taken out appropriately, right? So that's something that we don't want to happen in our patients, right?
As you all know, it's not a pleasant feeling, to have a Foley catheter replaced for example, you know. Putting it in the first place, so that is a balancing measure, that we want to have. So this is the best way to illustrate the outcome measure, process measure, and a balancing measure that we all need to keep an eye on, not just for catheter associated infections, but then also our glycemic efforts in our, in all our hospital.
Example is in the outpatient area, the outcome measure can be, um, mean HbA1c for your patient population and a process measure can be the number of patients in which the HbA1c is tested, right? And then the balancing measure can be the percentage of patients where the unnecessary labs have, uh, are being drawn.
And the same thing with the outcome measure of hypoglycemia when we are trying to improve hypoglycemia, we also want to measure the process measure of blood glucose timeliness for patients on IV insulin. But at the same time, when we are really passionate about decreasing hypoglycemia we need to keep an eye on if that is impacting the hyperglycemia rates, because that is an unintended consequence one might experience.
So now, you know, we talked about data, what people, you know, the thoughts people get when, when thinking about data, the importance of signal and noise and the definition and understanding of process outcome metrics and then balancing measures. So I know Dr. Messler and Dave Cooper gave a fantastic presentation really diving deep into glucometrics.
So I'm not going to dive deep into GlucoMetrics, but just wanted to say that the GlucoMetrics platform provides that fantastic resource for us to keep an eye on those outcome measures and then process measures as you see on the slides. You know, we can slice and dice by, we have like 19 different hospitals, right?
Or 20 hospitals in our system. So using GlucoMetrics platform, I can slice and dice by hospital and different units, different service lines, how I am doing with the outcome measures of hypoglycemia. How am I doing with outcome measures of hypoglycemia? And not only there is data out there, but the good thing about GlucoMetrics is how it is visually demonstrated.
It's very easy and nice and easily can be interpreted, but then not only these outcome measures of hypoglycemia and hyperglycemia that can be sliced in different ways for better understanding. And then also that helps in our improvement efforts. There are process measures like blood glucose timeliness.
Right? You can also slice and dice that data by hospitals and different units so that you can find out the opportunities that your team can work on. So, just wanted to say that GlucoMetrics is a fantastic resource for all of you that are on this journey to measure yourself in terms of process and outcome metrics.
And this is a scatterplot on the GlucoMetrics website that I really love, right? You know, these are all, we took out the names intentionally, but these are all the different, you know, hospitals in a health system. And then you can nicely see how these dots move in the right direction, which is the left corner, left bottom corner, as you try to improve those processes, as you try to improve these outcomes, these dots nicely move towards the left lower corner, which basically indicates that you're going in the right direction.
That is decreasing hypoglycemia and hyperglycemia and all the downstream effects of, you know, on different things, right? Length of stay, financial metrics, and then basically patients are more likely to do better.
So, to kind of summarize on the data part, you know, we all talk about data driven decisions, right? Or if not data driven decisions, data informed decisions. We talk about data driven decisions and data informed decisions, but... I think it's an open secret that we don't use data most of the time to, to drive decisions or inform our decisions.
So it's really important for us to start with the data and it takes time, right? In my previous job, just to develop the dashboard, a very simple dashboard, it took six months. Right? You need to work with the right stakeholders. You need to work with your quality department. You need to work with your IT department.
And even within the IT department, there is a data analytics department. So it takes time to build these dashboards and then to find the right partners and to have that finally established that you can work with. So the great thing about GlucoMetrics is Glytec team already, you know, they've already done it for us so that we don't need to spend all this time trying to collaborate with all these stakeholders to build a dashboard.
So the GlucoMetrics platform is something that's already done for us.
So, as we talked about data, we'll just, I'll just briefly share the experience of Ballad Health with Glucommander. So, as I said, you know, it's a 21 hospital system and we've done it in two phases. We've, for a while, for years, we have IV Glucommander in place that our physicians and APPs and nurses are very satisfied with, but now we are embarking on the journey of SubQ Glucommander, and also upgrading the IV Glucommander.
And we started as phase one in two hospitals, two small hospitals, even though I'm talking about two small hospitals, we're talking about 200 nurses that we had to educate, and about 50 providers that we need to educate. And the reason we started in a phased manner is to have all the learnings that we need to before we kind of deploy and expand in the rest of the facilities, in the rest of the hospitals.
So now we've completed phase one and now we are in phase two. Our phase two IV upgrade and SubQ expansion is going to start. You know, it's going to actually start happening in the next two weeks and it's the culmination of the result of six months of hard work of our teams and the Glytec team.
So, the pilot was successful in these two hospitals. Again, we found plenty of opportunities that we could, you know, address some concerns, questions, you know, whatever needs to be done, that way we don't need to deal with that in the bigger expansion for phase two. And we are very comfortable and pleased with the situation we are in.
And we would not have been in that situation if we didn't do a pilot phase with phase one. So I don't need to, you know, preach you on the SubQ Glucommander. Dr. Messler already has, talked a bit to it. But SubQ Glucommander is also a tech driven approach to basal bolus.
You know, the way I would like to tell our physicians and APPCs that, you know, every day you go into the hospital as a hospitalist, I go into the hospital, check my blood glucose levels, and then try to adjust that basal bolus on a daily basis myself, right? It's a lot of work when you have a full load of patients.
So I'll tell them that Glucommander SubQ, or IV, these are the tools that do that job for us. So it takes out that cognitive load so that we can better take care of our patients. So these are our facility groups, you know, as I said, some hospitals are using only IV. Some hospitals don't use, don't have any experience with Glucommander.
And some hospitals, the group A hospitals is where we piloted our phase one of Glucommander expansion. So this is our timeline now. I think I'll switch back to the previous slide, but this is, you know, this is the high level project management plan, you know, with the system as big as ours with about 20 hospitals, you know, it can be, it can look like a lot of work, but it's all worth because at the end of the day, when we establish our systems well by working at the front end well, you're going to have dividends eventually, over the years in terms of, you know, what our patients benefit.
So this is our high level project plan, started in July, been having weekly meetings, again, not only developing these order sets, the technical work, the background upgrade work that happens, so we, we are at the tail end of this project plan now, so now, this is our timeline to upgrade IV in all the facilities on November 6th, and then we are doing the SubQ expansion phased manner, starting on December 4th, a couple of hospitals at a time.
And because of the, you know, how big of a health system ours is we need to really have a structure for planning and implementing, and then working with our Glytec colleagues. We had an enterprise level, system level steering committee that was basically directing how this project is planned and then how this is going to be run.
And under the direction of the steering committee, we had, you know, technical core teams and then clinical core teams and education teams at the system level with the goal that all this information flows through our hospital CMOs, the facility champions, and they'll take full ownership to have this implemented in hospitals.
And I was telling people that, you know, as part of our technical and clinical code teams, on a weekly basis, we had about 50 people meet every week to finalize those workflows, finalize those order sets, and define those workflows and kind of optimize EPIC in terms of coming up with the BPA or refining the glucose summary tab or the sidebar, you know, all these things to make this successful as we are nearing our goal updates.
So, I know I spend a lot of time talking about process metrics, outcome metrics, and balancing measures, so I better have something for our health system. And we already do, which is the process metric is, you know, the process metric that we identified for our health system is at the end of the day, if this tool is not used by our physicians and APPs, our patients are not going to benefit, right?
You know, that is the process that needs to happen so that our patients can benefit from this particular tool and help with overall glycemic control. So the process metric that we chose is the adoption of Glucommander order sets by our providers. And the outcome metrics, as you can imagine, are the hypo and hyperglycemia, right?
Without the process metric of these order sets or Glucommander not being used by our providers, we're not going to see the outcome metrics that we want to see in terms of reducing hypoglycemia and severe hypoglycemia rates. And, in an effort to rightfully use this Glucommander tool, you know, it's quite possible where this Glucommander is, you know, can be used inappropriately in some patients where it is not indicated.
So we are using that as a balancing measure so that we have that as part of checks and balances as a balancing measure to ensure that we are using it appropriately. So, finally, I don't, I don't know how much, you know, where we are with the timeline, but we are in the final slide.
You know, I would reiterate on some of the key takeaways as we end this session. Not only this needs daily tactical planning, but it's very important to have that strategic, you know, six months or one year vision plan. As you embark on this journey of implementing Glucommander or doing glycemic control in general at your institutions or health systems.
And always going back to people, process, technology, right? Each and every part of this component is important, not only technology and process, but at the end of the day, it's people and keep in mind the scalability and sustainability, right. There is a reason we started with phase one and then scaling up after we learned all the components.
Learning lessons that we had to, and then talk, uh, think about sustainability. It's, it's can't, uh, overemphasize the importance of building systems and hardwiring your processes so that this particular tool is used in a consistent manner so that our patients would benefit from it.
Like I talked about EPIC optimization, BPS to remind physicians so that this tool can be used, things of that nature. And the importance of standardization and actually having a goal that is measurable, right? At the end of the day, and it's not uncommon for it to happen, right?
We embark on an improvement project or an improvement initiative, do multiple things, and at the end of the day, only to feel like, oh, actually we have improved things without having a measure, or without, without having a measurable process. So it's very important to have a goal that you can measure with which you can define success.
And it's quite easy with glycemic control because Glytec already supports it. So as this talk is about data, I would emphasize again how invaluable data is and without which I wouldn't, I would not even take this journey because you wouldn't know how you are benefiting our patients.
So I'm going to stop there, Dr. Messler, if you'd like to add anything or if our audience that you'd like to chime in.
Jordan Messler: Yeah. Thank you so much. That was really fantastic, Dr. Talari. I think the audience, you know, really can begin to understand why you're the Chief Transformation and Outcomes Optimization. And, you know, this is, we know that glycemic management can be complicated. We here at Glytec try to make it simpler, but it's impacting 40, 50 percent of patients.
So a couple of questions. One, you know, you're talking about, right, all the steps to drive change, the really a structure for success, standardization, scalability, getting these metrics, which has been a big part of this talk. I'm curious before I get into a metrics question, can you talk a little bit about what got you into this kind of role and the hospital needing this kind of role?
I think it's evident as you're talking on really driving change and having someone, a leader like yourself, understanding the concepts is important. This is essential, but this is not a typical role at a lot of health systems, and it seems to be a really exciting role. This is a complex place we work in.
So I'm curious about how you got into this role and maybe, you know, even talk about some other projects that you do.
Preetham Talari: Yeah. Thank you for that question, Dr. Messler. Again, it's very important. I'll actually, I think it's okay to mention two individuals here, you know, who kind of got me into what I'm doing now in this clinical transformation role is the, the first individual is, you know, he's our both, our both boss at one point of time, Dr. Mark Williams.
He was instrumental in instilling this idea of not only, you know, treating patients one-on-one individually, but then how do you actually change systems? How do you change processes and then work with systems and other stakeholders so that you can benefit larger and larger patient populations.
At Ballad Health, Dr. Amit Ashish, who's the Chief Clinical Officer for the health system, has been very instrumental in envisioning the Center for Clinical Transformation at Ballad Health and this role to take on these initiatives like glycemic control or other system wide initiatives to drive change for what's best for patients and mainly relying on the fact that, yes, we are all into PDSA cycles, you know, making small tweaks to, bring change, you know, kind of taught at the residence level or, during our earlier career. In addition to that, how do we actually transform, right? Bring those big changes in addition to those incremental changes.
So that's what kind of got me into this role. Dr. Mark Williams’ mentorship and Dr. Amit Vashisht's vision at Ballad Health.
Jordan Messler: Yeah. Mark Williams would be excited to see us both. You know, he's certainly a mentor, really my first boss as well. The other question around metrics, our team, here that's been listening.
I know the audience has been really excited about how you've been using our platform, GlucoMetrics, the importance of getting the right data. You mentioned at the beginning of the talk, you know, delivering that data to the front line. One of the common challenges that we see is that, all right, we have the data, we get the data, the team has it, they see the great results, but then sometimes there's a translating that to the front line that they didn't realize that, oh, we're actually are seeing a reduction in hypo and hyper.
Do you have that kind of structure in place so that the front lines also see when things are going well or not going well and, and, and really, you know, how they see the data as well?
Preetham Talari: Yeah. So again, great question, right? Because, that's... You know, we can always over communicate, because we always under communicate, right?
So, sharing data is one of those things. So far, again, as we are embarking on this journey of SubQ expansion, there is a plan in place to... use the GlucoMetrics and other data platforms to, so that the data can be shared automatically with the facility leaders from the corporate. And the facility leaders would then share that data with the service line directors or hospital directors, so that they can share with their individual physicians or APPs.
So, there's a plan to automate that process so that, you know, when people are doing all this hard work, right, giving input to build these order sets, to establish these workflows, working with Glytec and all that, they would want to see the data that this is actually changing things for the betterment of patients.
At the end of the day, if our frontline providers can see that data, that's the biggest motivator for them to continue to do what's right, which is using this particular tool. So yes, Dr. Messler, there is a plan in place to achieve that goal.
Jordan Messler: Yeah, yeah, that's fantastic, right? I mean, two big things there.
One, you have to have a plan for this, like other things. We've got to plan for success. You mentioned it. What's the plan? And how can you make it easier instead of having to pull reports? How can we push it out, automate things, certainly make things easier. Really, this was a fantastic… I think that's really all the questions that we have time for. Really appreciate you taking the time today.
We appreciate everybody in the audience taking the time to stick around as we really, largely wrap up our Time to Target conference for those that want to stick around a little bit longer. At five o'clock, we do have our post Time to Target huddle to hear a wrap up of all of our sessions of the day.
Thanks again, Dr. Talari, and thanks everybody out in the audience for joining us today. Have a great afternoon.
Preetham Talari: Thank you. Thanks everyone for sticking around and have a good rest of the day.
SOP #34