On-demand webinar

Watch the recording

Learn how to align analytics with planning milestones to shape your 2026 risk adjustment strategy.

Download slides

WEBVTT

00:00:03.620 --> 00:00:12.150

Bailey Fields: All right, hello everyone, welcome. We'll give everyone just a few seconds to join us. I see everyone joining in. Hello, welcome, welcome.

00:00:14.780 --> 00:00:16.379

Bailey Fields: Hello, everyone.

00:00:17.630 --> 00:00:29.820

Bailey Fields: All right, we're going to go ahead and get started. So, welcome everyone to today's webinar, sponsored by Optum. Today, you're going to be, hearing about the analytics in action, building a strategic framework for Risk Adjustment.

00:00:29.820 --> 00:00:39.110

Bailey Fields: Before I turn it over to the Optum team, I just want to go over a few housekeeping items. So, as a attendee, you will notice that your camera and mic are going to be off.

00:00:39.110 --> 00:00:50.479

Bailey Fields: for the duration of the webinar. This is just for your privacy. If you do have any questions throughout the webinar, we encourage you to put those in the Q&A that you'll see in the box at the bottom of your screen.

00:00:50.480 --> 00:00:59.499

Bailey Fields: If the Optum team can't get to all the questions, during the duration of the webinar at the end, not to worry, they'll follow up with you directly afterwards.

00:00:59.500 --> 00:01:03.720

Bailey Fields: So, without further ado, I'm going to turn it over to Vijaya to get it started.

00:01:04.250 --> 00:01:28.430

Vijaya Vishwanathan: Hi, everyone. Thank you so much for joining us today. I'll start with a quick introduction. I'm Vijaya Vishwanathan. I lead the risk adjustment and analytics SaaS platform over here at Optum, and I come from a health plan background and the vendor side both, so I have a mix of knowledge across Medicare, Medicaid, FEA, and all the government programs. Nice to meet you all today.

00:01:28.430 --> 00:01:30.260

Vijaya Vishwanathan: Virtually. Over to you, Ramnik.

00:01:31.320 --> 00:01:41.980

Ramneek Kaur: Thanks, Vijaya. Hi, everyone. Thank you for joining. This is Ramnik. I have about 10 years of experience in healthcare technology, in analytics, strategy, and growth teams.

00:01:41.980 --> 00:02:02.250

Ramneek Kaur: At Optum, I partner with health plans and health systems all across Medicaid Advantage, Medicaid, as well as commercial plans, helping them create strategic risk adjustment programs that can drive better outcomes. Next on, I believe we have the agenda. Vija, would you like to go to the…

00:02:02.250 --> 00:02:27.080

Ramneek Kaur: Agenda slide, I think we quickly covered this. Yes, so very excited for today's webinar. We're going to start with the life cycle of analytics, talk about different steps, why each of these are important, and the timing of those. Moving on, we will look at a data-driven roadmap for 2026. What does a roadmap in terms of timeline and milestones look like for risk adjustment in the

00:02:27.080 --> 00:02:28.650

Ramneek Kaur: year 2026.

00:02:28.650 --> 00:02:48.360

Ramneek Kaur: And lastly, the third, which is usually ignored, we have the data, we have the roadmap, what else do we need to do? How do we manage the program so that they give us the outcome that we're expecting? So we look at program orchestration. So to start with, I am going to hand it over to Vijayar to talk about the lifecycle of analytics.

00:02:48.700 --> 00:03:02.210

Vijaya Vishwanathan: Yeah, so, thank you, Ramnik. So, let's start with the life cycle of analytics. We say life cycle of analytics, but what it is, is more the life cycle of risk adjustment, with analytics kind of driving it, or being the engine behind it.

00:03:02.380 --> 00:03:16.170

Vijaya Vishwanathan: So let's look at this slide and what we start with. So the first three parts, as you see, talk about data collection and consolidation, gap identification, and member-provider profiling. So let's break this down a little bit to say.

00:03:16.460 --> 00:03:23.710

Vijaya Vishwanathan: how everything begins in the continuum of risk adjustment. Everything starts with data collection and consolidation.

00:03:23.970 --> 00:03:35.600

Vijaya Vishwanathan: it is kind of like building our foundation, right? Bringing together claims data, EHR data, lab results, pharmacy data, and member-provider demographic data, all into one unified view.

00:03:35.620 --> 00:03:54.639

Vijaya Vishwanathan: without kind of having this foundation, it's really hard for us to move to the next step, which is the gap identification. Once we have a good, solid foundation, you can actually go to the next step of identifying gaps. And these gaps, although we, you know, there are multiple ways to call it gaps, opportunities, suspect.

00:03:54.920 --> 00:04:11.099

Vijaya Vishwanathan: it actually talks about two things, right? One, gaps in your data sets themselves. Are you missing claims? Are you missing pharmacy data? Are you missing any kind of member demographic information? So that's one kind of data gap that we look at, and the other is truly opportunity identification.

00:04:11.180 --> 00:04:28.969

Vijaya Vishwanathan: Where do your members have gaps? Where do your providers have gaps? And then looking at… which is all driven by completeness of data. So, catching these early and efficiently in the lifecycle means that we can act on them before they become bigger issues.

00:04:29.080 --> 00:04:43.130

Vijaya Vishwanathan: Next, once you have, kind of, your profiles identified for gaps, you move into categorizing them into a member or a provider profile. What do we say… mean when we say we are profiling members or profiling providers?

00:04:43.150 --> 00:04:52.050

Vijaya Vishwanathan: It's basically looking at segmenting those populations based on risk, based on utilization. It could be based on engagement, it could be based on

00:04:52.220 --> 00:05:00.250

Vijaya Vishwanathan: visits, etc. From a provider perspective, we can also look at a lot of stratification, like coding accuracy.

00:05:00.250 --> 00:05:03.749

Vijaya Vishwanathan: Performance trends, recapture rates for providers.

00:05:03.750 --> 00:05:21.310

Vijaya Vishwanathan: Again, when you build out a provider profile and a member profile, it acts as this 360-degree view that lets you identify each of these actions further down. They help us tailor our outreach and education so that appropriate interventions can be

00:05:21.510 --> 00:05:25.359

Vijaya Vishwanathan: Targeted, and they are meaningful based on data.

00:05:25.960 --> 00:05:31.139

Vijaya Vishwanathan: So that's the first three steps of data gathering and making sense of that data.

00:05:31.650 --> 00:05:44.819

Vijaya Vishwanathan: Next is, great that we have analytics and data, right? But what do you do with it? And unless you actually act on it, there is no value to it, right? So, program orchestration, so program design and implementation.

00:05:44.820 --> 00:05:59.530

Vijaya Vishwanathan: So, now that you know what your member profiles and provider profiles look like, how do we launch those programs? Is it creating member engagement workflows? Is it provider training, or provider outreach, or provider education?

00:05:59.920 --> 00:06:15.750

Vijaya Vishwanathan: Each of these you can launch based on the data that you see. And it could vary per customer, right? It could vary based on your plan data, it could vary based on your provider information, it could vary based on what you want to invest in as a plan or a provider.

00:06:15.850 --> 00:06:34.820

Vijaya Vishwanathan: The key is also that these programs are all data-driven, so there is no guesswork, right? You're building based on your foundation and going to the next steps. Finally, launching a program is great, but again, there needs to be that continuous loop of tracking the program outcomes, tracking the program KPIs.

00:06:34.910 --> 00:06:53.789

Vijaya Vishwanathan: impact of the programs that you're launching. If you're doing a member engagement program, is it yielding the results that you expect? Is it giving you the ROI? Is it giving you the patient engagement, member health outcomes that you expect? If it is a provider program, is it giving you the tracking and everything that you can,

00:06:53.890 --> 00:07:09.549

Vijaya Vishwanathan: measure so that you know that you're investing in the right programs. So, all of this is driven by analytics, is what this indicates right over here. So, measuring those KPIs, measuring those ROIs, helps us get it to the next stage.

00:07:09.670 --> 00:07:29.299

Vijaya Vishwanathan: Based on all… and again, all of this plays into prospective risk adjustment, looking ahead for 2026, as you'll see through our webinar, looking back for 2025 dates of services, looking at services across the board for coding, submissions, and every continuum of risk adjustment. Quality measurement plays into this as well.

00:07:30.500 --> 00:07:31.430

Vijaya Vishwanathan: So…

00:07:31.540 --> 00:07:37.469

Vijaya Vishwanathan: Moving on to the next slide, let's expand a little bit about the inputs that we just talked about.

00:07:37.720 --> 00:07:45.750

Vijaya Vishwanathan: Integrated analytics is the key to comprehensive programs. So what this means is that at the onset, our inputs are

00:07:45.830 --> 00:07:59.959

Vijaya Vishwanathan: all the clinical data sets that we all know about, CMS files, MMRs, MORs, as it applies to each line of business, edge server data as it applies, claims information, 837s, lab data, pharmacy data.

00:08:00.220 --> 00:08:11.319

Vijaya Vishwanathan: Further moving on, when you want to target specific provider groups, or member groups, you need to have member provider information, and that needs to go behind

00:08:11.470 --> 00:08:26.169

Vijaya Vishwanathan: or beyond any specific demographic data. Especially for members, we want to look at SDOH data, healthcare databases, any publicly available data that we could use to augment our analytics, any kind of

00:08:26.390 --> 00:08:39.500

Vijaya Vishwanathan: transportation information, all of that that falls into the member-specific categories, we really want to start tracking them here as well. Now that we have kind of looked at what our inputs are and the life cycles are.

00:08:39.600 --> 00:08:54.129

Vijaya Vishwanathan: we are looking at it to see what outcomes it can drive. And here's why AI and machine learning can come in, right? AI and machine learning can kind of take all of this data and help us get to the next step to say.

00:08:54.130 --> 00:09:02.899

Vijaya Vishwanathan: here's the gaps that we know are chronic, can occur via recapture, but what can AI, and especially as we are growing into the agentic AI world.

00:09:02.900 --> 00:09:19.540

Vijaya Vishwanathan: how can AI help us identify predictive gaps before they even come in? How can we get to the next step of profiling these members and prioritization using all of this? So turning those into actionable insights, is what Ramnik is going to talk about next year.

00:09:20.750 --> 00:09:45.610

Ramneek Kaur: Thanks, Vijaya. That was a great overview of analytics and all the steps involved. I want to talk a little bit about the output, tracking and, you know, closing the loop a little bit. You touched on it a little bit, but I want to talk about why it is so important when we're looking at analytics as a whole. So, a lot of times, member profiling is just seen, like, collating all the data we have for a member

00:09:45.610 --> 00:10:10.550

Ramneek Kaur: or a provider, but I think it's more than that. It's understanding if we have all the data points that are required to get the right program for the right member or the right provider. Two members who might have a same number of chronic conditions or even same chronic conditions, like three chronic conditions, they might not have the same segmentation or might not lie in the same category because of one factor

00:10:10.550 --> 00:10:35.069

Ramneek Kaur: maybe, you know, they're in a rural zip code, or they are not as engaged with their provider as the other one is, you know? Are they completing their annual wellness visits? So things like that, small things and small data points, is the key to getting accuracy to, are your… is your prioritization and your profiles as accurate as they should be from a program perspective? Are you going to

00:10:35.070 --> 00:10:44.689

Ramneek Kaur: achieve that success in the program you're designing. So keeping in mind the strategy and program design aspect of it, even while you're looking at analytics is really important.

00:10:44.860 --> 00:11:09.710

Ramneek Kaur: Similarly for providers, two health systems of, same size could be really different based on if they're seeing a lot of Medicaid members or Medicare Advantage members, because they could be on different EMRs, maybe some Medicaid in some rural areas, they might even be, dealing with paper records. So, just knowing a lot about them is not enough. You need to know the exact, data points that help you understand

00:11:09.710 --> 00:11:12.840

Ramneek Kaur: What programs will give us the maximum success?

00:11:12.840 --> 00:11:37.799

Ramneek Kaur: The second thing that I want to focus on is the feedback loop. Like you mentioned in the last slide as well, I think that is something that we see missing from more than a few analytic systems or processes built out there. You get… you do all the programs, everything you're supposed to do, but what doesn't happen is that you close the loop for the data that goes into the next year. Again, any kind of goal or target setting

00:11:37.800 --> 00:12:02.330

Ramneek Kaur: for the programs, understanding what the KPI should be, or what is the goal we should be meeting, largely needs data from prior years. Any, you know, correction in data that we might have done, maybe provider addresses, phone numbers, as we see during retrieval, they change. Are we getting the right data back into our system so that next year, when we get to those programs, we are

00:12:02.330 --> 00:12:08.790

Ramneek Kaur: closer to the KPIs that we want to have, and we're closer to the source of truth for analytics that we want to do.

00:12:09.050 --> 00:12:29.769

Ramneek Kaur: I think it's a good, segue for the next slide as well, where, we're going to… Okay, before that, we have the poll. So, before we talk about the analytics capabilities, we have a poll here for you, and we'd love to hear from you on how… how mature is your organization's analytics capabilities?

00:12:31.420 --> 00:12:35.970

Ramneek Kaur: I think it'll be on the screen for about 30 to 45 seconds.

00:13:00.150 --> 00:13:03.280

Vijaya Vishwanathan: We should start seeing some responses come in.

00:13:14.610 --> 00:13:22.110

Vijaya Vishwanathan: Okay, great. So, looks like we have all our responses in, and… It's great to see that

00:13:22.650 --> 00:13:36.949

Vijaya Vishwanathan: solid, I want to say about 60% are sitting in the… we use analytics for program tracking, which is great. So, I think what this means is that you are in a place where you do have, good analytics.

00:13:36.990 --> 00:13:49.230

Vijaya Vishwanathan: good program orchestration, but it has to be now taken to the next level, where it is integrated end-to-end and is AI-powered. For organizations that you are starting to just collect data,

00:13:49.570 --> 00:14:05.550

Vijaya Vishwanathan: Here, I think the biggest challenges we see are data silos, inconsistent data formats, and healthcare data is one of the most complex data sets, I think, out there, given that it is so diverse, and given that it is so many different sources.

00:14:05.570 --> 00:14:12.820

Vijaya Vishwanathan: starting to consolidate those using AI where possible to help, Eliminate those data silos.

00:14:12.880 --> 00:14:20.860

Vijaya Vishwanathan: Are great. And then we have basic analytics and reporting. We do have about 20-30% people sitting in that category. Again.

00:14:20.860 --> 00:14:34.070

Vijaya Vishwanathan: that's where you want to take it to the next level to say, okay, I do have, you know, analytics and reporting, what do I do with that next? And then, Ramnik, if you want to add anything else about the program tracking aspect here.

00:14:34.170 --> 00:14:36.619

Vijaya Vishwanathan: We can close out the poll and move to the next one.

00:14:37.240 --> 00:14:59.629

Ramneek Kaur: Yeah, it was great to see that a lot of, audience here are working with program tracking and improvement, and I was excited to see we had about, I think, 8 or 9%, with the AI, leveraging AI already, which I think, might increase, keeping in mind the governance around it, but I think it can prove to be a great tool in decision making.

00:15:01.750 --> 00:15:02.510

Ramneek Kaur: Thank you.

00:15:02.930 --> 00:15:03.910

Vijaya Vishwanathan: Go ahead, Romney.

00:15:04.280 --> 00:15:29.249

Ramneek Kaur: Of course. And then, next year, I think this is… this ties into what we were talking about, and also how mature is your, analytics capabilities for anyone. So, I think this, we thought could be a good litmus test, you know, for all of us as an audience to look at and understand, are we there where we want to be? Like I was saying, a member profile is not just collating all the data.

00:15:29.250 --> 00:15:54.190

Ramneek Kaur: have for a member. It's looking at those key attributes that help us decide, what those… what the program can be, getting the right care to the right member at the right time, as our goal always is. Do we have enough data to make sure, that we have that information for both members and providers? So, member profile can be anywhere from, like, should have data about demographic, chronic conditions, utilization patterns.

00:15:54.190 --> 00:16:08.049

Ramneek Kaur: and preventative care, and also their engagement with their providers. And similarly for providers, we need information about their specialty, their size, how they're doing in terms of coding and, you know, quality metrics.

00:16:08.050 --> 00:16:31.069

Ramneek Kaur: for the audience here, and also, from Vijaya, since you work with analytics so closely, I want to understand from you, are any of these metrics more easily available than others that are not, and how can plans make sure that they have the right metrics? Which ones of these, is there a priority here that you can help share with us, and which ones are the most important?

00:16:31.440 --> 00:16:45.810

Vijaya Vishwanathan: Yeah, so rather than a priority, I'd almost call it… so I always like to say this statement, not all data is created equal, especially in healthcare data, right? Some elements, like demographic data, claims data, so…

00:16:45.810 --> 00:17:04.840

Vijaya Vishwanathan: again, chronic conditions and utilization patterns kind of tie into it, but because all of that can be derived from claims, right? All of that is relatively easy to access. They're structured, we all know we have all moved to EDI now, so structured, standardized, and so you can kind of derive a lot of things from those.

00:17:04.839 --> 00:17:09.009

Vijaya Vishwanathan: They're also very well documented in EMR systems, etc, too.

00:17:09.089 --> 00:17:15.799

Vijaya Vishwanathan: But then you get into areas like what you just said, social determinants of health and actual gap in care.

00:17:15.990 --> 00:17:27.680

Vijaya Vishwanathan: Which could be due to other reasons, right? And that's where some of these complexities come in. Knowing all of that data is going to help us drive care management decision making.

00:17:27.780 --> 00:17:46.810

Vijaya Vishwanathan: These often require pulling from external sources, or even mining unstructured political notes, right? And that's, again, where it can be so useful to look at tech and analytics as a partner here. So, for example, just transportation challenges or housing instabilities

00:17:46.810 --> 00:17:59.620

Vijaya Vishwanathan: So many times are overlooked, and these are the patients that you could prioritize for either in-home assessment cares, or in case of transportation stability, or providing those transportation vendors, etc, partnering with those.

00:17:59.620 --> 00:18:07.169

Vijaya Vishwanathan: These may all appear only in case management notes, which are not structured, and you need advanced text analytics to surface them, and

00:18:07.830 --> 00:18:22.410

Vijaya Vishwanathan: use it as input into your analytics. On the provider side, specialties, locations, etc, although provider location tends to be really, really complex data to maintain, because it can change so frequently, there's still

00:18:22.530 --> 00:18:33.819

Vijaya Vishwanathan: relatively straightforward to acquire. However, inferring from that data coding accuracy, or where is the gap in provider,

00:18:34.210 --> 00:18:45.810

Vijaya Vishwanathan: life cycle, so to speak, for risk adjustment, performance trends, recapture rates, those can be relatively harder. So all of this depends on timely capture and documentation, and

00:18:46.180 --> 00:18:59.170

Vijaya Vishwanathan: inferring the data accurately from that point. Quality metrics is another one. They often vary by program and payer, so harmonizing those across different data sets is also key for us to handle.

00:18:59.170 --> 00:19:16.480

Vijaya Vishwanathan: So, looking at all of this, while our goal is a 360-degree view, we kind of start off with what is the easiest to obtain and kick off the programs, and then look at sophisticated analytics to kind of augment those and launch the programs that we need.

00:19:16.650 --> 00:19:23.100

Vijaya Vishwanathan: Great part about this is, although some of this data is really complex to obtain and requires a lot of effort.

00:19:23.100 --> 00:19:43.810

Vijaya Vishwanathan: when we do get it right, it does show in the programs and results pretty immediately. It shows in the prioritized interventions, it shows in the outreach, it shows in provider relationships, and it shows in… especially as we are moving towards the value-based care world, it definitely shows in the provider relationships and risk-bearing entities, as well.

00:19:45.410 --> 00:19:47.309

Ramneek Kaur: That's great to know. Thank you.

00:19:48.490 --> 00:19:58.500

Vijaya Vishwanathan: So, moving on, that's actually a good segue. So, moving on to the next slide, what are our key steps? So, we have the data, we set up the data and analytics.

00:19:58.500 --> 00:20:10.650

Vijaya Vishwanathan: Now, what do we… how do we do that in a structured manner? So, we look at it essence, and again, this is not all-encompassing, this is just what strategies we have found works best for us with our partners.

00:20:10.650 --> 00:20:27.490

Vijaya Vishwanathan: Is identifying members with cohorts of care, right? So, what do we… we talked a little bit about it. So, stratifying the members by risk, by opportunity, by utilization, all of those, and identifying those in your care cohorts.

00:20:27.550 --> 00:20:36.549

Vijaya Vishwanathan: Designing specific upgrade strategies by partnering internally with your, other service partners, right? That's part of it.

00:20:36.690 --> 00:20:56.509

Vijaya Vishwanathan: with the member outreach. And the third part is the provider partnerships. So having those provider engagement relationships and coordinators where they can help you drive those interventions, be it outreach, be it education, be it coding accuracy, any of those pipelines that need to be set up.

00:20:56.650 --> 00:21:01.329

Vijaya Vishwanathan: Those then become really easy to do once you have the data supporting it.

00:21:01.550 --> 00:21:03.390

Vijaya Vishwanathan: So…

00:21:03.460 --> 00:21:12.740

Vijaya Vishwanathan: it's more than just saying… reaching out to your provider partners and saying, here's a part… here's your data, here's your members who are at highest risk. It's…

00:21:12.740 --> 00:21:30.589

Vijaya Vishwanathan: more than that, to say, we can help you do the outreach, we can help you do the engagement, and taking it to the next step. If they need education, being that partners with them, if they need scheduling, making that data available to them, I think that's the next step that we talk about over here.

00:21:30.590 --> 00:21:31.060

Ramneek Kaur: Let's go!

00:21:31.740 --> 00:21:42.610

Vijaya Vishwanathan: This actually leads us into what we want to do, because talking about all of this in abstract is great, right? But what do we want to do as we are wrapping up 2025 and moving into 2026?

100

00:21:43.740 --> 00:21:46.950

Vijaya Vishwanathan: So, data-driven roadmap for 2026, Kamik.

101

00:21:47.760 --> 00:22:12.529

Ramneek Kaur: Perfect. Before we talk about, you know, when to do what programs, and when to initiate them, and why a particular time in the year is a good time, I wanted to quickly go through all the milestones when it comes to risk adjustment that plans will have, in 2026. So this, this is for Medicare Advantage. As you can see, in, you know, in any given year, there might be multiple,

102

00:22:12.650 --> 00:22:37.579

Ramneek Kaur: program years, dossiers that plans might be working with. So, in the start of the year, just wrapping up dates of service in 2024, and then throughout 2025, looking at post-encounter programs or retrospective charge reviews for, dates of service in 2025. And then in 2026, working on pre-encounter and point-of-care programs. You know, we all know there could be multiple

103

00:22:37.580 --> 00:22:46.669

Ramneek Kaur: of them there. So, I feel like a lot of times I work with plans that can feel that, you know, we're driving in multiple lanes at any given point of time.

104

00:22:46.690 --> 00:23:11.530

Ramneek Kaur: There's a final 2025 final run in 2025… in February. Then in March comes the mid-year risk adjustment run for 2026, and then in September, the initial run for 2027. And adding on to it, there are some estimated timelines for audit memos as well, which go to program years, you know, even prior to 2024. So all of this, just want to, you know, make a point here that

105

00:23:11.530 --> 00:23:36.120

Ramneek Kaur: in a given year, there's a lot that plans, need to do and should be doing, so key planning is, you know, the key message here is to plan your resources, your processes, so that you're not ignoring any of them, because each of them have an importance for that, your, dates of service year, and we can't ignore if we want to achieve the maximum potential for risk adjustment programs.

106

00:23:37.250 --> 00:24:02.209

Ramneek Kaur: now that we've looked at the timelines, what the milestones are, we are starting to get a sense of when should we start planning for these deadlines, and even given a… within a particular year, what all programs should we be doing? And that's our next step. So, Vijaya is going to start with the post-encounter programs, or the retro programs, and then we'll go further into the pre-encounter and point-of-care programs.

107

00:24:02.210 --> 00:24:06.960

Ramneek Kaur: at different milestones, and especially how it connects to analytics, I'm looking forward to that part.

108

00:24:06.960 --> 00:24:30.399

Vijaya Vishwanathan: Yeah, and as Rameek said, this is focusing on MA, because currently MA is the one that allows us the most retro, right? And other programs as well. Medicaid has its own complexities with state-to-state variances. Reach out to us if you'd like to learn more about Medicaid ACN, we are happy to share, but our next few slides are focusing on the MA timeline specifically.

109

00:24:30.610 --> 00:24:41.140

Vijaya Vishwanathan: So, yes, moving on to the retrospective, so we broke out the calendar year 2026 into two aspects, because, as you know, we run multiple things in parallel.

110

00:24:41.330 --> 00:24:54.330

Vijaya Vishwanathan: for risk adjustment. So, for 2025, what this slide is focusing on over here, if you see, is the retrospective chart reviews. So, how analytics can support the decision-making and

111

00:24:54.370 --> 00:25:10.299

Vijaya Vishwanathan: for key milestones specifically, is what this is focusing on. So as we kick off the new year, we already are completing the 2025 retrieval year, so we can actually get a leg up and review retrieval insights, make sure

112

00:25:10.380 --> 00:25:27.390

Vijaya Vishwanathan: we have an idea of, where digital retrieval works best, what providers are most responsive, what is… what timelines of those, locations that we have cleaned up or identified. So that kind of analytics can already be done in January and be ready for us.

113

00:25:27.560 --> 00:25:52.379

Vijaya Vishwanathan: By March, we always recommend about March or April for kickoff of the first MA wave, and one of the reasons for that is there is a lot of lag that can happen in claims, so usually waiting around March, we have seen, yields us the best results for wave one of the chase. We do always want to do two waves. There was a recent KFF study that came out too, and it actually proved this point to say that

114

00:25:52.710 --> 00:26:02.050

Vijaya Vishwanathan: Multiple waves, or at least two waves of retrospective chases, helps, close the most gaps and increases ROI.

115

00:26:02.270 --> 00:26:10.910

Vijaya Vishwanathan: So, around July, August, we want to run another wave of chase, and then use that Chase to compare with the first one to see

116

00:26:11.050 --> 00:26:22.969

Vijaya Vishwanathan: If there is any course correction that is needed, if there is any reprioritization that is needed, are there any new data points that we need to consider, any new members that came in, etc.

117

00:26:22.970 --> 00:26:46.540

Vijaya Vishwanathan: doing this in phases, you also don't want to wait too late and start in June, July, because then you're losing one part of the year, and two, it also delays some of your prospective programs. So, all of this is cyclical in nature. Anything that we are doing from a retrospective perspective… so, for example, in, say, in March, we kicked off a program, and for member A, you were able to find

118

00:26:46.540 --> 00:26:55.399

Vijaya Vishwanathan: four codes in the chart, and that's added to the member's profile, right? That automatically informs now the 2026 year.

119

00:26:55.530 --> 00:26:57.090

Ramneek Kaur: Prospective care.

120

00:26:57.090 --> 00:27:17.100

Vijaya Vishwanathan: right there to say these four additional codes should be tracked for either in-office assessment or when the member is being seen by a provider. So, whenever you're doing CDI programs or prospective programs, providers can have more complete information if our retrospective programs are accurate and drive that data forward.

121

00:27:17.250 --> 00:27:25.160

Vijaya Vishwanathan: So again, that is why it's important to also start early, but then do a phase later on as well.

122

00:27:25.300 --> 00:27:28.130

Vijaya Vishwanathan: Again, the other part that could help here is

123

00:27:28.340 --> 00:27:50.720

Vijaya Vishwanathan: when you know, based on the retrieval insights, when you have providers, you know where digital retrieval yields results sooner, you may want to prioritize those sooner, and then anybody that takes longer time, we might want to adjust accordingly. That way, you have enough time to close gaps, you're also able to realize the ROI sooner.

124

00:27:50.750 --> 00:28:05.100

Vijaya Vishwanathan: So, where analytics, again, plays into all of this is this bottom line right over here. So, ongoing cadence of generating the chase, prioritizing your members, and then, as you launch the programs, tracking the KPIs.

125

00:28:05.100 --> 00:28:13.970

Vijaya Vishwanathan: Is the chase yielding what you want? Is the ROI there for you to get the most for your members and providers? The coding accuracy.

126

00:28:13.970 --> 00:28:29.549

Vijaya Vishwanathan: how many charts are being coded, what's your new, net new added codes, all of that can be tracked through analytics, claims verification, validation, and finally, through the end of the year, how many net new codes did you submit, right? And then, as you…

127

00:28:29.580 --> 00:28:36.669

Vijaya Vishwanathan: evaluate all of this through the year, it kind of feeds into your next year risk adjustment program,

128

00:28:36.740 --> 00:28:38.230

Vijaya Vishwanathan: Driving those better.

129

00:28:38.870 --> 00:28:49.409

Vijaya Vishwanathan: So, moving on… so this is retro that we just talked about, right? Ramnik is going to cover the prospective aspect, which is happening parallelly, and then we'll go on to the next topic.

130

00:28:50.350 --> 00:29:15.350

Ramneek Kaur: Yes, thank you, Vijaya. So, we looked at the retro programs and the timing, and why that timing is important, the importance of two phases. Some of that ties, very smoothly into, the prospective programs, the pre-encounter and point-of-care programs. Before we just talked about retro, we looked at milestones, and a lot of times, we can, you know, we were stuck

131

00:29:15.350 --> 00:29:38.739

Ramneek Kaur: in that mindset of, we need to complete something before a particular deadline. But, you know, when we think of risk adjustment and care management going hand-in-hand, it's… the deadline is not to finish before, like, a particular month, or the deadline of submission. It is to start as early as possible, because when you're in 2026 and you're designing the pre-encounter and point-of-care programs, if you're waiting till, say.

132

00:29:38.890 --> 00:30:03.799

Ramneek Kaur: September, October, understanding that we will still have done something before December. From a member point of view, and from a care management point of view, you've lost most of the year. When is the care management happening? So the, you know, we are working with Medicare Advantage members who have multiple chronic conditions. Some of them could be… a considerable portion of them could be high-risk, and, you know, high-cost members, so if there are any

133

00:30:03.800 --> 00:30:28.349

Ramneek Kaur: interventions that need to be associated with analytics, or any insights we get from analytics. If we get them in, you know, September, October, we don't have any action from a care management perspective. There are members who need to work with their PCPs earlier in the year to have a care management plan, just to schedule their visits with their specialists, how many times they need to come in.

134

00:30:28.360 --> 00:30:37.610

Ramneek Kaur: So, you know, that is why the prospective programs for any particular year, dossier, it starts in the first month, as you would see. So…

135

00:30:37.640 --> 00:31:02.529

Ramneek Kaur: looking at data from retro. So, like Vijay said, retro and prospective, the data we get from each of those programs feeds the other programs. So, once we have the data from prior years, we know the, you know, we know the gaps that exist that were not even covered last year after all the programs, or that still need to be covered this year as they were last year. So getting those gap letters and starting to send these out to providers

136

00:31:02.530 --> 00:31:20.490

Ramneek Kaur: starts in, you know, the first month, or maybe the end of first month, because the providers need this info when members visit them, during the point of care. If there's information that is sitting with us, and it's not with a provider who can actually take an action on it, you know, that's not the

137

00:31:20.490 --> 00:31:44.309

Ramneek Kaur: the best use of it. So, that's why we start… we recommend to start as early as January. The second program that I want to talk about is the IOA program. Again, we want the providers to be educated and enabled during, the member visits. So, the appointment scheduling for IOA so that the members can see their providers as early as the first quarter to ensure that they have a care plan for the year.

138

00:31:44.310 --> 00:31:47.150

Ramneek Kaur: That needs to happen in first quarter.

139

00:31:47.150 --> 00:32:11.959

Ramneek Kaur: again, we understand that there could be some delay in data, and you know, also there are lessons that come with every program implemented. You work with maybe, an initial set of providers as you are, you know, gearing up your program, and then you want to include more providers. So we do recommend having two waves, or two phases. The first is in, you know, the first quarter of the year, and then maybe as you begin the second half of the year around July time.

140

00:32:11.960 --> 00:32:36.750

Ramneek Kaur: time frame, so that you can actually pivot and be agile based on the insights you receive as when you do the first phase of the program. You want to gear it up, you want to add more providers, so all of that will still… can still happen if you're doing Wave 2. In fact, we've seen, with our clients and, you know, some data that if the IO program is started after July-August timeframe, you know, you actually might even

141

00:32:36.750 --> 00:33:00.880

Ramneek Kaur: miss out on some of the potential and effectiveness of this program, and you don't receive the best results that you can. So, the two waves allow us the flexibility to be agile, to scale up the program, to get lessons and pivot wherever we want to be, to enhance that plan design and implementation, but starting early is still key if you want to achieve the maximum potential of the program, and

142

00:33:00.880 --> 00:33:11.539

Ramneek Kaur: If it's important in retro, it's even more important in perspective, because you want to be there, you know, you want to have that data utilized when the member is visiting the provider.

143

00:33:11.690 --> 00:33:36.179

Ramneek Kaur: The other program we talk about here is, in-house assessments, again, which starts a little bit later than I always, because we want to identify members who are not able to make it to the office, because of, some SDOH factors, looking at their, prior year's data, so we start a little bit later, but still in the Q2 or in the first half of the year, and then a wave to recommended, you know, in the second half of the year.

144

00:33:36.180 --> 00:34:00.619

Ramneek Kaur: When we are almost at the end Q4, which is, you know, October timeframe, we recommend that we collate, continuously collate data to look at how providers are doing, but being in October is that, kind of that last push. Who are the providers who are doing better with the help of our programs, but can we still assist them with member engagement, with, coding programs to, enable them to do, you know, their best when it

145

00:34:00.620 --> 00:34:25.519

Ramneek Kaur: comes to the ROI or outcome of this program. The ongoing cadence, like Vijaya mentioned, is something that analytics keeps on fueling all the time, which is, you know, you want to look at quality of coding, you want to look at pre-visit clinical claims, which is updated, you know, every week, every month, and that is provided to the providers, for their point-of-care programs, and even for pre-visit planning, you look at eligibility

146

00:34:25.520 --> 00:34:48.160

Ramneek Kaur: eligibility and enrollment verification, more often, of course, not every month, if not every month. And then documentation review is an ongoing process. You want to make sure that any program you're doing, that's accurate and has quality results. So all of these milestones, be it ongoing or, you know, disparate milestones, analytics is going to be the key, but

147

00:34:48.159 --> 00:34:57.800

Ramneek Kaur: just having data is not enough. Planning when do we have that data, and what data do we need to make sure that our plans, our programs are running smoothly, and giving us

148

00:34:57.800 --> 00:35:02.330

Ramneek Kaur: The outcomes, better outcomes for our members, for our providers, for everyone.

149

00:35:02.410 --> 00:35:12.609

Ramneek Kaur: we need to have all of this, you know, this chart in mind, is when do we need the data and what data do we need? Vijay, anything that you want to add to that?

150

00:35:12.610 --> 00:35:21.999

Vijaya Vishwanathan: No, I think the only other thing I wanted to say over here, and you made a great point, that it's almost more… timing is almost more important and prospective,

151

00:35:22.000 --> 00:35:46.769

Vijaya Vishwanathan: this is talking about timelines for MA, but if you look at ACOs, or value-based care entities that are risk-bearing right now, etc. too, there's only a 365 degree… a 365-day timeline over there, right? There it becomes even more important that as you go through the year, you're looking at prospectively giving care so that you don't… because you have a very small window to go back and close those

152

00:35:46.770 --> 00:35:47.810

Vijaya Vishwanathan: apps, right?

153

00:35:48.030 --> 00:36:02.659

Vijaya Vishwanathan: So, and again, I like how both these slides kind of talked about the full life cycle right in them, because retro fuels prospective, and then prospective sets you up for a next retro year, successful or not.

154

00:36:04.030 --> 00:36:04.740

Vijaya Vishwanathan: Okay?

155

00:36:04.740 --> 00:36:05.940

Ramneek Kaur: Good point.

156

00:36:06.460 --> 00:36:23.650

Vijaya Vishwanathan: Okay, moving on to the next slide. So, now we talked about… we've talked about the timing of it, we've talked about the data gathering and setting the foundation up from an analytics perspective. Let's go into program orchestration. And, this kind of

157

00:36:24.580 --> 00:36:34.800

Vijaya Vishwanathan: highlights, we spread it out in separate slides, but in reality, this is what it is, right? There are multiple processes happening at the same time, multiple programs running in parallel.

158

00:36:34.800 --> 00:36:44.730

Vijaya Vishwanathan: So, what this is trying to indicate is how disparate programs can limit potential value if you don't have a bird's eye view of all of it.

159

00:36:45.050 --> 00:36:54.459

Vijaya Vishwanathan: So a common challenge that we see is programs are managed in silos. So there is a team that handles retro, there is a team that handles prospective, within…

160

00:36:54.460 --> 00:37:14.129

Vijaya Vishwanathan: and depending on the size of the plan, payers, providers, etc, it could all be segmented even further, right? But what that leads to is limited potential value. You can see some of the issues that we are highlighting over here. If you don't have centralized reporting, it's going to be

161

00:37:14.240 --> 00:37:36.530

Vijaya Vishwanathan: really fragmented and hard for you to get an overview of our entire ROI. Non-integrated data sets could mean that you're consuming the same data multiple times and increasing, decreasing efficiencies over there. Provider abrasion. If we are reaching out to the same providers time and again for different programs, though, they are not going to be happy with us.

162

00:37:36.530 --> 00:37:39.380

Vijaya Vishwanathan: Multiple EMR platform supports.

163

00:37:39.710 --> 00:37:55.630

Vijaya Vishwanathan: That we also need to handle over here. So with all of this said, what does this mean in practice, right? If we don't effectively implement the programs, two big risks that we have seen, both from a member and a provider perspective, are

164

00:37:55.630 --> 00:38:06.069

Vijaya Vishwanathan: From a member perspective, we could engage the wrong members, right? We could engage the wrong members, because we don't have a full picture of their health, because we don't have the right member profiles.

165

00:38:06.070 --> 00:38:17.149

Vijaya Vishwanathan: or because we are reporting in silos, and so, system A is not talking to System B, and that's why the data is not feeding each other. So outreach can go to someone who doesn't have

166

00:38:17.660 --> 00:38:22.209

Vijaya Vishwanathan: A higher, need for, patient care at this time.

167

00:38:22.350 --> 00:38:41.239

Vijaya Vishwanathan: We could also run into the opposite problem, which is over-engaging members. We are contacting them multiple times for different programs. We identify the same members for an IHA, IOA, and then so we are reaching them multiple times, because data is not connected, and so that leads to frustration and reduced trust, right?

168

00:38:41.240 --> 00:38:56.720

Vijaya Vishwanathan: From a provider perspective, too, without orchestration and without understanding what programs we are running for the data as a whole, they can receive what I said earlier about overlapping requests from different teams.

169

00:38:56.720 --> 00:39:00.260

Vijaya Vishwanathan: That creates provider abrasion and loss in trust.

170

00:39:00.320 --> 00:39:08.689

Vijaya Vishwanathan: When data is fragmented, it kind of becomes really hard to provide… to prioritize staffing also. So.

171

00:39:08.830 --> 00:39:17.669

Vijaya Vishwanathan: it leads to inefficient staffing ratios, and our programs become less efficient. So, this is generally to indicate that

172

00:39:17.830 --> 00:39:34.229

Vijaya Vishwanathan: you know, the takeaway here is it's not just about the data and the programs you launch, but how you launch them. So, program orchestration is not just about operational efficiency, it's more about improving the member and provider experience as a whole.

173

00:39:34.230 --> 00:39:42.959

Vijaya Vishwanathan: Reducing the provider abrasion, and making sure that every touchpoint we have with a provider or a member is meaningful and data-driven.

174

00:39:46.030 --> 00:39:47.820

Vijaya Vishwanathan: Aramnik, you want to take this one?

175

00:39:48.080 --> 00:40:13.030

Ramneek Kaur: Yes, so, as Vija was talking about, the realities of working with programs, I think one of the biggest realities that we're all facing, you know, in this world is the CMS changes that we've been seeing recently. So, in the last 3 years, we saw the risk adjustment model change from V24 to V28. There has been a redesign in Part D risk adjustment. There's new star ratings methodology.

176

00:40:13.030 --> 00:40:37.909

Ramneek Kaur: you know, there's new interoperability rules, and also there's evolving compliance and audits and the requirements for that. So, you know, all the issues that Vijaya highlighted when the program management and reporting is in silos, if, you know, that was, like, a big problem, you know, usually then with all of these changes and coordinating with all the stakeholders, if they are not communicating

177

00:40:37.910 --> 00:40:47.679

Ramneek Kaur: with each other, about all of this, and there's no standardization, in all the data sets or all the processes you have, it becomes an even bigger problem.

178

00:40:47.890 --> 00:40:50.950

Ramneek Kaur: And it's not just, you know, it…

179

00:40:50.950 --> 00:41:15.190

Ramneek Kaur: just doesn't add complexity. It can even lead to gaps, you know, there could be a gap in alignment with, in different teams, so a strategy gap, there could be gap in communication, the data sets, managing quality, and, you know, even data validation. All of it impacts the ROI, the timeliness of the programs we're doing. There's a lot of effort that goes into it, as we've seen, you know, there are multiple programs running.

180

00:41:15.190 --> 00:41:39.510

Ramneek Kaur: So, all of these have a big impact on how programs are run, and gives us insights on how programs should be run. Even though we can, you know, talk about it, have a different separate webinar for that for an hour, I think we likely want to touch on why program orchestration is as important as making sure that your analytics is in place, and it's exhaustive.

181

00:41:39.510 --> 00:41:55.269

Ramneek Kaur: And knowing what data to use how. This is equally important to make sure that your processes, tools, resources, they're all aligned, they are in the center of everything, which, brings us to the next slide, where we,

182

00:41:55.270 --> 00:42:01.619

Ramneek Kaur: want to talk about, you know, where should program orchestration sit, or program management sit, so…

183

00:42:01.750 --> 00:42:26.740

Ramneek Kaur: You know, the… if you look at the graph on the left side, the client leadership and the risk adjustment operations, these are the two entities that, you know, we most commonly see. In reality, risk adjustment operations is not one cycle, but, you know, smaller, multiple cycles who are working with different, teams, even within the clients, like Vijaya mentioned. There could be a team who are looking at provider relationships versus

184

00:42:26.740 --> 00:42:42.679

Ramneek Kaur: retrospective or, you know, retrieval. So, how do we ensure that all teams at the client side or the health plan or health system side are talking to and are aligned on the same page as all the different teams who might be managing risk adjustment?

185

00:42:42.680 --> 00:43:05.290

Ramneek Kaur: operations. And the answer to that is that at the heart of it, at the center of these… all these entities is program governance and compliance, which, you know, takes us to program orchestration. So it could be a separate team, it could be a vendor, it could be within a client, whoever is doing… there is a team who needs to do… who needs to make sure that all the risk adjustment operations and

186

00:43:05.290 --> 00:43:30.240

Ramneek Kaur: and other goals from a health plan or a health system perspective, they're all aligned well, that strategy is created for risk adjustment end-to-end. So, program orchestration, this team, you know, is dependent on client leadership, of course, for guidance and support and understanding the KPIs and making sure that those KPIs are met, reporting on them, and they work with the risk adjustment operations for strategic insights

187

00:43:30.240 --> 00:43:37.500

Ramneek Kaur: they provide these insights, and they also manage any escalations that could impact the ultimate KPIs.

188

00:43:37.570 --> 00:43:51.639

Ramneek Kaur: And to ensure that all of this is in place, you know, how do you do it? How do you make sure that your program management, and even beyond that, your orchestration is central and integrated with all the elements?

189

00:43:51.640 --> 00:44:06.040

Ramneek Kaur: And for that, we usually follow a five-phased approach, as you will see on the right. So, discovery and alignment, strategic alignment, you know, that's the first stage of understanding why are we doing something, like, what are we doing? And then.

190

00:44:06.040 --> 00:44:29.800

Ramneek Kaur: Next, coming to design and mobilization, making sure that the program is designed in a way that we are able to work through different milestones, we have what we need, and if not, what are the risks? Highlighting any risks at that point. Just after that comes execution and monitoring, which is, you know, working with the risk adjustment operations team, making sure that those KPIs, the program tracking aspect that we highlighted.

191

00:44:29.800 --> 00:44:34.180

Ramneek Kaur: Making sure that we are monitoring the programs. Now, what…

192

00:44:34.350 --> 00:44:59.350

Ramneek Kaur: what is the success of this program one week into the start versus six weeks? Do we have those, you know, metrics ready for us for comparison otherwise? How does a 10% different from a 20%? Like, what is the answer to that? Where should we at any given… where should we be at any given point? Lastly, optimization. If you're not meeting any goals that we should be meeting, how are we ensuring that we are doing better? Are we starting?

193

00:44:59.350 --> 00:45:17.460

Ramneek Kaur: starting earlier next year? Are we, you know, scaling more? Are we making the focus more niche? Are we starting with a limited number of providers to make sure that we get it right? Are we doing two phases, or are we doing one phase? What is it that will get us to the stage where we want to be?

194

00:45:17.630 --> 00:45:42.610

Ramneek Kaur: And then, you know, a lot of times, it'll go back to 1. So from optimization, we go back to next year, and usually it takes, you know, it can take 2 or 3 years if you're just starting out, or if you're not as mature in these capabilities to get to the stage, stage 5, which is Nirvana, or steady state, which is where we all want to be, is that, you know, we… all our programs are giving us their maximum efficiency.

195

00:45:42.610 --> 00:45:58.599

Ramneek Kaur: or output, that we want to have, and so it usually takes a lot of iteration, so making sure that the orchestration teams sits at the middle of everything and is ensuring that everything is aligned and in place is really important.

196

00:45:58.820 --> 00:46:08.040

Ramneek Kaur: I believe that was the last slide from us before we, stop with a quick poll. Vijay, anything you want to add here?

197

00:46:08.250 --> 00:46:27.659

Vijaya Vishwanathan: I think, we mostly covered it. I think just to, you know, close the loop on this, what, we wanted to cover today was, analytics is your engine, so, you know, we use it quite efficiently, I think, within our space, and would, encourage you all to also use it

198

00:46:27.660 --> 00:46:46.039

Vijaya Vishwanathan: as the engine, powerful engine, it can be for program orchestration. I believe it provides visibility and intelligence that is needed to make informed decisions at every stage of this five-pronged approach that Ramnik just shared. So, during discovery, it can help with identifying gaps and opportunities.

199

00:46:46.040 --> 00:46:49.380

Vijaya Vishwanathan: In execution, it helps you monitor,

200

00:46:49.620 --> 00:47:07.599

Vijaya Vishwanathan: you know, the KPIs, and ensure that you're compliant. So that's the other part. I think we would all love to reach that state of nirvana, but given the changing compliance aspects, building that into your processes in the execution phase makes it such that we are able to react quickly, right, which could

201

00:47:07.600 --> 00:47:27.890

Vijaya Vishwanathan: help us get to that, you know, ideal state, steady state, even if there are changes thrown our way constantly. Without robust analytics, what we have kind of seen is orchestration becomes more reactive, while our goal is to be proactive, build it into our budgets to proactively launch all of this.

202

00:47:27.890 --> 00:47:50.589

Vijaya Vishwanathan: integrating data from multiple sources, as we covered, I think, in closing, just truly makes, analytics shine and ensures alignment across stakeholders. There are no… you don't have… there is no guesswork, right? You can go with data points to say, here's why we want to launch a retrospective, or here's why we want to launch a prospective program for a value-based care client, etc.

203

00:47:50.710 --> 00:47:59.750

Vijaya Vishwanathan: So I think, in closing, orchestration that is supported by analytics is ideal state, and that's what we would encourage.

204

00:47:59.920 --> 00:48:03.199

Ramneek Kaur: So with that…

205

00:48:03.200 --> 00:48:05.710

Vijaya Vishwanathan: Towards our last poll.

206

00:48:05.910 --> 00:48:06.460

Ramneek Kaur: Yep.

207

00:48:06.740 --> 00:48:22.800

Ramneek Kaur: So, yeah, before we wrap up today, we'd love to hear from you, so we want to start with a poll that tells us more about the challenges that your team might be facing, ineffective risk adjustment, orchestration, or anything related to program management that is

208

00:48:22.800 --> 00:48:35.680

Ramneek Kaur: preventing you from getting, the most from your risk adjustment programs. And so this is going to help us, guide us, design, you know, further content and resources

209

00:48:35.680 --> 00:48:39.110

Ramneek Kaur: For, the barriers that are phased out there.

210

00:48:39.170 --> 00:48:53.029

Ramneek Kaur: I believe that was the last slide, it'll be there, for a few seconds, and then as soon as we start getting in, start getting the responses, and we get all the responses, we're going to move to the Q&A section of the webinar.

211

00:48:54.530 --> 00:48:55.650

Vijaya Vishwanathan: Thank you, everybody.

212

00:48:56.920 --> 00:48:57.990

Ramneek Kaur: Thanks, everyone.

213

00:49:22.190 --> 00:49:25.220

Vijaya Vishwanathan: Okay, with that, I think we can open up for Q&A.

214

00:49:39.960 --> 00:49:44.229

Vijaya Vishwanathan: We are starting to see a few come in. Ramnik, I can,

215

00:49:45.190 --> 00:49:48.350

Vijaya Vishwanathan: Let me make sure I have the Q&A up as well.

216

00:49:51.290 --> 00:49:54.860

Ramneek Kaur: I can, yeah, I can, keep a track and ask you questions.

217

00:49:54.860 --> 00:49:55.750

Vijaya Vishwanathan: Yeah, go ahead.

218

00:49:56.610 --> 00:50:07.650

Ramneek Kaur: Perfect, okay. So, the first question we have is, how can risk-bearing organizations leverage advanced analytics to effectively orchestrate their risk adjustment programs?

219

00:50:08.790 --> 00:50:32.819

Vijaya Vishwanathan: So, we talked a little bit about it. I'm hoping that it answered some of the questions, but, from a risk-bearing perspective specifically, I think having those advanced analytics, because risk-bearing entities are looking more at prospective care and the CDI workflow, having that constant feedback loop at every stage, where we are able to generate those gaps prospectively early and often.

220

00:50:32.940 --> 00:50:44.629

Vijaya Vishwanathan: Specifically based on member scheduling, even. So you're not… we have to move, I think, from a, batch mindset to a really,

221

00:50:44.630 --> 00:51:08.679

Vijaya Vishwanathan: there's a word for it, and I'm forgetting it, but on-the-fly, on-demand analytics phase. So we have to have almost real-time data so that we can generate those gaps for members that actually are scheduled for this week, for next week, etc. Surface them, and then having a really strong coder workflow, I think CDI workflow over there, where there is an intermediate step of

222

00:51:08.680 --> 00:51:12.089

Vijaya Vishwanathan: Who are validating that data for you.

223

00:51:12.360 --> 00:51:25.109

Vijaya Vishwanathan: who are then feeding it back into the analytics to make it better for next time, and then surfacing the right kind of gaps. With integration into EMRs, I think is the perfect way to,

224

00:51:25.110 --> 00:51:32.889

Vijaya Vishwanathan: have advanced analytics driving your risk-bearing organizations. That's how we have seen it be most successful.

225

00:51:34.710 --> 00:51:43.759

Ramneek Kaur: That's a great point. And then, I think just related to that, I see, another question. How do you suggest we address data silos?

226

00:51:44.850 --> 00:51:50.310

Vijaya Vishwanathan: Okay, data silos are interesting, so here's what I would say.

227

00:51:50.310 --> 00:52:09.690

Vijaya Vishwanathan: and we are dealing with some of this in our own world right now as well. I think looking at it holistically instead of looking at it piecemeal. So when you are looking to build a member profile, what are all the data points you need, and how do you get them at an incremental level and not make it a pipeline that is

228

00:52:09.690 --> 00:52:23.979

Vijaya Vishwanathan: has to be enhanced each time you build a data point, right? Same from a claims perspective. So looking at it from that overall entity level, having a few really good data architects who can help you bridge those gaps.

229

00:52:23.980 --> 00:52:32.540

Vijaya Vishwanathan: And build data meshes and a CDO-type arch… central data office-type architecture is what I would encourage in order to break those data silos.

230

00:52:34.370 --> 00:52:59.180

Ramneek Kaur: That's a good point. One of the questions we have is with frequent CMS updates and interoperability mandates, what strategies can help maintain compliance without slowing innovation? I can start, Vijaya, and then you can add as you… if you think something's missing. I think the way we should see it is not that compliance and innovation are in conflict with each other. In fact, I think they go hand in hand. So, Vijaya talked about

231

00:52:59.180 --> 00:53:11.569

Ramneek Kaur: about analytics and making sure that the data you have is, you know, the right, the right source of truth. It is exhaustive. It has all the data elements that are needed to make sure,

232

00:53:11.570 --> 00:53:35.689

Ramneek Kaur: that, you know, the programs are the right programs for that member and for that provider. And then, you know, tying onto that, we talked about, you know, planning, which is important when it comes to compliance and making sure the data validation in general is, you know, well thought of, well planned out. So, anything that you're doing should have, you know, should be, focused, on with quality.

233

00:53:35.690 --> 00:53:53.789

Ramneek Kaur: with getting things right, and making sure that, data is, you know, the right source, and data is true. So, AI, ML, any kind of innovation, interoperability, analytics, advanced analytics is everything that helps meet those standards.

234

00:53:53.790 --> 00:54:18.649

Ramneek Kaur: Rather than go against it. I think the only thing which adds onto it is planning and, you know, making sure that program management and orchestration is the center, rather than looking at compliance as a separate arm, that's something that goes along with everything else, along all the other programs. And once we have that mindset, into the processes, the systems, training, everything, you know, we don't have to look at it

235

00:54:18.650 --> 00:54:26.140

Ramneek Kaur: as something different or conflicting than the program outcomes, but something that goes along with all the programs.

236

00:54:26.600 --> 00:54:33.200

Vijaya Vishwanathan: And also, that's a great point, Ramnik, and also building compliance into your analytics, right? So,

237

00:54:33.200 --> 00:54:47.840

Vijaya Vishwanathan: OIG is really good about, releasing some of our, initiatives that they are actually tracking plants and payers and providers on, so we can actually start to use those to build those out, the conditions that they are looking at.

238

00:54:47.840 --> 00:55:05.960

Vijaya Vishwanathan: the RADV-related details that they're actually looking at. So we can build those reports and build the compliance functionality as part of our analytics. So you're looking at it, so when you submit codes and when you submit documentation, ensuring that it is compliant, ensuring that it has all the supporting data that it needs.

239

00:55:05.960 --> 00:55:10.869

Vijaya Vishwanathan: Right at the point of submission will help us avoid it in the future.

240

00:55:12.820 --> 00:55:14.769

Ramneek Kaur: That's a great point.

241

00:55:14.800 --> 00:55:35.379

Ramneek Kaur: I think, lastly, we have, another question on AIML that you talked about. So, how do you see AI and machine learning, analytics, and interoperability, transforming risk adjustment in maybe, like, the next 12 months, or, you know, next 3 to 5 years? Give us your, short prediction, and, and, like, the odds are long, but…

242

00:55:35.910 --> 00:56:00.820

Vijaya Vishwanathan: It's going to be interesting. It's so rapid, the changes are coming in so rapidly, and the use of it in healthcare data and technology, it's evolved so rapidly that it's going to be really interesting to see how it shapes up. But I think the biggest way, AIML in a safe and moderated way can help us is with predictive insights. So, early identification of risk and care gaps.

243

00:56:00.820 --> 00:56:02.580

Vijaya Vishwanathan: For prospective members.

244

00:56:02.580 --> 00:56:16.460

Vijaya Vishwanathan: I think that's the biggest way we can utilize it. From a retro perspective, using NLP and other machine learning, AI, ML algorithms to review the charts and highlight the codes and

245

00:56:16.460 --> 00:56:23.900

Vijaya Vishwanathan: make it easily apparent to coders, I think is another way AI ML can be used. With,

246

00:56:24.130 --> 00:56:40.019

Vijaya Vishwanathan: SaaS application, specifically using agenting AI to surface some of these interventions that we talked about, right? None of this should be an afterthought, right? As data is loaded, launching actions to say, here are your members, you need to do a prospective.

247

00:56:50.400 --> 00:56:56.910

Ramneek Kaur: I think we might have, lost Vijaya, the…

248

00:56:57.040 --> 00:57:01.550

Ramneek Kaur: working from home conundrum. Each other… Yep.

249

00:57:01.870 --> 00:57:08.719

Vijaya Vishwanathan: Yeah, allowing them to interact with each other will also mean seamless exchange for holistic members, etc.

250

00:57:10.620 --> 00:57:35.609

Ramneek Kaur: We lost you there for a few seconds, but I think we captured the essence of it. I just wanted to add this, another interesting aspect that we're seeing more and more now is, how do you get clinical insights into advanced, you know, analytics and machine learning? You know, in the past few years, and when these were growing, we saw that there, there was data training, data, and, you know, giving us those insights, but as we see in the last few years.

251

00:57:35.610 --> 00:58:00.580

Ramneek Kaur: and, you know, even coming forward, there's more on how to get the clinical insights, training your data models more, implementing that, and we've seen, some of the results that come, you know, the positive results that we can see. Like you said a while ago, when you do it right, you can, you know, see the results just there and then, is the output of the analytics that we get when it has clinical insights.

252

00:58:00.580 --> 00:58:10.279

Ramneek Kaur: It's implemented in it, and the data is trained well, and there's governance and transparency around it. The kind of results we get are really different and very positive.

253

00:58:10.510 --> 00:58:32.489

Ramneek Kaur: So I'm excited, too, to see how it changes more, just both from the administrative part of the job when it comes to coding programs and making sure we have all the data, you know, it makes up a little bit for time, and then we have more data, and how well we are utilizing that data to give us the insights that are needed, and going back into the system, so I'm excited as well.

254

00:58:33.080 --> 00:58:38.230

Vijaya Vishwanathan: Yeah, and you can bring up a great point about transparency. Having AI ML

255

00:58:38.340 --> 00:58:50.760

Vijaya Vishwanathan: But having that transparency to say, here's why AI thinks this is a gap, and having the explainability behind it goes another step towards gaining that provider and member trust.

256

00:58:51.620 --> 00:58:54.049

Ramneek Kaur: That's true.

257

00:58:54.170 --> 00:59:10.090

Ramneek Kaur: we have a lot more questions, but I want to look at time and be mindful of everyone's schedule, so I think that would be the last question we're taking. I want to give it back to Bailey, just to close out, and if there's anything that we missed.

258

00:59:11.060 --> 00:59:35.889

Bailey Fields: Yeah, so thank you so much, everyone, for attending, and thank you so much to the Optum team for a great webinar today, and partnering with us to give us all that valuable information. Before you log out, I just want to give a gentle reminder, that you will see a survey pop up on your screen. If you could just complete that, for us, that would be great and much appreciated. But with that, that concludes today's webinar, and we look forward to

259

00:59:35.890 --> 00:59:37.390

Bailey Fields: Seeing you all again next time.

260

00:59:37.390 --> 00:59:38.430

Bailey Fields: Have a good one!

261

00:59:38.660 --> 00:59:39.330

Vijaya Vishwanathan: Thank you, everyone.

262

00:59:39.330 --> 00:59:39.709

Ramneek Kaur: Thanks so much.