Driving efficiency and accuracy across the middle revenue cycle

Hello, everyone, and thank you for joining today's webinar, Reimagining Your Middle Revenue Cycle with Optum Integrity. One name is Brittany Terman with Optum, and I will be your host today. Before we begin, Please note the following housekeeping items. At the bottom of your audience console are multiple application widgets that you can use to customize your viewing experience. If you have any questions during the webcast, you can click on the Q&A widget at the bottom of your screen to submit a question. We do capture all questions and we'll provide follow up to questions as appropriate. If you experience any technical difficulty, please click on the Help widget. It covers common technical issues. You can expand your slide area by clicking on the Maximize icon on the top right of the slide window or by dragging the bottom right corner of the slide window. There is a survey widget which you can use at the end of the webcast to provide us with feedback on today's presentation. Additionally, this presentation does use streaming audio. You may listen to the audio through your computer speakers or headphones. To ensure the best possible system performance, please be sure to shut down any VPN connections and connect directly to the Internet. Presenting today will be Lori Sides, VP of product management for Middle Revenue Cycle. I'll go ahead and turn it over to Lori now. Thank you and hello everyone. I thought it would be fun to kick things off with a quick survey question. When these survey questions present throughout the presentation, take a moment to put in your answer. We're going to give you about 30 seconds there and then we'll review the poll results and keep moving through the presentation. So the first question, what do you think of when it comes to transformation of the middle revenue cycle? I love the responses. So some of you said, you know, I think it's time we become more proactive, interesting enough with ICD 10 and access to information with the install of your EMR, there really is an opportunity to rethink things that often occurred after discharge or after coding automated. It is all the buzz and we're going to talk a lot about that. And during today's presentation connected, there really is a move happening across the market where what have traditionally been siloed teams. We're starting to see teams come together to help create that streamline approach from point of whether it's medical necessity decision making to point of documentation, integrity and coding connected and finally, outcomes. Really starting to think about from a connected standpoint, being able to follow patients as they move in and out of your health system, whether it's physicians providing at risk care and risk adjustment models or trying to help Dr. outcomes from pay for performance issues such as risk of readmission. I'll keep moving in today's presentation and thank you for participating. I'm, I'm excited to talk about transforming middle revenue cycle operations with you today. We have all heard the adage, if it isn't broken, don't fix it. But I'd like to challenge this thought by asking a thought provoking question. Maybe your operations are not broken, but are there opportunities to do things more efficiently to drive new or improved outcomes or shift to a more proactive versus reactive approach to your biggest challenges? The truth is, for many of us, ICD 10 was a monumental event in the world middle revenue cycle. What if I told you that market dynamics are creating yet another seismic shift that will force us to reevaluate our people, our process and our technologies? Let's talk about some of those market drivers and industry challenges. More than 50% of hospitals will have negative margins this year. There is both a coder and a clinical shortage driving up cost, and that is if you find the resources. Hospital CE OS actually ranked personnel shortages #1 on their list of immediate concerns. Revenue integrity has also moved to be top of mind due to increased payer scrutiny, pay for performance pressures that are forcing additional transformation. And then what? Everyone's talking about the promise of new AI technologies. Advancements in large language models really have the market talking about not just automation but humanless automation driving top down initiatives to advance your automation. It is not uncommon to see chief automation officers within a house system and and and the building of an infrastructure to also monitor and certify the use of AI in your in your healthcare setting. There's also the clinician reported burnout is at a all time high with two or three physicians, according to AMA. Feeling burned out, talking a lot about the administrative time they're spending inside the EMR, selecting charges and codes, clarifying documentation queries, etcetera. You know, in support of those things, average denial rate is up almost 12%. And while denials are not new for healthcare, the rise in clinical encoding, denials and audits has people concerned. 25% of those denials occur at point of middle revenue cycle. Middle revenue cycle staffing shortages are 62% of open positions for the revenue cycle. These can include medical coders, clinical documentation, integrity specialists, and coding auditors. Interesting enough, when it comes to automation, a survey of middle revenue cycle leaders were asked if they plan to invest in automation and roughly 56% of those middle revenue cycle cycle leaders answered yes. However, according to a study by the HFMA, 95% of those survey respondents said they're planning to invest and the number one place was in that middle revenue cycle area. Why the need for transformation? Now that we've grounded ourselves to market challenges, let's dig a little deeper. Interesting enough, there's been 25 years since the implementation of IPPS and OPPS guidelines. For many of the challenges introduced under those guidelines, great leaders and innovative thinkers like yourself have figured it out with people processing technologies. And while we make great headway, disjointed teams, disjointed processes, and disjointed technologies still allow some things to fall through the cracks. We are seeing more and more organizations daring to collapse the gaps between teams such as utilization review and clinical documentation improvement as well as coding because there's recognition that critical decisions made every step of the way must come together to tell the story on a patient claim, not just a point in time. In our process, we're hearing customers looking to reduce the vendors needed for wrap around solutions requiring multiple solutions to support and multiple staff to support those systems. We also hear organizations talk about the cost of reconciling work across discipline, whether it's CDI, encoding, reconciliation or utilization review, looking at those critical inpatient observation decisions as compared to what finally was determined in the clinical documentation integrity process. The task and process that collectively make up the middle revenue cycle represent the essential lifeline linking clinical care and financial reimbursement for organizations such as yourself. So how is often approaching middle revenue cycle transformation? We believe it is critical to connect the middle revenue cycle for improved outcomes through a single technology platform that not only connects the dots and critical components such as charging, coding and documentation, but that every step of the way there is a check and balance between these critical elements. In order to achieve results, we've invested heavily in automation organizations the need both autonomous capabilities and more efficient tools for partial coding of complex cases and ultimately enhancing the user experience, especially when it comes to interacting with AI. In our fast track to enhance automation, it is also critical that automation does not improve middle revenue cycle efficiencies to simply displace the burden of accuracy downstream for the back end of your revenue cycle. It is our mission to automate without sacrificing actors. This means understanding when a human should look at a case due to reimbursement or pay for performance risks and creating those wrap around technologies to facilitate that work. It means leveraging the traceability of documentation and coding or coding and charging to provide the insights needed to avoid downstream denials or downstream audits. In addition, we have surrounded our tools with powerful business analytics to assure you have the ability to manage your operations, report on your outcomes and ultimately extensive tools for identifying opportunities for improvement or additional impacts. We're also addressing point of care strategies. These are top of mind with more and more organization taking steps to reduce that administrative burden for their physicians. We're looking at how ambient documentation and greater integration into EMR workflows can take away that administrative burden, whether it be from code selection or answering queries at home. Finally, last but certainly not least, we've made investments in our AI infrastructure to fuse together our years of clinical language intelligence experience in the middle revenue sector space with generative AI solutions and large language models. These have created higher accuracy rates and enabled enhanced autonomous output. So why don't I introduce you to Optimum Integrity 1? And how does this vision on the previous slide we talked about translate to transformation, our fully integrated, newly launched technology for enabling your middle revenue cycle operations and assisting your teams in achieving outcomes. I'll start this conversation off with Transformation support. It is never just as simple as choosing a technology transformation requires attention to detail. Evaluate evaluating your current versus future workflows, educating your teams and creating a motivation for change. Establishing baselines to measure your success. In addition, it is key to evaluate your current staffing models and evaluate the current scope to identify opportunities to expand and advance your operations with enhanced automation. Operating at top of license is a term you likely hear a lot from service and technology vendors, but transformation means establishing a plan to get to things you couldn't get to before, to evaluate and audit your automation vendors, to assure compliance, and to rule out risks, all leveraging your most valuable resources in a new way in this automated world. Experienced and modernized AI Optum Integrity 1 leverages our experience coding for healthcare organizations for over 20 years. While automating that with modernized technology that complements and advances automation. Reporting is also important. So through the advanced analytics, scheduled and ad hoc real time access to data leading to actionable insights, and a single data repository that combines the power of clinical and coded data, unified and comprehensive platform. It's a platform that truly bridges the gaps between critical middle revenue cycle operations. Whether it's a longitudinal view of a patient as they travel through your health system or the ability to code a facility in professional case, or assuring the claim represents the acuity of your patients, the medical necessity for the care you provide, and represents the quality of care your patients receive. So I'm going to move into what I mean by connecting the middle revenue cycle. The optimum integrity One solution is comprised of middle revenue cycle capabilities intended to connect previously disjointed processes to drive improved outcomes. Clinical documentation integrity is powered by clinical language intelligence to evaluate the story being told in the clinical record against the stated diagnosis, hence understanding what hasn't been stated or areas where additional specificity is needed. This 24/7 evaluation of information prioritizes cases with a documentation opportunity, identifies potential quality events that have occurred, and evaluates the coding against the documentation for clinical validation. This drives a market differentiating exception brace prioritization for your teams. Charge Capture Leverage leverages our proprietary algorithms for determining high value and high complexity outpatient charge element such as your Ed visit levels, observation, charging, clinic charges and infusion and injections. Coding is also powered by clinical language intelligence and evaluates what the clinician has stated and translate translates that into coding nomenclature, your IC10CPT, PCs codes, etcetera. This capability includes autonomous coding capabilities for professional and now expanded to hospital. The good news is we're not limited to volumes that can be autonomously coded like some new companies in the field, but offer a full suite of computer assisted coding, autonomous coding, and rules driven coding by exception that you'll hear more about as we move forward. Audit Compliance supports your team in evaluating risk for your middle revenue cycle. This fully integrated tool allows for operational and outcome driven audits. Whether you want to review the code assignment by a new coder you have hired within workflow tools for education or rules driven pre bill audits for high risk denials or financial impacts, or in support of automation and pay for performance initiatives. And finally, the analytic capabilities to measure those outcomes, monitor your operations and key performance indicators and identify ongoing opportunities as they arise. So as we wrap up this section, I would reinforce 3 critical elements of the future transformation Efficiencies in automation, both full and partial automation capabilities, single vendor strategy, reducing your total cost of ownership by having a single tool that stretches across your middle revenue cycle and financial optimization with accurate and defensive charging, coding and documentation. So let's talk about what I mean by enhanced automation. We'll start off with a poll question. Again, we'll give you about 30 seconds. I have not seen. I see some results, but I may not have given it enough time here. We will keep moving. As we dive deeper into automation. I want to take a moment to revisit the automation available from Optum Integrity 1. So let's start with coding. Think about this is understanding what the physician has stated in their documentation. So an example would be the physician document sepsis without any clarity of Organism and clinical language intelligence codes. A 41.9 sepsis unspecified CDI is where we're finding what the physician did not say. From a coding standpoint, CDI evaluates clinically significant facts and findings indicative of an information gap. For example, the following clinical indicators were found on a case documentation of an elevated BNP lab, medication administration of lasix and pulmonary congestion on a chest X-ray result, but yet there's no diagnosis of CHF. In addition, CDI evaluates documentation for the mention of a quality event or a condition that was not present on admission. And this really feeds strongly into that proactive approach you indicated in the first survey, where instead of a waiting until we've coded a quality condition on a case, we are proactively asking quality to confirm this in fact is a quality event. Finally, we leverage a check and balance between the clinical facts and codes on a claim for clinical validation. For example, let's take that diagnosis of sepsis for clinical validation. So now the physician has stated sepsis and we're going to test that against the clinical facts on a case. So clinical elements pertinent to sepsis to identify documentation support is present. For example, the patient had documentation of IV antibiotics, altered mental status, tachycardia and an elevated serum lactate and A and a lab result of low platelets. In that event, this certainly is going to meet the support for that diagnosis of sepsis, hence passing that clinical validation check. Enhanced automation is more than simply identifying a term in a clinical record and mapping that to a code or clinical outcome. It's about assessing the indications, attributes, relationships, translating to industry coding, nomenclatures and then reasoning over these those findings as a check and balance. For example, indicators are the core clinical concepts found within your clinical records, signs, symptoms, diagnosis, labs, vital signs, etcetera. Attributes are the detailing factors and contexts needed to drive accurate meaning from these indicators. A history of versus acuity severity or previously diagnosed with. For example, the patient record may contain a mention of atrial fibrillation which on its own could trigger a code of I48, but by discerning the attributes surrounding that term, it recognizes that the patient said his father has a history of atrial fibrillation and hence Noah code assignment would occur. Relationships are inherent and contextual associations. What is the principal diagnosis or indicator for or symptom of? I think that we can all understand what the translation coding level means, but let's talk about the last box, clinical markers. That's how we reason across a clinical profile. Imagine your work list was a list of every patient in your organization with an abnormal hemoglobin. You would overwhelm your teams with documentation reviews that yield no result. But instead there's a patient on your work list with a hemoglobin of 6 who's receiving blood products where the physician has not written a diagnosis of acute blood loss anemia. This is just a great example of how indicators and attributes help sort through the noise in your clinical relationship. But by being able to reason over those those findings, we're able to find deeper and reduce phosphor, find deeper intelligence and reduce those false positives by understanding both what the clinicians have stated in the form of a diagnosis, but also those clinical indicators suggesting a story that hasn't yet played out in the clinical documentation. So why choose Often as a middle revenue cycle automation vendor, you want to choose a vendor with a flexible architecture and built to scale on pace with technological advancements. Innovation is occurring at lightning speeds and you want to assure you invest in a vendor who will not become obsolete overnight. Many newcomers in this market require resource heavy model training from your staff as well as validation. Our 20 years of experience means you get the benefit of an experienced model with large data sets to evolve new AI capabilities and the in the event of coding and regulatory changes, which we all know happened and there's no data to train a model, you're not left zero with 0 automation abilities. Optum models are validated by a combination of clinical coding, HIM, medical record experts and AI experts. And that position us positions us uniquely as your partner today and in the future. Let's talk about shifting left and preventing rework. We've heard this for years, but what does that really mean? We'll start with a poll question. I'm going to try to give it a little more time this time. So let's take 30 seconds to answer a poll, plug a polling question, and then we'll keep moving. OK, I still didn't get results. All right, I will keep moving, shifting left and preventing rework. Many middle revenue cycle leaders face a real challenge. You have to choose between going fast and implementing tools and processes to solve for clinical and coding related denials, years of being productivity driven and everything pending on your discharge, not final coded numbers. It's often hard to say slow down and get it right. Transformation means assessment of the cost compared to the benefits of shifting left. How does it impact your productivity versus avoiding downstream rework and the potential for lost revenue? A cost benefit analysis will help you make the right choice. Having automation at the point of documentation coding creates a view into the potential claim every step of the patient journey. As we think about I CD10 and an implementation of computer assisted coding tools, you know, when IC 10 was over and we'd kind of gotten through that change, many of us said I'm not sure that that computer assisted coding makes us an faster enough to cover the cost. And so as we think about this, this conundrum, right, what we really accomplished during that ICD 10 time frame was putting tools at the point of care with ongoing assessment for gaps and documentation and assignment of codes. Hence, creating a solution of concurrent coding that we've all tried before and allowing us a view into a claim or a view into a potential quality event long before a patient leaves the hospital. Transformation means assessment of the costs and determining if the benefits are there for your organization. Having automation at the point of documentation and coding creates that view into the patient journey. How you surround these capabilities with people and process affords you to drive new and impactful outcomes for your organization. Our audit and compliance module that is part of Optima Integrity 1 is one of those tools that that would help you do this. This can be your organization's first line of defense for tackling challenges that originate in the middle revenue cycle. This support, this affords a pre implementation assessment from your 835 and 837 claims data to pinpoint those various denials to help quantify those as clinical or coding denials. Audits that you may be receiving that also make their way back to the point of origin audit compliance not only enables that pre bill correction but also provides immediate feedback to the teams to course correct different error patterns. Another option is a tool that affords middle revenue cycle to code by exception for high volume caseloads in the ambulatory environment where you may have made a decision for physicians to select codes within the EMR. This affords a rules driven exception based work cube tailored to your organization. For example, if the physician failed to capture a hierarchical condition code, but the technology identified A hierarchical condition code, then it would route for validation by your coder or auditor or perhaps the physicians assigning an ENM Level 3 where the technology extracted elements equating to an ENM level 5. Again, your rules drive the things you look at and you're not at the mercy of what's selected in the EMR, but you're on and also not having to ship back to an all or nothing approach by having coders touch every ambulatory visit. A simple graphic highlights this approach further. Imagine a 50 year old patient presented to the emergency department with chest pain and a cough. As expected, full cardiac workup ensues with frequent assessments of the patient condition. At point of charging a visit, Level 5 is a sign. However, upon completion of the assessment and the ancillary findings, it was determined the patient had an intercostal strain with a reason for visit of cough. The validation step recognizes that a visit level 5 for a cough and intercostal strain is subject to scrutiny and may not support the medical necessity of the testing that was done. The coder recognizes the missing reason for visit of chest pain and upon revalidation the case can be sent to visit to billing. These are just some examples of shift left capabilities leveraging data and analytics. There was a time where we didn't have enough data and now we're overwhelmed with data. So let's pause for a moment and ask one more survey question. Which of these is most important to your organization's middle revenue cycle transformation? I'm going to keep things moving. Leveraging data and BI analytics. We've talked a lot about leveraging data and BI analytics along the way today. The graphics here reflect real time reporting and dashboards. There are numerous real time reports, the ability to bookmark and interact visuals with extensive drill down and export capabilities. And one of the most popular features among our customers is the ability to decide what reports you need and schedule those reports for automated distribution to the stakeholders in your organization who need them. In addition, we offer ad hoc customer report designer and the ability to export your data to the reporting platform of your choice. Point of care strategy. As we think about the future, Optum recognizes the rapid adoption of ambient scribing technology to help ease the administrative burden for physicians. The market is certainly excited about these technologies and physician adoption is growing. With that comes new complexities to documentation practices. They're impacting revenue cycle outcomes while also creating opportunities to better integrate within the physician workflow and to work with ambient vendors to assist in strong documentation output for helping you achieve your automation goals. This example is a combination of capabilities being leveraged today with an exciting future to partner closely to accomplish unprecedented outcomes by not only creating automation in the middle, but imagine that same automation in the front, middle and back revenue cycle, taking these more simplified transaction and being able to expedite patient responsibility and payment. OK, we're going to go a little bit deep on automation advancements in this section. We've talked about a advancements and the rapid evolution of AI technologies, but what does that have to do with transforming your middle revenue cycle? AI advancements and the promise of higher volumes of automation is all the hype. This can be largely tied to large language models, a type of artificial intelligence algorithm that uses neural network techniques with extensive parameters to process and understand human languages or text using self supervised learning techniques. These models are trying to vast amount of data, enabling them to perform a wide range of tasks. I happened to notice on social media last night that a friend had asked a large language model to plan their meals for the week, and it was pretty impressive. Not only the meals they suggested, but tips and tricks for making those meals more fresh if that's what you're into. And then of course, there was also several people who had large language models roast their family photos. I found that interesting. In our fused architecture, we bring to our customers the best of both worlds. On the left you see the large language model benefits and on the right the benefits of our symbolic AI approach. So large language model benefits include knowledge and Internet scale, fuzzy content based understanding of text and natural language prompts, and a powerful ability to perform limited reasoning on simple multi step problems. Often, CLI and the approach we discussed earlier in the presentation leverages indicators, attributes and reasoning to create a more explainable and verifiable clinical detail. The platform is supported by 26 years of cumulative subject matter, expert clinical knowledge, resources and experience reading health information to drive outcomes. In our testing, we found that in fact the large language model could not compare to the explainable experience and knowledgeable model we offered for that. However, when combined together, the two produce a double digit improvement to precision and recall metrics and increased autonomous outputs. Why? When you think about the complexity of a clinical record knowledge base of things like coding regulations or clinical meanings attributes, the risk for hallucination and modern models goes up. What we discovered is the ability to leverage our existing models to build a foundational knowledge graph for the large language middle model that provides a powerful combination. So in this diagram is an example of an often clinical language intelligence derived knowledge graph taking the indicators from a clinical encounter and breaking that down into critical elements or decision making. As you follow the graphic left to right, you will know the critical granularity required for middle revenue cycle automation elements. If you haven't attempted to code a medical case or apply guidelines using a large language model model such as chat, PGPT, or Microsoft Copilot, I highly recommend it. And so I'll give you an example and I hope you can imagine this with the diagram on the screen. I asked the large language model what is the code for the Cubitus ulcer. It quickly returned an ICD code of L89. Of course I realized that location and laterality matter. So I re asked my question of a decubitus ulcer of the left heel and it quickly returns an L 89.629. But there's just one other element needed to drive the correct code. So I asked to code the decubitus ulcer on the left heel. That is the stage 4 and of course it returns the L 89.624. What is impressive is the accuracy for the data presented. So if I ask a specific enough question, they return inaccurate code. Where the magic happens is when the knowledge graph has associated the appropriate attributes to those indicators that help create a higher level of accuracy and reduce the hallucinations for misassociated terms or interpretations. So if you look below, you can see CPT codes. The patient received a cortisone injection. But to accurately code that, we need to know when the laterality and all the different elements needed to assign that code. And so this is where leveraging our experience allows us to. It's almost like taking the entirety of a clinical record and creating the right associates associations to either prove a term is relevant or to identify that in fact was a mention of a history of or a previous work up or a work up they're going to do today versus a confirming a diagnosis. The optimum autonomous coding models are built on the fusion of multiple AI capabilities. Grounded in the knowledge graph we saw on the previous screen in order to produce high performance across specialties and all coding categories. These results are also accompanied by case and code level confidence scores, creating powerful configurability and enhancing automation. This fusion of models means as technology evolves, we can test and add capabilities and even retire lower performing capabilities to render the best results. It feels like every week new technology models are entering the large language model space and it's moving rapidly. That means you need to partner with a vendor who has an architecture that can respond quickly to technology advancement. But it's also not limited to just one approach. Some of these models are better at different service types or different code types. And so it's not enough to just bring one form of AI. The fusion of multiple AI capabilities comes together, analyzes that knowledge graph to produce the best results. So let's talk about partnering for transformation. I have a survey question. I dare to try it, but I'm going to do it. How likely are you to change your middle revenue cycle technology or service vendors in the next year? Thanks so much for your responses. I have never been good with the pregnant pause. Maybe I rushed the 32nd timer. All right. Partnership opportunities, measuring success and driving outcomes. At Autumn, we take your transformation serious. We're not just selling a service or a technology and walking away. We are partnering in your success. Opt in value assurance starts in the sales process and with a pre sale assessment and it follows you through the build training and go live and continues throughout the life of the partnership. As we go into a a sales evaluation that becomes foundationally the baseline metrics that allow you to hold us accountable for the cost benefits we commit to. We provide you with access to coding and automation experts with deep product knowledge and all subject matter expertise experts in their field. We provide baseline to target reviews and KPI tracking. In addition to the advanced analytics I shared with you. We provide yearly health checks, strategy sessions, and goal alignments. This is going to be super critical in an automated world where you need to constantly evaluate new documents to coming into your organization and how to streamline those for the best automation results. We also provide national and Optum peer group comparisons and action planning. When you fall below those expectations, we do on request analytic deep dives. If there's a new objective for your organization and you want to analyze how the middle revenue cycle can help make an impact, we will do these deep dives with you, help you create the dashboards you need to measure and schedule those reports for the stakeholders in your organization. Partnering for transformation is, is more than just the install and it's more than just the reporting. And based on comparisons we provide the product experts as I mentioned we we also provide a relationship liaison. This is your advocate, this is your voice into our enhancements issues that you may encounter and overall and support our product opportunities. We also provide industry webinars webinars that your staff can attend they ought. We also offer CE us for those webinars. Client Learning Community is an online comprehensive computer based training site that allows you as you onboard new coders or need to re reiterate user adoption with the existing coder. CDI staff other users and we provide our release notes and user guides there to help you get up to speed on your product's most useful capabilities. We also do product spotlights. These are quarterly user group sessions to share your experiences and best practice with organizations similar to yourself and to drive Rd. maps and innovation through that sharing. Each month we highlight a specific product feature providing in depth insights and usage tips. This concludes today's presentation. I would like to thank you for your time and just would love to hear back in the comments on any feedback you have that I may have missed and where your thought process is on transformation. I look forward to seeing you out and about and industry events. And that concludes my portion of the presentation. Thank you. Thank you so much, Lori for sharing your insight and expertise with us. We again would like to thank you, our audience, for sharing your valuable time. We know that you're all busy and we appreciate your attendance. As you leave a reminder, we do have a survey widget to provide us with your feedback on today's presentation. We appreciate any feedback that you would like to give. Thank you so much and have a great rest of your day