Key Takeaways:

• Each year Iodine conducts multiple cohort studies, focused on productivity improvements and overall impacts to performance experienced by clients throughout their journey with Iodine
• In 2022, 94% of facilities experienced a lift in productivity with Iodine, with the average facility seeing a 136% lift in normalized query volume
• With Iodine, physician response rates either stay steady, or significantly improve in cases where Interact has also been deployed, even in the face of increased queries
• Iodine also measures the impact queries have, including CC and MCC capture volumes, CMI, and GMLOS. The vast majority of facilities experienced an improvement in MCC capture with Iodine, with a median of a 27% increase in the number of cases that have an MCC
• The increase in MCC capture resulted in significant financial impact: an additional $3.5 million in annual reimbursement

Methodology

Iodine’s Cohort Studies started as an internal initiative in 2017 in an effort to understand the impact clients saw with AwareCDI and how certain key metrics were trending. Each year Iodine conducts multiple cohort studies, some focus on productivity improvements, others report on overall impacts to performance clients experience throughout their journey with Iodine. Cohort studies are completed by comparing pre-Iodine data as a baseline against the most recent CMS fiscal year’s data with Iodine. Every facility for which there is at least two months of baseline data and two months of measurement data is included in the study. Only like months are compared to account for seasonal variety (i.e. January-March of 2021 is compared to January-March of 2022) and for DRG based cohort studies, only that year’s CMS DRG tables are used.

The intent is to examine: if a CDI team performed at the same level as they did pre-Iodine, what would that performance have been, and what is the difference between that modeled performance and what was actually observed with Iodine.

Productivity

Iodine examines productivity through two main means, unique case query rate and normalized query volume. Iodine measures both query rate and query volume to provide as complete a picture of CDI performance as possible. Since unique case query rate is an indicator of whether CDI specialists are looking at the right cases, it is possible to have a high query rate by only reviewing those cases that you feel confident you will query, but as a result your query volume will drop. Conversely, you can drive up your query volume by increasing staffing, but that’s not very efficient. Examining both query rate and query volume ensures that Iodine has positive impacts on both and is driving as much value for clients as possible.

Iodine measures normalized query volume as a measure of the overall output from a CDI program, as more queries lead to more accurate documentation which has a variety of downstream affects including accurate reimbursements, quality reporting, and more. Iodine adjusts for changes in CDI staffing over time, and compares the incremental queries sent in the fiscal year 2022 against the baseline period, the year immediately prior to a facility adopting AwareCDI.

Higher productivity and more efficient and effective processes enables clients to expand the scope of their programs and accomplish more with their existing staff. One client had a team of 17 CDIS and after implementing Iodine they increased their query rate from 21% to 35% using only half of their staff, enabling the remaining staff to be redeployed to performing retrospective reviews. Other orgs have leveraged the freed up man-hours to increase collaboration with other departments like coding, implement second-level reviews on mortality or PSI cases, and improve job satisfaction as CDI specialists work at top of license.

Physician Response

When reviewing and responding to a query is no longer a burden for physicians, response rates and times can improve, even in the face of additional queries. On average, there is an almost 17 hour savings in response time, with CDI specialists waiting 31 hours or less for a physician to answer a query, meaning CDI specialists don’t have to spend as much time and energy chasing down responses. The average physician spends 60 seconds or less reviewing and responding to a query, and the median physician response rate for an Iodine client with Interact is 94%.

Impact of a Query

In addition to lifts in productivity, Iodine also measures the impact those additional queries have, including CC and MCC capture volumes, CMI, and GMLOS, with the theory that as documentation gets addressed, it more accurately reflects the true acuity of the patient population, and as a result these measures naturally go up. The more the productivity of CDI teams improves, the more one can expect MCC capture to increase, as well as CMI and GMLOS to a certain extent.

In line with our productivity metrics, the vast majority of facilities, 90%, experienced an improvement in MCC capture with Iodine. On average, their MCC capture volume improved by almost seven percentage points, which ultimately resulted in a 27% increase in the number of cases that have an MCC – more than one out of every four cases.

While CMI is a common metric across the industry, especially for CFOs, it’s influenced by a wide variety of factors, many of which are outside CDI’s control. This includes things like changing patient populations, shifts in med/surg volumes, changes to service line volumes and more.

Almost as many facilities experienced a lift in GMLOS as MCC capture (84%). GMLOS is a good indication as to whether or not a hospital is getting credited for how much effort it takes to care for a patient. If a patient is under documented, it will appear that they should have a short hospital stay, and then there can be a gap between expected LOS and actual LOS.

Financial Impact

Iodine utilizes two different methods for calculating financial impact, a “bottoms up” approach based on the value of a query, and a “top down” approach based on increase in MCC volumes.

The bottom’s up approach is calculated by examining: what are the number of queries issues by a hospital, how many of those queries are likely to have a financial impact, what was that impact measured in CMI MS-DRG relative weight points. Using this approach, a fictional hospital with 10,000 discharges, a 30/70 med/surg split and a $6,000 base rate would see an additional $2.4 million in appropriate reimburse.

Using the top down approach, looking across our entire cohort (which includes everything from hospitals that have been with Iodine for just the minimum two months to hospitals who have been an Iodine client for seven years) the average is $3.5 million in annual additional reimbursements based on improved MCC volumes.

Key Takeaways:

• Documentation integrity is the foundation of good quality rankings
• OhioHealth was able to expand their scope and implement new workflows even at historically low staffing levels due to efficiencies introduced by Iodine’s Concurrent
• Improving documentation accuracy and capturing all patient conditions led to improvements in SOI, ROM, and Vizient quality rankings, including OhioHealth jumping from 112th to 14th in the nation for Trauma Care

In the world of healthcare, hospital quality rankings hold immense weight, guiding consumer decisions on where to receive care, impacting hospitals’ reputations, and influencing the language healthcare providers can include in contracts with payors. Ultimately, the foundation for achieving good quality rankings lies in capturing accurate and complete documentation. Documenting patient care and medical procedures accurately and comprehensively directly impacts the assignment of codes, which, in turn, affects risk adjustment and external metrics.

As CDI programs shift their focus to quality, they must also shift their processes and workflows. Whereas in the past CDI specialists might have focused exclusively on capturing CC’s and MCC’s, it is now increasingly important to capture all conditions and treatments, regardless of whether they qualify as a CC or MCC, because many of these conditions ultimately impact risk adjustment. “It becomes important to know not only did it happen, but when did it happen?” said Fran Jurcak, Iodine’s Chief Clinical Strategist “Was it something the patient brought in with them, either from a previous encounter or something that happened to them in the outside world of the hospital? Or is it something that actually happened during the inpatient encounter? Which is what we’re really trying to capture with some of these quality metrics.”

Tonya Motsinger is the System Director of Clinical Documentation Integrity at OhioHealth, and her program has instituted a variety of workflows with the intent of completely capturing all patient conditions and improving their quality. She echoed Fran’s sentiment, relating that her CDI department found instances where HACs and PSIs were being driven by documentation, not the care being give. “If we could change the words in the chart, there wasn’t really a PSI or a HAC that happened” said Tonya, “It was just that it was documented incorrectly.” In today’s world, it’s not enough for health systems to provide good care, they also need to document the care they’re giving accurately.

OhioHealth is a nationally recognized, non-for-profit health system based out of central Ohio, composed of 14 hospitals and over 200 outpatient and physician offices, and has been recognized as one of the top five large health systems in America by IBM Watson Health six times. The CDI program at OhioHealth has 54 full-time positions, including a director, two managers, an educator, and an informaticist, although currently the program has positions open and is down ten staff. They originally invested in Iodine out of a desire to increase the efficiency in their workflow; they had found that they were performing a lot of re-work and weren’t seeing a ton of benefit from that work. Continuing, and improving upon, the success that they had seen in the financial and quality spaces was also key – around three to four years ago OhioHealth’s CDI program had ramped up their focus on quality, including instituting some new processes and workflows, and wanted to sustain that progress. OhioHealth felt that Iodine’s product Concurrent, with it’s artificial intelligence and prioritization, could help them achieve their goals without comprising their already established success.

As a quality focused program, it’s very important that Tonya’s team is not only accurately capturing the patients’ severity of illness and risk of mortality, but also ensuring the chart is explicit on what occurred with a patient so coders can accurately capture all diagnoses and treatments. As part of this effort, OhioHealth has a number of workflows in place, including an extensive second review process for expirations, DRG mismatches, and low acuity which can help ensure complete capture of how sick a patient truly was. “I believe that these practice are really essential growth avenues to drive the success of any CDI program, and I think Iodine really helped us to expand on these” said Tonya.

 We’re about ten CDS down…but we’ve been able to, even at our very lowest staffing ever in the history of our department, take on more work and meet the organization’s board goals as a result of the help that Iodine’s provided for us through the prioritization.

– Tonya Montsinger

While post implementation OhioHealth’s overall review rate has stayed the same, their query rate, especially among high priority cases, has increased. Despite spending time writing more queries, their workflow is so much more efficient they’ve been able to maintain their collaboration with coding, and took on even more work from a quality perspective, implementing additional workflows. This led to an improvement in their O:E ratio, their SOI and ROM, and their Vizient ranking. In 2019 OhioHealth was ranked 112th in the nation for trauma care, and they knew that wasn’t reflective of the care they were delivering. Today they are 14th in the nation, and number one in the state of Ohio.

Tonya noted that job satisfaction has improved as well. By enabling CDIS to review the right case at the right time and saving them from unnecessary review, OhioHealth has freed up staff for additional workflows like using the Vizient calculator, performing mortality reviews, and collaborating with coding and other departments. Describing the impact of the increased efficiency and effectiveness of their program due to having a prioritization tool, Tonya said, “I believe it creates a higher competence level in the CDS team, a little more of a strategic lens, because they can see the impact of their work beyond just case review, and they can sometimes even connect it with patient care.”

This expansion of skillset also gives CDI specialists opportunity for career advancement. As Fran noted, “We now have clinical ladders in the CDI space where five, seven years ago, there was no such thing.” In a world where hiring new staff can be challenging, retaining the staff you already have is key, and ensuring CDI specialists are satisfied and have room to grow within your organization can entice them to stay within your program.

Key Takeaways:

• Health systems need to move beyond cost cutting strategies to weather the current financial climate, they need new strategies for generating revenue, including relying on automation to scale scarce clinical resources. BJC Healthcare focuses on people, process, and technology: where can they automate where previously they relied on people to do the work manually
• Whereas previously labor was the primary driver of economic growth in healthcare, there is real opportunity for leveraging technology to capture additional opportunity. Mueller cautions healthcare orgs to ensure they have processes in place to support new technology implemented; and Damschroder pushes the importance of standardizing work and embedding new tools into the workflow.
• The onslaught of AI powered technologies on the market makes evaluating and selecting a tool for investment confusing. Both HFH and BJC leverage Project Management Teams to ensure they get returns on their investments

Iodine Intelligence tackles a new challenge in healthcare’s mid-revenue cycle every month, shedding light on problems and solutions and sharing valuable insights from industry experts. Listen to Episode 13 Opportunities and Obstacles: a CFO Conversation on Health System Financial Resiliency to learn more.

This month’s episode of Iodine Intelligence brings you a conversation we had in March, Iodine’s Chief Revenue Officer Troy Wasilefsky was joined by Henry Ford Health’s Executive VP and Chief Financial and Business Development Officer Robin Damschroder and BJC Healthcare’s VP of Revenue Management Harold Mueller to discuss the challenges healthcare systems are currently facing and strategies for building financial resiliency in the face of economic uncertainty.

Leveraging Technology to Capture Opportunity

The financial challenges facing healthcare providers today has led to an uptick in the use of consultants, particularly in the revenue cycle space, to try and stabilize health system economics. Overwhelmingly, the recommendation from these consultancies is cost reduction. CFOs are looking to slash budgets, with 75% stating they were planning on decreasing operating budgets as a cost savings measure.¹

However, cutting costs doesn’t generate revenue. McKinsey has mapped out impacts to profit pools if no new sources of revenue are introduced. While passing on higher costs to payors and patients lessens the impact, profit pools are eroded in all three scenarios. Hospitals and health systems need a new source of revenue, or at the very least to be effectively capturing the revenue they’ve already earned.   

This report from McKinsey is pushing the idea that cost reduction alone cannot be the answer here. We need to find a way to also generate growth and new revenue, either new revenue, or at least be capturing revenue appropriately for the work that we’re already doing.”

–Troy Wasilefsky, Chief Revenue Officer at Iodine Software

When asked how he approaches balancing efforts to manage costs with growth strategies, Mueller responded, “I think from the standpoint of things that we can’t control. So, if you think about the last 24 months, the price of some of the travel nurses…you wouldn’t have believed them five years ago, if you were to look at the balance sheet. That being said, from a revenue cycle standpoint, we are focused on the people, process, and technology. So, how do we automate things that can be automated, that are manual?” Currently BJC has a number of projects in flight looking to automate tasks which are currently done manually, and hopefully, these projects will enable BJC to ultimately redeploy staff to other areas.

Damschroder echoed these thoughts on implementing automation; noting that as the cost to procure a patient has gone up and staffing shortages persist, opening the “digital front door” is more important than ever. “Patients want to be seen faster than we can often get them through our process,” says Damschroder, “So the health system that can get that gate open in a market faster, and turns the faucet on, I think, wins there.”

Prior to the COVID-19 pandemic, Henry Ford Health decreased their revenue cycle costs while improving their yield by tipping their revenue cycle on its side and inserting automation to cluster homogeneous work (ex. grouping denials by demographics, regardless of payor), enabling staff to complete work faster and more efficiently. Henry Ford Health implemented Iodine with the same idea: increasing the productivity of the staff they currently do have. Damschroder noted that they have a lot of openings in the revenue cycle, and they’re holding out on those openings in the hopes that Iodine’s product will enable them to cover their current workload without the need for additional hires. 

Healthcare Providers Need a New Driver of Economic Growth

Research from the Bureau of Labor Statistics shows that 90% of the economic growth in the healthcare space comes from labor, with more than two-thirds of labor’s contribution coming from workforce expansion (4M net jobs were added) whereas other sectors’ growth was primarily driven by capital or innovation.² However, McKinsey would argue that there is a $1T opportunity in healthcare stemming from accelerating and scaling innovation in four key areas: care delivery transformation, administrative simplification, clinical productivity, and technology enablement.³ 

However, it’s not enough for healthcare providers to merely find and acquire the right technologies, they must also be mindful of how they’re implementing them. As Wasilefsky explained, “Innovation, technologies powered by AI and machine learning…they can bring a tremendous amount of opportunity to organizations, and yet, if you don’t use them, or build your processes around using them, and manage that change management, you’re not necessarily going to get all the results.”

Mueller was quick to point out that ensuring processes are in place to support the actual technology implementation is key. As an example, before 2019 the mid-revenue cycle at BJC was decentralized, many departments including CDI, HIM operations, and coding were not part of a shared service, and different hospitals deployed technology in different ways. BJC centralized services right after the start of the COVID-19 pandemic, and this enabled them to deploy Iodine, which BJC used as an opportunity to educate staff consistently, query physicians in a consistent manner, and gave them a tremendous lift in query volume. BJC facilities that deployed Iodine saw a 15%-20% lift in query rates, with physician response and agree rates staying the same. Although CDI specialists were reviewing the same volume of charts, Iodine’s prioritization ensured they reviewed the right charts at the right time, resulting in improved query volume.

“You want to make sure that your technology supports your actual administrative processes, and the processes that you have in place… From an Iodine standpoint, when we had facilities that we had centralized (and we centralized them in waves) we saw a 15%-20% increase in query rates, with physician response rates staying the same and agree rates staying the same. And these folks were reviewing the same volume of charts, but they were actually in the in the right charts, “

-Harold Mueller, VP of Revenue at BJC Healthcare

Damschroder touched on change management; while some may interpret “standardization” to mean “you don’t trust me to do my work,” she was quick to push back that standardization is not about lack of trust, but rather about elevating staff to their top of license and focusing them on where they can make the biggest impact. 

Damschroder also talked about the importance of ensuring new tools are embedded in the workflow and that the new workflow has buy-in. Henry Ford’s Health trick is robust monitoring and transparency: ensuring everyone can see how everybody is performing. “When you can barely find the other workflow or the workaround somewhere else, it’s been fully adopted.” said Damschroder, “And when new people come into the organization, they don’t even recognize that there was an old workflow out there.”

Realizing the Promise of AI

In fact, 60% of healthcare leaders report being confused by the range of automation and AI solutions.⁴ In the words of Wasilefsky, “AI is the new shiny bauble, and everybody uses that term, probably to an exhaustive level.” The confusion surrounding AI can make selecting an AI powered tool, and ensuring that investment will have a financial return, challenging. 

These feelings were reflected in the webinar attendees. In an Iodine survey of those attending the webinar, only 19% felt very confident in their ability to effectively evaluate and select the best AI tech needed to improve their financial performance, and only 28% of respondents felt they were getting quantifiable ROI from their current mid-revenue cycle solution. In fact, 80% of respondents felt they are missing out on earned revenue in the mid-rev cycle.

“This is something we hear in the market a lot. There’s an inherent skepticism of: have we seen the ROI prove out on some of these AI applications?” 

– Troy Wasilefsky, Chief Revenue Office at Iodine Software

The confusion surrounding AI is exacerbated by the glut of AI-powered solutions on the market. Between buzzword inflation, AI’s nebulous definition, and the vast range in AI technologies, their capabilities, and results, it can be difficult for healthcare leaders to truly wrap their arms around: what am I buying, what is it doing, and what outcomes can I truly expect? 

Mueller touched on the onslaught of AI-powered tech in the market, saying “Every vendor that we deal with that has a computer is “AI” now.” When it comes to evaluating AI powered solutions and measuring their impact, BJC has implemented Project Management teams, including some specifically in the revenue cycle space with revenue management experience, who evaluate business cases for new technology investments, and post implementation do a look-back to ensure they are seeing a return on their investment.

Damschroder echoed this, discussing HFH’s rigorous due diligence process regarding implementing new solutions. Regarding ensuring you see return on your investment, Damschroder emphasized the importance of ensuring adoption and that the new tool is embedded in the workflow. Whenever Henry Ford Health evaluates a new solution, they pay particular attention to: what is the lift to get this embedded in the workflow. For healthcare leaders out there who believe they’re not seeing the promised return on a technology investment, Damschroder’s advice was look at the workflow and ensure it’s embedded in the process, and then look at your adoption rates, because staff may not be using the tool or may only be using parts of it.

1.   Academy IQ, CFO Forum Debrief, December 2022

2.  The Productivity Imperative for Healthcare Delivering in the United States. McKinsey & Company. February, 2019.

3. Claiming the $1 Trillion Prize in US Health Care. McKinsey & Company, September, 2013.

4. The Academy Research and Analysis

Key Takeaways:

• There is enormous potential for leveraging AI in healthcare, including removing toil from staff workloads, increasing efficiency and productivity, improving consistency in results, and
• It can be difficult to truly evaluate AI powered solutions due to buzzword inflation, and the fact that AI is an umbrella term covering a wide range of technology
• Asking some foundational questions can be key to truly understanding if an AI powered solution aligns with your business need.

---

Artificial Intelligence (AI) has been an area of significant interest for the healthcare industry for years, and that interest is only growing in the face of prevailing financial headwinds and staffing shortages. However, in a crowded field of AI powered solutions promising to solve healthcare’s more pressing issues, it can be difficult to truly evaluate their capabilities and potential. In this month’s episode, Priti Shah, Iodine’s Chief Product and Technology officer, provides a framework for making sense of AI and its potential, and some of its applications in the mid-revenue cycle space.

The Promise of AI

The real world applications for leveraging AI to solve some of the pain points in healthcare can be sorted into a few broad buckets:

1. Automation: One of the most basic applications is simply automating tasks that you don’t actually need a human to perform.
2. Efficiency: If you have a task that can’t be automated entirely, it still requires human judgement, you can still supplement and enhance your staff with AI-powered tools.
3. Timeliness: You can ask an AI model to actively evaluate your entire patient census 24/7, which, realistically, you will never be able to staff humans to that degree.
4. Consistency: An AI model can help establish a baseline level of competence, rather than relying on strengths and weaknesses of individual staff with different experiences and clinical background.

AI is Confusing

Although recognizing the promise of AI is easy, evaluating an AI powered solution can become confusing quickly, because the truth is not all AI is the same, and not all AI can solve all of healthcare’s unique problems. 

Today, there are two main barriers to understanding and evaluating AI. The first is that AI is just a hot space right now and the term gets bandied around a lot. Everyone is trying to claim that they use some form of AI, and they all mean something slightly different when they make that claim. The second problem is even within the field of computer science AI is not well defined, it’s an umbrella term that covers a lot of different tools and technologies for solving a lot of different types of problems. The only real common theme is: applying computer systems to perform tasks that normally require human intelligence because they’re too hard or complicated for computers.

AI models are constantly walking that tightrope, balancing precision and sensitivity, and you can’t over-pivot on either axis because it essentially renders that model impractical to use. But that also means we have to understand no model is perfect, and you have to chose what balance of false negatives and false positives you can live with.
” AI’s accuracy, or success, is always on a spectrum, and there’s always tradeoffs that I think we should be aware of.

PRITI SHAH, CHIEF PRODUCT AND TECHNOLOGY OFFICER

De-Mystifying AI

Knowing the benefits that AI can bring you, but acknowledging the challenges of truly understanding and evaluating AI, Priti Shah offers the following framework for thinking about AI, with some basic, fundamental questions to have in the back of your mind when thinking about making an investment in an AI powered tool.

Use Case

The first question to ask yourself is: what problem are you trying to solve?

This is important for two reasons, one, there’s a wide range of AI tools and they’re not all equally good at all tasks, and two, almost every AI model you’re going to encounter right now is trained to do one very specific task. 

There are different tools available in the AI space that are better suited for solving some types of problems than others. 

• Documentation Interpretation: NLP or large language models
• Classification Problems: Gradient boosted machines or neural networks
• Image Recognition: Deep learning models

Different tools have different applicability to different problems, so you shouldn’t just focus on the latest and greatest. ChatGPT is currently the talk of the own, and while there are some things it does very well, its limitations have been well demonstrated. While powerful in its domain, no one’s trusting it to make a clinical diagnosis.

There is no silver bullet, you should not expect any one technology to solve all your problems, so when selecting an AI for investment, focus on: what is this AI actually trying to do for me, and does that match up with the business problem that I am trying to solve.

Performance

The second question to ask yourself is: how well does the model actually perform?

You can’t assume that just because a solution is powered by AI, the AI is highly successful. While some may think, “If the AI is trained to do this, it’s doing it perfectly,” that’s rarely the case. The reality is AI’s success is always on a spectrum, and generally that comes with trade-offs that you need to be aware of.

Artificial intelligence models are always balancing sensitivity and precision. You can create a model that’s so precise it has no false positives, but your criteria will be so narrow you’ll miss most of what you’re searching for, and have a ton of false negatives. Conversely, you can create a model that is incredibly sensitive, but by casting a wider net you’ll also catch a bunch of false positives.

Dialing up either precision or sensitivity too high can result in a model that is essentially useless for practical purposes, so you have to choose what balance of false positives and false negatives you can live with.

Timing

The third question to ask yourself when evaluating AI is: when is it capable of making its predictions. 

An example of this coming into play would be giving your sepsis coordinators an AI model that can predict sepsis. If the model is only capable of making its predictions post-discharge, it doesn’t actually fit the use case of your sepsis coordinators, who want to identify sepsis within 24 hours of a patient’s stay. Timeframe is critical when making a decision about deploying AI.

Explainability

And the final thing you should consider is: how much insight can the model actually give you into why it’s making its predictions. 

There are use cases where all you care about is the answer and how confident the model is in the answer. But when it comes to healthcare, where you’re interacting with other people, and you’re dealing with someone’s health information, it cannot be a black box. You have to be able to discuss and explain why is the AI making this determination. 

Especially because, consider what we discussed earlier: that no AI is perfect, that it’s always a balance of sensitivity and precision, and therefore also a balance of false positives and false negatives. You need clear explanations of the predictions so you can know when to trust the system’s predictions, and when to apply your own judgment.

Iodine and AI

To further help conceptualize this, below are example of an AI company answering these four fundamental questions.

• What problem are we trying to solve
  • At Iodine, we leverage machine learning models to take in raw clinical data (lab results, ordered medications, performed treatments, clinician observations, etc.) and look at all those disparate pieces of data in conjunction with one another to make predictions about various disease conditions. We leverage those clinical predictions in a variety of ways. In the CDI space, we compare the clinical reality of the patient against what’s documented, and then look for gaps in between that can be clarified. So when it’s time to bill and code, the documentation is complete and accurate, and health systems will get paid. In the utilization management space, we compare the clinical reality of the patient to the level of care we would expect a patient like that to receive, to aid UM nurses with determining the appropriate level of care and admission status for patients
• How well do our models perform
  • Iodine is fortunate enough to have access to about a quarter of all inpatient visits in the US, and this enormous set of clinical data is fueling our models. Having more data enables the ability to target more rare diseases. We’ve also been iterating, and experimenting, and improving on our models for seven years. Data science is a process of discovery; for some of the more complicated disease states, we have gone through seven or eight different generations, each new version building upon previous advancements to increase performance. Across our client cohort, we’ve found that we’ve been able to substantially improve client’s metrics: 92% of facilities saw an increase in productivity (query volume) with the average facility generating more than twice as many queries per CDS as they did before Iodine. With our models surfacing those cases with the greatest likelihood of opportunity, CDI specialists were two-thirds more likely to query a reviewed patient. And this has real, measurable impact, including increased MCC capture rates, increased CMI, and, on average, an additional $3.5M in annual appropriate reimbursements per 10k admission.  
• When are our models making their predictions
  • Our models are working concurrent with the patient stay, and our models are constantly reevaluating in real time as new information becomes available
• How much insight do we give into why the model made its predictions. 
  • We bubble up the most relevant clinical evidence to our users so they can see why we think this way about a patient. We’re not speaking about a patient in the abstract, or this general type of patient, it’s specifically: why does the model think this patient, Jane Smith, has sepsis based on the way she’s presented so far. 

Hopefully this framework is helpful for evaluating AI solutions, and having these four fundamental questions in the back of your mind will help to demystify the hype around AI, explain the different types of AI, and why it should be something you’re considering, but considering for the right reasons. 

Key Takeaways:

• ChatGPT is a free online chat bot that creates custom content at users requests, debugging computer code, composing songs, writing student essays, and more
• While some tout ChatGPT as the next solution for the CDI space, it has a long way to go before it will have real utility to clinical documentation integrity
• ChatGPT is based on Natural Language Processing (NLP) which is not well suited to interpreting and analyzing data to determine what’s missing
• In its current iteration, ChatGPT isn’t capable of producing documentation with enough detail or specificity to replace physician written progress notes or CDS composed queries

---

Iodine Intelligence tackles a new challenge in healthcare’s mid-revenue cycle every month, shedding light on problems and solutions and sharing valuable insights from industry experts. Listen to Episode 11 – Beyond Buzzwords: Understanding the Real Potential of ChatGPT in CDI to learn more.

ChatGPT is currently a free online program that works like a chat bot: you type in questions or requests and it spits back answers. Although it’s designed to mimic human conversation, users have found it to be incredibly versatile, using it to write and debug computer codes, compose poems and songs, write student essays, and even take the bar exam.

While ChatGPT is impressive, and certainly fun, it has a long way to go before it’s ready to revolutionize the clinical documentation space.

ChatGPT is founded on Natural Language Processing (NLP), which is already deployed widely in the CDI space. And while NLP can be a powerful tool for taking written text and interpreting it into a form a computer can understand, it is not well positioned to help CDI with their chief concern, namely: what is missing from the documentation. NLP isn’t capable of looking at information and applying any kind of logic or understanding to the underlying meaning. A medical record can speak very specifically about signs and symptoms of a medical condition, but if that medical condition isn’t articulated in the documentation, NLP isn’t capable of finding it. ChatGPT has the same limitation: it’s not trained to identify what the documentation should be. A very different type of AI is needed to interpret and understand documentation.

Additionally, some users are encountering instances where ChatGPT has falsified its answers. Users will ask ChatGPT a question, and get an answer that sounds very scientific and even includes citations, but upon digging deeper, users will discover the answer provided was fake and ChatGPT invented the citations. Additionally, ChatGPT has revealed some concerning biases in the data it’s been trained on, for example answering “describe a good scientist” with “a white man.” Both of these issues can have serious consequences when applied to people’s health and medical records.

Perhaps the largest barrier to ChatGPT in the CDI space is it doesn’t get to the root of the problem – mimicking the way a physician writes progress notes is likely to produce the same gaps we find in documentation now, and with the physician a further step removed there’s the potential for the documentation to be even less accurate – because the physician never touched it. In it’s current iteration, if you ask ChatGPT to write you a progress note or a query, it’s not capable of producing the level of detail necessary because it doesn’t have the right data or learning as a foundation. It would require a huge clinical data set, composed of millions of medical records to effectively emulate a clinician’s brain when creating documentation.

Further, even if ChatGPT was capable of writing a progress note or a query, if its outputs are just large blocks of text, that has limited utility in a modern healthcare setting where data must flow downstream to multiple programs and softwares. Additionally, no matter what ChatGPT requires the author to review its outputs and ensure what was translated into the document is accurate – even with AI you can still end up with typos or words and phrasing not appropriate or accurate for the situation. (Just ask anyone who’s blindly relied on auto-correct in a text message, or auto-completion in an email).

Some better uses for ChatGPT in healthcare might be handling scheduling and frequently asked questions from patients, helping doctors and patients with translations, and writing or drafting emails, including for fighting insurance denials. Use cases in the outpatient setting might also have a lower hurdle to clear, as patient encounters in outpatient are much shorter, so the details needed to be included are more streamlined and limited to this specific encounter, rather than the overall picture of the patient. Even helping with the day-to-day communications between providers, coders, CDI nurses, can be helpful and a time-saver, but as it stands today, at best ChatGPT is getting you a first draft, a human is still required to refine and finalize.

Key Takeaways:

• Concurrent can be a disruptive technology requiring some key shifts in thinking and processes to unlock its full potential
• Auto-assignment allows for complex cases to be distributed evenly across a team and for better coverage when there’s a gap in staffing
• When setting quiet periods, shoot for a time range that’s appropriate 80% of the time
• When artificial intelligence and machine learning are leveraged to provide a prioritized list of cases for review, priority review rate is a more important metric than number of initial reviews and re-reviews
• In healthcare’s ever changing landscape, it’s key to periodically reassess your configurations to ensure they’re still best fit

---

Iodine Intelligence tackles a new challenge in healthcare’s mid-revenue cycle every month, shedding light on problems and solutions and sharing valuable insights from industry experts. Listen to Episode 10: Unlocking the Power of Concurrent to learn more.

Concurrent is Iodine’s flagship software, and it leverages artificial intelligence and machine learning to prioritizes cases for CDI teams to review concurrently before patients are discharged based on misalignment between the clinical evidence and existing documentation.

In this month’s episode we sat down with Iodine’s Client Services Operations Manager, Justin Gerardot, and Iodine’s Clinical Product Consultant Manager, Diana O’Connor, for some tips and tricks for our users to get the most out of Concurrent. Concurrent, and it’s prioritization, can be referred to as a “disruptive” technology, and there are some key shifts in thinking and processes required to unlocking the full potential and return of Concurrent.

1. Auto-Assignment

Concurrent distributes cases to CDS’s for review using auto-assignment, which can be a shift for CDI teams who are used to cases being distributed by location or service line.

Iodine advocates for a “generalist” rather than a “specialist” approach. Some service lines are, by their nature, more complex than others, and auto-assignment allows for those complex cases to be distributed evenly across a team. It also helps when a CDS is on vacation or there’s a gap in coverage: everyone is able to cover and review those cases.

2. Quiet Periods

A Quiet Period is the amount of time a case must “incubate” before it can be considered for auto-assignment. It can be a delicate balance walking the line between a quiet period that is too short, and CDS’s receive cases for review with very little information, and a quiet period that is too long, and the patient is discharged before the case gets a chance to be prioritized and reviewed.

When selecting a quiet period, Iodine recommends shooting for the 80/20 rule: 80% of the cases are prioritized at the appropriate time or within the appropriate window.

3. Metrics

Concurrent provides teams with an intuitive, prioritized work-list in order to get the right cases in front of CDS at the right time, and because of this, Iodine is not as concerned with how many initial reviews and re-reviews a CDS does on a given day or during a given week. Instead, Iodine focuses on priority review rate: how many high priority cases did you get to.

It can be difficult moving away from longstanding, traditional KPIs, especially when they’re used for projections and staffing needs. Iodine has a robust reporting platform to support these new metrics, allowing CDI teams to easily track their progress.

4. Reassess

Healthcare is an ever changing field, and as a result it can be very helpful for hospitals and healthcare providers to reassess their Iodine configurations on at least an annual basis. Changes to the size of a CDI team, adding additional service lines or payors, and software updates within Concurrent itself can all impact the way CDS interact with Concurrent. Reevaluating your configurations, especially once you have six month’s to a year’s worth data to help with your evaluation, can help ensure you’re staying at top functionality.