Interview at AHIMA 2020: Reducing Revenue Leakage with Cognitive Emulation

Iodine Software was interviewed at AHIMA 2020 on how HIM leaders can leverage AI and machine learning to reduce revenue cycle leakage. Iodine has pioneered a new machine learning approach called Cognitive Emulation™, and most recently launched the AwareCDI™ Suite. Listen to a recording of the interview here and read the full excerpt below. 

AHIMA: Can you talk about the problems that Iodine is seeing when it comes to mid-revenue cycle leakage?

IODINE SOFTWARE: When it comes to the mid-revenue cycle, it’s critical that the full clinical picture as reflected in the evidence is correctly, accurately, and with detail documented, and then fully represented in the code. Unfortunately, this can cause problems due to the fact that humans are involved at every step, and that the underlying legacy software is focused only on workflows that aren’t holistically solving any of these problems. 

For example:

  1. There aren’t enough CDI personnel to review every case every day, which is necessary to ensure documentation integrity. 
  2. Even when pointed to and reviewing the right case, there’s a substantial loss of integrity at the point of decision to query. 
  3. When the query is written, there are fall offs both in physician response and agree rates. 
  4. And finally, there’s further loss of integrity at the coding step due to lack of clinical competency, poor communication, and failure to cross-connect evidence / documentation/code.

What this results in is lost “earned revenue”, which can significantly impact organizations. 

AHIMA: Could you help us better understand the magnitude of this leakage? 

IODINE SOFTWARE: Prior to the start of COVID-19, health systems were already operating on generally thin margins, with many finance leaders acknowledging that a significant root cause was leakage from their mid-revenue cycle and that “average performance” was still well below optimal results. For the average 250-bed hospital, that is $4.7-11M1 in revenue each year

Today, the world is different. Complacency has been fast replaced by a new urgency, and the traditional approach to solving this problem — hiring more staff — is no longer feasible as highly trained and specialized staff to do clinical documentation are in short supply.

We can no longer afford to effectively ‘earn dollars’ only then not to realize them, solely because of unintentional, clerical and clinical human error in documentation and coding. Failing to get this right could mean the difference between positive or negative operating margin, which impacts our real mission – delivering the highest quality clinical care, sustainably.

With this new normal as our backdrop, finance leaders are looking at how to best leverage technology to do things differently – now – and ensure their organizations are financially resilient for the next decade and beyond.

AHIMA: Can you tell me about Iodine’s Cognitive Emulation approach, and what makes it different from others on the market? 

IODINE SOFTWARE: Today, most healthcare technology solutions that support revenue cycle billing, coding and documentation teams use systems and workflows that “think” like computers – not clinicians. They leverage rules and check-lists, which only consider narrative documentation and can lead to unforeseen errors given the many nuances of the healthcare revenue cycle. 

At Iodine, we take a different approach. Cognitive Emulation applies physician-like judgment to the clinical evidence in a patient’s chart and leverages previous learnings to more accurately determine the likelihood a condition exists. Conditions often present in a variety of ways, and by relying on clinical evidence rather than ambiguous thresholds, Iodine is able to identify and learn from these unique instances.

We’re the only organization with the capability of quantifying the magnitude of this problem with precision. And now, we’re the organization uniquely equipped to address it. For each of the leakage points that I talked about earlier, we’ve built and deployed software modules, with each one emulating clinical judgement to solve this earned revenue leakage problem. All these components seamlessly integrate in a unified suite that we call AwareCDI, and powered by our core AI/machine learning technology, Cognitive Emulation.

AHIMA: How could an HIM leader leverage the AwareCDI Suite?

IODINE SOFTWARE: One of our newest products, and an example of how we apply Cognitive Emulation to the mid-revenue cycle, is Retrospect. Retrospective reviews are often the last opportunity to resolve documentation and coding issues prior to final submission of codes for billing and quality reporting purposes. With up to 25% of post-discharge reviews resulting in meaningful education opportunities or code changes that can lead to revenue impact, this final inspection is business-critical. However, this would require the review of every single discharged record to ensure full integrity of each and every outgoing code—which is impossible to do without technology.

At Iodine, we ease the burden on CDI and coding teams by automatically reviewing every record prior to billing. Retrospect provides reconcilers with clear and actionable information to review the right cases at the right time, calling out specific opportunities to clarify documentation and/or final codes in order to improve review confidence and query quality.

We have several clients that are currently utilizing the first version of Retrospect, and the results are pretty amazing.  What we are seeing in our early adopters is that about 30% of cases reviewed in Retrospect resulted in coding changes that impacted the final DRG. Through the use of our CognitiveML engine and prioritization, we were able to support a post discharge workflow that impacted final codes in greater than 60% of cases reviewed.  

To learn more about Iodine and the AwareCDI Suite, click here

¹ 2016 ACDIS Advisory Board Study 

The Truth About Organizational Definitions

By Cheryl Ericson, MS, RN, CCDS, CDIP

Let me start by saying I’m a proponent of organizational definitions. I have long advocated them as a Clinical Documentation Integrity (CDI) best practice, but I do think their purpose is frequently misunderstood. Although we like to think organizational definitions are a tool to minimize denials, they are really just an organization-wide strategy to promote consistency and have little to no bearing outside your organization.   

Contrary to popular beliefs, Centers for Medicare & Medicaid Services (CMS) does not “define” conditions like sepsis or malnutrition or morbid obesity. CMS provides guidance around when a particular condition is considered medically necessary so it will be covered by Medicare through National Coverage Determinations (NCDs) (e.g., gastric bypass defining morbid obesity as a BMI of 35 with the presence of complications due to morbid obesity) or Local Coverage Determinations (LCDs). But these NCDs and LCDs do not necessarily “define” the referenced conditions when it comes to publicly reported data. The same is true for quality measures adopted by CMS.  

The CMS Quality Measure titled “Severe Sepsis and Septic Shock: Management Bundle” supports best practice for the treatment of severe sepsis and septic shock which includes processes associated with Sequential Organ Failure Assessment (SOFA); however, the population eligible for this measure is defined by the assignment of either a sepsis or severe sepsis code. If you need evidence of a lack of CMS guidance defining a particular condition, look no further than the recent Office of the Inspector General (OIG) findings related to severe malnutrition where the OIG audited cases to “determine whether providers are complying with Medicare billing requirements when assigning diagnosis codes for the treatment of severe types of malnutrition on inpatient hospital claims” (https://oig.hhs.gov/reports-and-publications/workplan/summary/wp-summary-0000258.asp).  It is even hard to pin down commercial payers other than CMS when it comes to defining conditions. Often organizations receive information informing them a diagnosis was removed with little, if any explanation, of the criteria used to make that determination.  

So why bother with organizational definitions? To promote consistency across physicians, CDI professionals and Coding professionals. Often organizational definitions are a great way to engage physician leadership so they can become CDI advocates and to train CDIs. The reality is that making a diagnosis is a complex process and there is often disagreement across providers treating the same patient. Something else to consider is whether or not organizational definitions are too stringent and promote under-coding within an organization which can negatively impact financial goals. Just think about the debates that are occurring within the CDI profession over the use of Systemic Inflammatory Response Syndrome (SIRS) criteria vs. SOFA criteria for sepsis. Many organizations made the shift to SOFA criteria when it was first released only to return to SIRS criteria after the volume of sepsis cases decreased within their organization due to the stricter criteria. And what about those diagnoses that don’t have an organizational definition? What criteria should be used? Are we getting too bogged down in discussions about how to define a condition that we can’t see the forest from the trees? In fact, the whole concept of clinical criteria was such an issue within the CDI and Coding professions that the Official Coding Guidelines added the Coding Assignment and Clinical Criteria guideline a couple of years ago:  

“The assignment of a diagnosis code is based on the provider’s diagnostic statement that the condition exists. The provider’s statement that the patient has a particular condition is sufficient. Code assignment is not based on clinical criteria used by the provider to establish the diagnosis.” (ICD-10-CM Official Guidelines for Coding and Reporting (FY 2021), Page 12 of 126).

Although this guideline was intended to provide clarity, I really think it just added another layer of confusion as some organizations mistook this advice as an excuse to stop clinically validating documented diagnoses. Really, this guideline only separated the coding function from the medical necessity function because those lines were getting blurred; however, the medical necessity requirement is still alive and well as demonstrated by OIG audits.  

It was once believed that organizational definitions could be helpful from a compliance standpoint, but that only occurs when the definitions are consistently used across all payers and for all situations. Unfortunately, what I’ve seen over the years is that CDI professionals often have one rigorous set of criteria they use before querying for a potentially missing diagnosis and a different threshold for clinical validation e.g., if the provider documents a diagnosis based on limited criteria. I get it, most CFOs view CDI departments through a financial lens and few CDI departments want to be responsible for removing a diagnosis complication/comorbidity (CC) or major complication/comorbidity (MCC), but this inconsistency is confusing and increases compliance risk.  

So that begs the question, “Have organizational definitions outlived their usefulness?” Perhaps it is time to re-evaluate the purpose of organizational definitions. Are they resulting in more harm than good? In particular, perform an analysis to determine if organizational definitions are yielding the desired impact, which is typically fewer denials. In addition, determine if organizational definitions have actually become a liability due to inconsistent application, or potentially leading to under-coding of valid diagnoses. Organizational definitions can be a great educational tool for CDI professionals, Coding professionals and providers, but they often have limited practical application outside of your organization, so be sure to evaluate their usefulness on a regular basis.  

Webinar Recording: The Documentation Improvement Process: Where and Why Leakage Happens

Each stage of the CDI documentation integrity process represents an opportunity for additional leakage of accurate and appropriate documentation, resulting in inaccurate coding of conditions being monitored and treated during the patient’s encounter. And despite massive investments in documentation and coding solutions, earned revenue loss continues to persist — to the tune of $5-11M in leakage for an average 250-bed hospital1.

Leakage occurs due to a combination of factors including both human issues and technological misalignment, and occurs even in the most mature and highest performing health systems in the country.  Minimizing this leakage is key to the success of any CDI program and top of mind for CDI and coding leaders.

View this on demand webcast in partnership with ACDIS featuring health system leaders from Iodine Software and The University of Kansas Health System for a discussion on:

  • The four distinct phases where leakage occurs (from finding the case, to getting it coded)
  • The challenges to managing a more accurate mid-cycle
  • Tactical steps and solutions to address leakage

Download recording.

12016 ACDIS Advisory Board Study

Webinar Recording: Reducing Payer Denials with Clinical Documentation in the Mid-Cycle

Determining the right fiscal initiatives are at the forefront of every healthcare finance leader’s mind due to COVID-19 and its impact on margins – from furloughs, to reducing capital expenditure and focusing on debt collection. But margins were thin even before COVID-19 due to leakage in the mid-revenue cycle, with a 250-bed hospital losing up to $11M annually (2016 Advisory Board Study). Now with payer denials up 25% due to the CARES Act (XIFIN claim data), hospital finance leaders need to think beyond short-gain fiscal initiatives to recover lost revenue.

View this on demand webcast in partnership with RevCycleIntelligence featuring health system leaders from Iodine Software and Integris Health for a discussion on:

  • Why mid-revenue cycle management is a key concern today
  • Strategies for reducing payer denials with clinical documentation
  • New technologies that focus on the mid-revenue cycle to fully capture earned reimbursement and drive quality

Download recording.

Why You Need to Start Looking at Your Mid-Revenue Cycle for Lost Revenue

Originally published on Becker’s Hospital Review: https://www.beckershospitalreview.com/finance/why-you-need-to-start-looking-at-your-mid-revenue-cycle-for-lost-revenue.html

By: William Chan, CEO, Iodine Software

84% of healthcare leaders believe the root cause of lost or decreased revenue is inaccurate clinical documentation and coding.* But the challenge of managing a more accurate mid-cycle is that this accuracy is fundamentally driven by ensuring that the patient’s full clinical picture (as reflected in the evidence) is correctly documented in detail, then fully represented in the codes.

And despite massive investments in documentation/coding programs, earned revenue loss continues to persist — to the tune of $5-11M in leakage for an average 250-bed hospital1

Where and Why Leakage Happens 

The overall leakage problem is a combination of the fact that humans are involved at every step, and that most software is focused only on workflows and not solving any issues that may arise along a given workflow. Why leakage happens: 

  1. There are not enough clinical documentation integrity (CDI) resources to review every case, every day, which is necessary to ensure documentation integrity. Being able to identify discrepancies between clinical evidence and documentation is the first step in minimizing mid-cycle leakage.
  2. Even when CDI teams are pointed to and reviewing the right cases, there’s a substantial loss of integrity at the point of decision to query. The reason for this is multifold — a lack of competence or confidence, a worry about physician response, or a concern about potential for impact. Regardless, the result is that CDI specialists are frequently deciding not to query even when there is a clear inconsistency between evidence and documentation.
  3. When the query is written, there are fall-offs both in physician response and agreement rates far in excess of what would be expected given the confidence in the root evidence. The causes here are again numerous: poor relationships between physician and hospital, poorly constructed or supported queries, lack of understanding in the importance and impact of better documentation, and a lack of ease of use in reading, interpreting and responding to queries can all drive down integrity at these steps.
  4. Finally, there’s an alarming loss of integrity at the coding step. Lack of clinical competency, poor interaction and communication with CDI, and failure to cross-connect the implications of evidence, documentation, and code can all be drivers of lost opportunity at this step.

How You Could Fix This — If Money and Resources Were Unlimited

With unlimited resources, it would likely be quite easy. For example: 

  • You would provide continuous education and resources to ensure physicians were doing their best to document at the point of entry.
  • Then you would review every single case, fully, every single day, during the patient’s stay with only the most qualified, tenured and experienced CDI specialists in the market.
  • Then, you’d find the easiest, most seamless way to transmit queries to the physicians in a way that worked in their normal workflows and was a minimal burden — perhaps even a pleasure — to drive quick responses. And you would query everything.
  • And then, finally, once everything was coded, you would look at every single case again— and not just for DRG mismatches, but for any documentation opportunity lost…and you would again use only the best retrospective CDI specialists with experience and strong tenure.

However, no health system has the resources to deploy these four strategies, and most legacy software solutions are not even capable of finding many financial and quality accuracy improvement opportunities because: 1) they cannot determine when patient information that is supported by medical data is not written in a patient’s chart, and 2) they cannot perform clinical validation, in which clinical evidence does not support something that has been documented, which increases the risk of audit.  

Machine Learning: A New Way to Address Mid-cycle Leakage

As Iodine started further examining ways to stop mid-cycle leakage, we realized that current solutions and technologies did not solve the problem. The objective of CDI is to determine whether the written documentation aligns with a patient’s clinical reality, and this requires more than tools with natural language processing (NLP) alone. Instead, this is where machine learning comes in. Machine learning is capable of considering the entire clinical picture and can make connections and predictions based on learnings from other data including lab results, vital signs, medications, radiology results, and other sources.

Iodine has built proprietary artificial intelligence technology and machine learning algorithms that “think” the way a clinician thinks and emulates clinical judgement. We call this approach Cognitive Emulation

The result of combining NLP and machine learning technology is a solution that evaluates complex medical data similar to a physician’s approach to diagnosing and treating patients.

Iodine applies physician-like assessment to the clinical evidence in a patient’s chart and leverages previous learnings to more accurately determine the likelihood a condition exists.

Results Across 500+ Hospitals 

We started with Concurrent™ — our first product, designed to increase query rates by prioritizing records that contain inconsistencies between clinical evidence and documentation — all without the inherent limitations of rules-, marker- and NLP-based approaches.

Our customers’ results with Concurrent have been impressive to date: 

  • $1.5 Bn additional appropriate reimbursement recognized annually²
  • $2.55MM additional incremental revenue captured on average (per 10,000 admissions, $6,000 average base rate and 70%/30% med/surg split)² 
  • 86% of clients experienced a growth in query volume² 
  • 21% increase in MCC capture volume² 

We recently released the AwareCDI Suite to identify and capture additional mid-revenue cycle leakage, beyond the Review and Query stages, all the way through to final billing. You can learn more about AwareCDI here

*HIMSS and Besler Revenue Cycle Management Research Report- Insights into Revenue Cycle Management October 20

1  2016 ACDIS Advisory Board Study 

2 Figures are based on a $6000 modeled base rate and actual measured MCC capture performance from the 2019 Iodine Performance Cohort Analysis of 339 facilities that compared measured MCC capture and CMI impact for the Iodine usage period 9/1/2018-8/31/2019 against pre-Iodine baseline performance.

Finding All Queries — not Just the “Right” Query

By: Fran Jurcak, MSN, RN, CCDS, CCDS-O, Chief Clinical Strategist

The goal of a quality Clinical Documentation Integrity program is to support documentation that identifies ALL of the conditions being monitored and treated during a patient encounter. Clear and simple. Yet many CDI Specialists (CDIS) in our profession spend significant time “shopping” for queries that only impact financial or quality outcomes (SOI/ROM). Not only does this waste significant and valuable CDI time, but the true goal of CD work is to ensure documentation integrity of the complete medical record.

If the clinical evidence supports a condition that is not clearly and consistently documented, best practice would dictate that communication with the provider should occur to ensure accuracy of the medical record. Period. For documentation integrity, the impact of the intervention shouldn’t matter. The need to accurately capture documentation of all the conditions that were assessed, evaluated and cared for in the medical record is essential to accurate coding of the care provided and results in accurate payment and reporting of quality metrics.  

CDI Specialists spend countless precious minutes searching for queries with impact. Thinking that the only way to engage providers is to only “bother” them with queries that matter is short sighted and misses the bigger picture of the value of documentation integrity. In a world with diminishing resources and where integrity of the medical record is so important to capturing true acuity of every patients’ condition, every query opportunity to support documentation integrity is vital.  

If the documentation does not reflect the clinical evidence in the medical record, any and all queries should be communicated with the provider. Picking and choosing the “right” query creates inconsistency that is not only confusing to providers but allows for inaccurate coding and reporting, which can lead to incorrect reimbursement and poor performance in quality metrics.   

Let’s also talk about the time wasted on searching for the right query.  Knowing that resources are limited and time is short (average length of stay is typically below 5 days) there isn’t time to spare for a search-and-seek mentality. CDI Specialists need to stay focused on the job at hand and not spend upwards of twenty minutes hunting for the query that drives a particular metric. All conditions being cared for and monitored should be appropriately documented in the medical record so they can be accurately coded and reported. 

During the current COVID-19 pandemic, we discovered that across our clients, query rates on COVID-19 patients are 33-42% higher than on non-COVID-19 patients. It’s not likely that provider documentation is materially worse for these patients than others. Rather, this seems to point to the fact that due to the perceived need for greater scrutiny on these records to capture all appropriate comorbid conditions for accurate reimbursement and reporting, CDI staff are actually querying for every co-morbid condition. Why just these patient records and not all patient records?  

Many CDI professionals spend additional time concurrently coding the record thinking this will assist them in identifying a query opportunity. While there may be other good reasons to concurrently code a record, doing so to specifically identify impactful query opportunities is often wasted effort as the impact of a query may change due to additional documentation that results later in the stay.  For example, consider a CDIS that spends time searching for a condition that will impact the DRG, say a Major Comorbid Condition (MCC).   So, the CDS searches for a condition that qualifies as an MCC and queries that condition based upon the clinical evidence but does not query other conditions that are also clinically supported but do not qualify as MCC.  Then, later in the stay an additional MCC condition becomes accurately documented, which changes the impact of the query.  The end result is wasted time concurrently coding, often leading to missed query opportunities on other conditions and subsequent negative impact to  accurate documentation, coding and reporting.  If a condition is being monitored and/or treated and not clearly and consistently documented, the documentation should be clarified regardless of the potential impact.

It’s important for CDI Specialists to  utilize their clinical expertise and judgment to determine if there are documentation integrity concerns and communicate with providers to resolve those concerns. While basic knowledge of code language and coding guidelines is important to assist in accurately capturing documentation, CDI professionals should be assessing the clinical evidence in the medical record to identify missed or inaccurate documentation of conditions to support the integrity of medical record documentation. How the condition final codes should be driven by the documentation and completed by professional coders.

So the catch line is this:  While it is important for CDI Specialists to understand coding language and be able to identify appropriate codes for conditions being monitored and treated, CDI Specialists need to focus their attention on supporting documentation integrity and query all documentation concerns, not just those that have financial impact. 

Let’s focus on documentation integrity in every record, not just where we can measure impact.

 

 

Do I Really Need to Query for that Third (or Fourth or Fifth) MCC?

Written By: Rachel Mack, RN, MSN, CCDS, CDIP, CCS
Clinical Program Manager

We all tend to have a number of “firsts” we never forget when it comes to our CDI careers: 

  • First positive interaction with a physician
  • First educational session with a physician group – and it goes well
  • First coder interaction where you have a light-bulb Coding Clinic moment 
  • First time you felt as though you truly impacted patient care

My first example of impacting patient care is when I saw a patient with a slew of clinical indicators for  malnutrition diagnosis (including significant weight loss, decreased PO intake, a pressure ulcer, and a BMI of 12). I did not yet see an order for a dietitian evaluation, and the patient had been in the hospital for several days. I took a risk and sent the query anyway; the next day the physician I sent the query to put in an RD order. That dietitian went on to document that the patient met criteria for severe protein-calorie malnutrition. I thought, “wow, I helped do that.”

But another instance I’ll never forget is opening a denial email from Coding and seeing that an insurance company was denying all three MCCs for one case. 

….three MCCs. No wonder the Coding department had reached out for CDI help!

It’s commonplace for insurance companies to deny whatever they can, whether clinically sound or not. But all three MCCs? I had to dive in and take a look. After review, I discovered that they were partially correct: one diagnosis was incorrectly coded (acute respiratory failure final coded, but was part of a previous visit only), and in our defense letter I acknowledged that yes, we should remove it from the final bill. However, the other two diagnoses – metabolic encephalopathy and septic shock – were above and beyond clinically sound. The patient was confused with a GCS score of 11 that improved to 15 by discharge and was treated with a head CT, sitter, and safety restraints. And the patient was clearly in septic shock and required Levophed s/p aggressive fluid resuscitation to improve their blood pressure. 

So what can CDI specialists do to make sure records are as safe as possible from denials? 

Here at Iodine Software we have a few best-practice suggestions:

  1. Query consistently for clarification of conditions when the clinical indicators are present but there is no associated documentation (or vice versa when it comes to clinical validation).  When we do this consistently, we help teach our providers to consistently document with a higher level of specificity. 
  2. Query consistently regardless of the financial impact to the record. Only querying for a first CC or MCC is no longer acceptable practice in CDI. If we only query for conditions when they impact the DRG, we are doing our physicians and providers a disservice. This behavior has potential positive downstream effects on quality metrics beyond the scope of typical CDI programs.
  3. Query consistently to determine present-on-admission (POA) status of conditions. It’s very easy for us (as clinicians reading the record concurrently) to make assumptions and assume that the coder will realize something is POA.
  4. Query consistently if a condition or conditions are documented in such a way that they are unclear, inconsistent, vague, or non-specific. If a condition is not clear for us reviewing a record concurrently, it will likely not be clear for the coder and is at risk for not final coding or requiring a retro query.

I don’t think we should query simply out of fear of denials. That’s no way to live or to work. But we do have to be aware of the healthcare climate of today. Hospitals are depending on us to prevent denials as much as possible – and confirming accurate documentation of diagnoses irrespective of impact is our responsibility. 

I hope this spurs some critical thinking for CDI specialists. At the least, I hope next time you hear someone say “Yeah, I’m not going to query, this record is maxed out,” you might encourage them to think again.

 

Cognitive Emulation: Insights from Iodine Data Scientists

By Lance Eason, Chief Data Scientist & Jon Matthews, Data Science Manager 

Q: Iodine has pioneered a new machine learning approach called Cognitive Emulation. How would you summarize this approach?

Lance Eason (LE): There are two sides of the picture to consider when thinking about clinical documentation integrity (CDI): 1) what is really happening to the patient, and 2) what has been documented and written about the patient. CDI is all about making those two pictures align, specifically making sure the documentation actually reflects what is clinically going on with the patient. Iodine’s Cognitive Emulation approach helps identify where there are discrepancies between these two sides by looking at the entire clinical picture, not just what has been documented or what aligns with certain rules or thresholds.

Q: Iodine has iterated on its intellectual property over time as new technology becomes available. How have different types of artificial intelligence been utilized, and what has been the experience with each?

Jon Matthews (JM) and LE: Iodine started off with a rules-based approach just using NLP, which calculated the probability conditions were present based on a simple set of rules. We started identifying challenges and limitations – issues with false positives and false negatives, for example – so, we decided we needed a more advanced technique. We explored coupling  Machine Learning and advanced Natural Language Processing (NLP) to holistically review the entire patient record, and found the approach worked well. This allowed us to move past surface-level documentation improvements constrained to specificity and capture the full spectrum of possible discrepancies between documentation and clinical reality. 

Q: What underlying technology does Cognitive Emulation rely on, and how is it different from legacy solutions on the market?

LE: Iodine’s combination of machine learning and NLP allows it to make separate judgments about the clinical state of the patient versus what is documented. This is a more statistical approach and does not rely on specific hard factors to determine if the patient’s symptoms are above or below certain markers. Instead, it calculates a statistical likelihood that a documentation opportunity exists based on clinical evidence. We let the technology find the patterns and connections between the data rather than arbitrarily defining rules ourselves.

Q: What is a marker-based or rules-based approach, and how is Iodine different?

JM: Rules- or marker-based solutions require either the client or the software authors to define what each condition means, so there is not a lot of flexibility. These approaches follow “if, then” logic and make simple yes or no decisions based on whether inputs are above or below certain values. The problem with this strategy is that one could spend forever iterating on those thresholds manually and still get poor results. You lose a lot of nuance in the data because what is apparently useful might actually be missing a large subset of important features in the models. Machine learning algorithms are able to find these features on their own and do so automatically, which takes the guesswork and years of iteration out of the picture.

 Q: How is Iodine’s Cognitive Emulation approach impacting healthcare?

JM: The biggest advantage of Iodine’s Cognitive Emulation approach is that it is able to leverage the experiences and knowledge of physicians, coders, and CDI Specialists (CDIS) across the country into one product. Iodine’s very broad data set allows it to define new and interesting features that are helpful for predictions, which you would not be able to do if you were just coming up with rules on your own. In this way, Iodine helps hospitals work with the resources they have to capture more opportunities for documentation improvement. With Iodine’s vast dataset, experience, and clinical expertise, we are able to build products that drive revenue integrity and labor optimization. So many functions in a hospital require guessing or people to spend time doing unproductive work. The sky’s the limit for what Iodine can do for health systems with our machine learning models. 

LE: There are a lot of patients in the hospital at any given time, and their medical records are each constantly being updated with new data. The job of a CDIS is to survey all of the patients all of the time. With an infinite amount of CDIS you would review every case every day, but this is unrealistic without the help of technology. Iodine does the job of filtering and prioritizing, emulating what the CDIS would have been doing, but constantly and for every case, so that CDI teams are spending their time on the cases that most require human review. This allows them to do more productive work, not just searching through all of the cases to find the ones to isolate for review. If you think of CDI work as searching for needles in haystacks, queries would be the needles and the hay would be all of the cases that do not require a review. Iodine removes most of the hay so the needles can easily be identified.

Machine Learning versus Natural Language Processing: What is the Difference?

Artificial intelligence is utilized for many use cases across the healthcare industry. However, just as all humans have different cognitive abilities, each type of artificial intelligence is distinct, and some applications are more advanced than others.

Artificial intelligence (AI) is a broad term referring to the field of technology that teaches machines to think and learn in order to perform tasks and solve problems like people. 

There are many types of artificial intelligence, but Iodine focuses on two: 

  1. Natural Language Processing, commonly referred to as NLP, interprets raw, arbitrary written text and transforms it into something a computer can understand. 
  2. Machine Learning, a form of applied statistics, solves problems based on large amounts of data by connecting the dots between many inputs without any human intervention. It answers questions similarly to how humans do, but automatically and on a much larger scale. 

What is the difference between the two? NLP interprets written language, whereas Machine Learning makes predictions based on patterns learned from experience. 

Iodine leverages both Machine Learning and NLP to power its CognitiveML™ Engine. These AI technologies work together to analyze, interpret, and understand the information within a patient’s medical record. Specifically, Iodine uses NLP to identify mentions of symptoms, diseases, procedures, medications, anatomical parts, and other key information. These mentions augment discrete clinical evidence including orders, results, medications, and demographic information and are evaluated by Iodine’s machine learning models. Machine learning models are mathematical representations of real world processes that are trained by analyzing vast amounts of data–billions of data points in Iodine’s case. Through pattern matching, these machine learning models emulate physician thought processes to determine the likelihood specific conditions exist and whether there are discrepancies between the patient’s clinical state and what has been documented.

Both NLP and Machine Learning are essential components of Iodine’s technology. Like CDI offerings available from other companies, Iodine uses NLP to determine what the documentation says which can help identify inconsistencies within the documentation and issues with specificity. By itself, however, NLP cannot find many financial and quality accuracy improvement opportunities because:

  • NLP cannot determine when patient information that is supported by medical data is not written in a patient’s chart. 
  • NLP cannot perform clinical validation, in which clinical evidence does not support something that has been documented, which increases the risk of audit. 

The objective of CDI is to determine whether the written documentation aligns with a patient’s clinical reality, and this requires more than NLP. This is where Machine Learning comes in. Machine Learning considers the entire clinical picture and makes connections or predictions based on learnings from other data, including: lab results, vital signs, medications, radiology results.

The result of combining NLP and Machine Learning is intelligent software that evaluates complex medical data similar to a physician’s approach to diagnosing and treating patients. Iodine’s Cognitive Emulation approach (via CognitiveML™ engine) augments the work of health system professionals with software that can actually “think” and learn from each new data point, just as physicians do. Iodine applies physician-like judgment to the clinical evidence in a patient’s chart and leverages previous learnings to more accurately determine the likelihood a condition exists. Conditions often present in a variety of ways, and by relying on clinical evidence rather than ambiguous thresholds, Iodine is able to identify and learn from these unique instances. 

HFMA Voices in Healthcare Finance Podcast: A Surprising Analysis Around Hospital Reimbursement during the COVID-19 Pandemic.

Troy Wasilefsky, Chief Revenue Officer, discusses Iodine Software’s recent analysis around COVID-19 and hospital reimbursement with HFMA for their Voices in Healthcare Finance podcast.

Download the free episode here.