Home » Interviews » Currently Reading:

An HIT Moment with … Andrew Kapit

February 2, 2009 Interviews 3 Comments

An HIT Moment with ... is a quick interview with someone we find interesting. Andy Kapit is CEO of CodeRyte.

CodeRyte is successfully attracting investor interest and increasing staff, both unusual in this economic environment. What are you doing differently from those companies that are shrinking instead of growing?

andrewkapit CodeRyte first identified a real pain, chose the right technology to cure that pain, and then executed it in a scaleable way. Investors have learned a lot since the 1990s and the technical due diligence process during our most recent funding shows that. It wasn’t just that CodeRyte developed a powerful NLP-Computer-Assisted Coding engine — it was how we did it that has enabled us to create investor and market confidence.

CodeRyte then built a solid based of clients by working diligently to appropriately set expectations, exceed them, and then provide the highest possible level of client service. When bringing a disruptive and sticky technology to the market, it is of crucial importance to let your clients know that you will stand by it and be there for the long term.

CodeRyte developed a strategic platform that builds on itself. By automating medical coding, CodeRyte is really providing structure to the language of medicine. Every medical record that passes through the coding engine, as nearly one hundred million have, adds to the knowledge base and enables us to provide increasingly greater granularity to the structure of clinical information in the report.

We started off with radiology and pathology, the diagnostic specialties of choice. With that linguistic foundation the technology can now code across the full spectrum of medical specialties. This brings me to the final reason CodeRyte has been able to build a successful investor and customer base: the technology leverages the expertise of our users.

CodeRyte is built around several fundamental tenets, one of which is to empower the experts, to allow them to help create the intelligence of the engine and then to create the “explainability” so they can not only become comfortable with the output but also understand how the engine arrived at its answers.

The CodeRyte team combines technological brilliance with a deep and abiding understanding of the complexity of healthcare — at the macro and micro levels. Engineers working with medical coders, subject matter experts and physicians have all created something that is truly relevant to the current and future of healthcare, which is what investors look for — relevance and the ability to execute.

But that’s only the beginning. That’s what gets today’s investors to the table. Once you have satisfied the basic criteria (real market pain, strong management that can execute and a scaleable solution), then investors want a real and meaningful upside — which is where the company’s ability to describe its mission in a meaningful and value-generating way comes in.

CodeRyte fundamentally believes, along with many, that we need machine readable, interoperable, and structured output. It is only by creating that type of output healthcare can enjoy an industry architecture that successfully aligns the flow of information and money and reinforces efficient, quality care – over the long term.

What we don’t believe, however, is that we should force physicians and other allied health professionals to create that structure. Not only will it reduce the quality of our source data for research and discovery but it also disrupts physicians’ intellectual workflow. CodeRyte’s technology can already, to a limited degree, automatically structure the output — without forcing physicians to change the way they provide and document their care.

You’ve worked in executive roles for providers, investment banks, and now vendors. What is your analysis of the healthcare IT market now and what changes do you predict over the next 1-3 years?

It is not just that healthcare is broken; more importantly it is that American healthcare is not scalable. I believe that every new patient coming into the system will be more expensive than the one before. To make the necessary changes we, as an industry and country will have to commit to doing the right thing — even at the expense of the some of the entrenched processes, technologies, and companies. This cannot be a self-serving, short-term thinking revolution as it has been in the past.

In short, predictions for the next few years of the HIT market depend largely on what questions are asked and addressed in the near term. If all we ask is "how do we get everyone covered" then we’ll end up bankrupting the system and ruining the quality of healthcare’s outcome. If, however, we ask "what architecture and infrastructure will allow us to best improve access and quality and what do we have to destroy and rebuild to get there?", then with hard work and true innovation, we can create a world-class healthcare system that will not only cost less but provide much better outcomes for the country — at the expense of some of the companies that have been winners at our expense over the last many years.

The government tries to fix what it can see, which is primarily CPT and ICD codes –- codes that have been polluted from clinical purity through the process of coding to get paid. The incentives for clinical improvement are through the filter of a system that financially rewards based on more procedures, not improved outcomes. The entrenched vendors support solutions that will benefit their bottom line — no doubt assuming that they will help in the process. Now, however, we have to be willing to look at the actual architecture of the system in order to make enduring progress. My fear is that with the crisis at hand a new and energized government will try to tackle the most visible problems — which will make the situation worse. Without fundamental change, we will only be piling on more expensive process, as we historically have done, without creating an ending.

What makes today different from the past is that there are new and truly disruptive technology companies and people running them who are not committed to the status quo. If the new administration is willing to ask the right questions, involve new people with innovative ideas, and is willing to create a new industry architecture, then and only then will we be able to enjoy the healthcare system we deserve. If, however, we stay with this basic model, then we will get the system and outcomes deserved by the incumbents to the system — and that would be a true shame.

Outside the possibilities of a new administration, the impact of primary care reform efforts cannot be understated. The legislation that recently passed in Massachusetts, aiming to make primary care more attractive, is another example, from a completely different angle, of how we can affect the actual architecture of the healthcare system. By legislating for loan forgiveness, home purchase help, improved reimbursement, and more, we can make primary care attractive again and quell many of the healthcare systems endemic problems — proper disease management, reliance on emergency rooms and more — earlier and more effectively.

Are providers doing everything they can to maximize the payment they’re entitled to through accurate coding?

No. And it is even worse than the question suggests.

In our travels around the country, we meet providers who intentionally down-code out of fear. They down-code because they are afraid to get audited, afraid that the system will not be reasonable. They are afraid to stand out and afraid that the fact that their patients are ‘sicker’ means that their higher codes will make them stand out. They are afraid — period. The truly ironic shame is that this adversarial culture not only reduces the morale of the physicians, it forces the data to be more flawed than it needs to be. The current flow of healthcare’s information and money do not support the gathering of life-saving data and high-quality outcomes. The data is systemically and systematically flawed.

Think about it — the most complex series of events most people endure in their lifetimes are reduced to three-, four- or five-digit codes – whatever follows the path of least resistance. These codes determine what’s likely to get paid and most easily treated, but never account for the actual patient. What’s best for them, what diagnosis is most accurate, what treatment they should pursue, what options they should consider — the list is endless. What is truly important in today’s system is what those codes are being used for — reimbursement. Not treatment. Not the most appropriate care. Not the patient at the heart of it all.

Physicians have these well-trained powers of observation and, with the full color of their narrative, describe what is wrong with us and what they are going to do about it. In that language are rich and complex concepts — some of which are negated, historical, related to a family member, or are equivocal because more information is needed. Does all of that valuable information get captured in the medical coding process?  Not even a fraction of it. The information captured in the record accurately reflects the actual health of the patients. The information healthcare uses to evaluate the quality of care and outcomes is inaccurate — out of fear and is both measuring and rewarding the wrong things.

Without truly capturing what is going on during a patient’s full episode of care, physicians are not only short-changing themselves, but the system is being hurt at the same time. There is no true discovery without having access to and being able to analyze all of the information. The people who can and truly want to make a difference are crippled by the inaccurate and limited information available to them.

What are your thoughts on ICD-10?

ICD-10 is an important and overdue advancement for the United States. It’s necessary to meet the needs of an increasingly complex, diverse, and electronic medical environment. By implementing ICD-10, the industry will advance from the use of a classification system in need of modern information to one that accurately reflects advances in diagnosis and care.

That said, it’s amazing that our great and competitive country is almost 20 years behind other parts of the world in this regard. Countries that are already using ICD-10 have access to important information about diseases that are being missed in the U.S. And why? It goes back to what I stated previously — there are some entities within the system that are so heavily invested in, and exert so much control over healthcare that without their "approval" the system is frozen. Big changes that would mean improving the system are put off in favor of the established status quo. If we let those few players control our destiny, then we are destined to get the healthcare system they deserve.

Furthermore, the pushback ICD-10 is receiving is not based on its utility or value, but rather the technical cost of the change. This is yet another example of how unscalable U.S. healthcare has become. If we continue to as we have, the problem will only get worse, and then at what cost?

Stepping off my soapbox, we need information. Good, granular, and specific information. The current coding system prevents that. ICD-10 will help, though it is not the complete answer. ICD-10 will give the healthcare industry a better chance to improve the way it handles documentation and coding operations and, at the same time, elevate the power of biosurveillance and pharmacovigilance.

The overwhelming scope of the disruption stemming from ICD-10 cannot be understated. The switch represents a change from 17,000 codes to more than 155,000. Given the diversity in size, specialty, and payer mix within the healthcare industry, the complications of the change are nearly impossible to properly measure. But so, too, are the new codes’ possibilities in terms of technology adoption, research, and discovery.

What makes you happy about running CodeRyte and what gives you satisfaction away from work?

There are so many things that I love about CodeRyte and being its CEO that it’s hard to list.  But, I’ll try …

The more than 100 people at CodeRyte are incredible. Given what we do, we have extremely talented people from myriad backgrounds — talented engineers, nationally recognized NLP experts, certified professional coders, driven sales executives, dedicated account managers, and others who have gravitated to CodeRyte out of a strong desire to revolutionize healthcare. Their passion, attitude, and intelligence are inspiring and exhilarating to work with.

What binds us is that we are all aligned around a common vision — CodeRyte and its technology can truly help revolutionize healthcare. Now we just need a seat at the adults’ table in order to show the industry that there are ways to achieve what they all believe to be impossible. I’ve said for years that we’ve been seated at the kids’ table while we’ve built and imagined the future possibilities for this company, its technology, and other HCIT vendors. The frustration of getting the ear of the change-agents is truly like tilting at windmills. We have the answer; it is just that the windmills don’t always want to listen.

You know that look when you give a child the present they have been waiting for? Well, that’s one of my favorite things about running CodeRyte — the look on our clients’ faces when we deliver on our promises. Increased productivity. A lowered cost to collect. Faster turnaround times. Appropriate revenue capture.

I also love the challenge of doing that which people say can’t be done. In the beginning, for example, we were told physicians know how to code, so why would they use an NLP coding application? Physicians should not have to assign codes — it takes away from their patient-facing time and should not be their area of expertise. Our application helps the physicians by allowing them to focus on delivering quality care and gives them back their day.

Away from CodeRyte, my priority is my wife and children. The importance of a work-life balance cannot be underestimated by the CEO of a company that aims to revolutionize an industry as important as healthcare. The time I give to one always takes away from the time I can give to the other. This means that for now my passion is balancing — giving the most I can to building CodeRyte and loving my family.

View/Print Text Only View/Print Text Only

HIStalk Featured Sponsors


Currently there are "3 comments" on this Article:

  1. The upgrade to ICD-10 will not change what the codes are used for. To quote Andrew himself: “…reimbursement. Not treatment. Not the most appropriate care. Not the patient at the heart of it all.”

    I agree that the pushback is almost entirely about the cost, and that it is wrongheaded. It is surprisingly easy to pushback on the structure of ICD-10-CM.

    The upshot of it all is that “accurate diagnoses” are impossible with a disease classification, as opposed to disease, coding system. So long as you code the class of diagnoses into which a diagnosis falls, and not the actual individual diagnosis, you have a problem.

  2. Interesting article raising a couple of questions:

    1. Can Andy share some metrics regarding CodeRye performance (such as precision, recall and their combined F-statistic) ?
    These metrics will provide an insight into CodeRyte capabilities for coding righteously…Minimum under-coding (loss of revenues) and minimum over-coding (auditing, penalties, etc.) Unfortunately as with other binary classifications techniques, a gain in one metric means a loss in the other and vice-versa. Thus the F-statistic (no pun intended).

    2. Does the NLP system “learn” in the process ? Does the algorithm performance improve in time?

    3. How long it takes to demonstrate a ROI for a medium size hospital ?

    Many thanks,
    Alex Scarlat MD

  3. Dr. Scarlat, thank you for asking these important questions and the answers are: yes, we are happy to share metrics about our performance; the system does “learn;” and the length of time to demonstrate an ROI depends on the hospital’s prior performance – but our rapid growth, 120 clients and almost 100 percent retention should make clear how powerful the ROI is.

    I agree that to evaluate the quality of an NLP application it is imperative to know its precision and recall scores. In common language, recall measures the engine’s ability to accurately identify the full range of correct answers and precision evaluates the accuracy of the answers.

    For example, when CodeRyte processes a note and identifies 10 ICD-9 codes and an expert coder agrees that each of those 10 codes were appropriately supported by the documentation, then the precision would be 100 percent. However, if the expert coder also thinks that the engine should have identified 20 ICD-9 codes, then the recall score would be 50 percent.

    As Dr. Scarlat points out, there can be a trade-off between precision and recall. If it were more important that every ICD be right, then an NLP vendor might be willing to have lower recall. On the other hand, if it were more important that no ICD code is missed, then an NLP vendor might be willing to have some incorrect ICD codes in order to make sure that all the correct ones were included.

    Thus, the F-Score…

    In addition to creating a high-performing coding engine, there was an additional challenge that CodeRyte had to overcome to successfully bring our NLP engine to the market. We had to put “accuracy” in context. In the scenarios above, for example, one would expect that three medical coders would agree that each of the ICDs selected by the engine were correct and, furthermore, would agree on exactly which ICD codes should have also been included. Unfortunately, that is not how it works. There is significant disagreement among coders.

    Because CodeRyte believes in transparency, and especially when bringing a disruptive technology like this to the market, we have already published metrics regarding CodeRyte’s performance. The papers, which get to the heart of your questions, were written by our team of engineers and NLP pioneers to dissect automated confidence assessment for compliant accurate coding.

    Because those papers were published two years ago and our scores have improved on every metric, I suggest that we set up a meeting during which we can provide you detailed answers to your questions.

Subscribe to Updates



Text Ads

Report News and Rumors

No title

Anonymous online form
Rumor line: 801.HIT.NEWS



Vince Ciotti’s HIS-tory of Healthcare IT

Founding Sponsors


Platinum Sponsors






















































Gold Sponsors
















Reader Comments

  • FRANK POGGIO: Mike, I agree w Mr. H. Your example of network management is just one application not a full business operational syste...
  • Eyes Wide Open: Re: "Has anyone ever analyzed the number of retread healthcare sales roles? It seem a bunch of folks just flit from comp...
  • AC: "323 hospital and clinic visits and 13 major surgeries" How do you get surgeons to perform 13 unnecessary operations?...
  • Desperado: "The vast majority of us conduct our business with integrity and understand that the more we give, we get back tenfold i...
  • Long time HIT salesman: Teri, I couldn't agree more. The vast majority of us conduct our business with integrity and understand that the more we...
  • Supporting good decisions: Mike, I read the original and I read your comments. You make some sweeping assertions without any basis in fact....
  • Teri Thomas: As a former vendor salesperson, I genuinely like lots of CIOs and would consider them friends. I was in my role a long t...
  • Mike Jude: Very interesting take on the Frost & Sullivan CDSS report. Since I wrote it, I thought I might provide some perspec...
  • Brian Harder: I get schadenfreude when learning of the woes of people who get too cocky (referring to the NULL license plate guy). ...
  • Math: Dittoing on how awesome Zenni is, I've been using them for a couple of years. Most people don't realize how much of an ...

Sponsor Quick Links