Mark Morsch is Vice President of Technology at A-Life Medical, and has published several papers on NLP clinical abstraction, text mining and computer-assisted coding. In his first post here on Healthcare-Exchange, he weighs in on the impact of ICD-10 on coders across the industry.
As the buzz at HIMSS this year indicated, ICD-10 is on the short list of things keeping hospital executives up at night. From 17,000 codes under ICD-9 to 155,000 under the new regulation, ICD-10 takes coding and reimbursement to a whole new level – and will exponentially impact the financial health of every hospital. Here is what every provider needs to know as they prepare for the transition:
1. Computer-assisted Coding (CAC) not just a “nice-to have” — It’s no question that coding in the ICD-10 world will be just about impossible without CAC. There aren’t enough coders in the world to support and maintain manual coding of 155,000 codes – making CAC an even more essential part of tomorrow’s health care delivery.
2. Not all Natural Language Processing (NLP) is created equal — The quality of CAC solutions is based on the power of NLP. Precision and recall can measure the performance of NLP, but there are also strengths and weaknesses to each approach. Some NLP technologies may only highlight words and not be able to produce specific codes. Some solutions may match a customized set of phrase patterns that must be maintained at each facility. Other technology may require a large amount of “training” data in order to emulate facilities’ coding practices. All of these methods require hospitals to rebuild their systems to handle the increased complexity of ICD-10 codes in the future. Providers need to closely evaluate CAC systems with an understanding that coding accuracy and consistency can vary widely based on the NLP technology that powers it.
3. Inpatient + Outpatient + Professional = Total coding solution — One of t
he biggest obstacles for health systems is consistency of coding across inpatient and outpatient settings. The technology must cover a broad variety of documentation types, clinical specialties and coding systems. Even within the same facility, clinicians may use different documentation systems including speech recognition, transcription services, electronic templates and structured input. One system that can extract information from all of these different sources with consistent accuracy, and support all venues of care including the professional landscape, will save time, money and significant headaches.
4. Coders need love too — With ICD-10, coders are still very much going to be part of the process — the human touch will still be required — but their lives are going to change dramatically. The use of CAC solutions will elevate the role of the coder to a reviewer or auditor, increasing the overall productivity and accuracy of the coding process. Bringing coders into the process as early as possible rather than waiting until ICD10 is imminent will ease the transition, promote proper training and reduced errors, and maximize the reimbursement potential for organizations in the long run.
5. Timing is everything — and the time is now — If providers aren’t up to speed by the October 1, 2013 deadline, charges will be rejected and resubmissions will be requested, wasting time and losing money. While every hospital has a number of IT projects currently underway, and it may seem daunting to add another, providers need to prioritize the projects that will ensure profitability and return on investment – and with ICD-10 driving every hospital’s reimbursement future, they can’t afford to wait. Hospital administrators should take note of lessons learned from other countries that previously adopted ICD-10. For example, Canadian hospitals experienced up to a 50 percent reduction in coder productivity when transitioning from ICD-9 to ICD-10. So, the sooner hospitals start preparing, the better.

5 comments
Comments feed for this article
June 18, 2011 at 3:21 pm
Naveen
Concise and well written. Out of curiosity, where does that Canada figure come from? I’d be interested in reading that study. Thanks.
Naveen
June 20, 2011 at 5:09 pm
Ingenix Provider Solutions
Thank you for the feedback! This data point was pulled from a HIMSS 2011 Presentation titled, “Implementing ICD-10: Lessons Learned from Canada,” which can be found here. We’ll have more to come on the topic of ICD-10 and Natural Language Processing in the coming weeks, so be sure to follow along and continue sharing your insight.
October 4, 2011 at 8:21 am
Measuring Up: Best Practices for Computer-Assisted Coding «
[...] care organizations. But once you decide which CAC to use — and as I discussed in my last post, not all Natural Language Processing (NLP) technology behind CAC is created equal — the next big question is: how do you measure its [...]
April 19, 2012 at 12:55 pm
Code Explosion: Can current NLP technology handle ICD-10? «
[...] Whenever the regulation goes into effect – currently proposed for October 1, 2014 – the expectation is that ICD-10 will be overwhelmingly complex for coding professionals. The shear increase in the number of new codes in the system – a five time increase in diagnosis codes and a 19 time increase in procedure codes – may be daunting to even the best of coders. However, most do not realize that the code explosion comes from a handful of additional attributes that are being applied to codes they already know. Hospitals and health networks increasingly rely on computer-assisted coding (CAC) systems driven by natural language processing (NLP) technology, but not all NLP engines are created equal. [...]
September 21, 2012 at 5:26 am
Ashok
There are many CAC solutions availanble in market based on NLP, but with advent of ICD 10 code which are very near to each other . There could be a change of single word between two different codes description . I think there shouls be more colloberation between Payer , Provider and regulator . Also please let me know the accuracy rate of your CAC software .