Discover the benefits of artificial intelligence for medical coders.
Those who participated in AAPC's RISKCON 2022, a virtual conference held June 22–23, got the opportunity to network with other professionals and hear the most recent developments in risk adjustment from industry experts.
Technology Solutions for Risk Adjustment (AI, NLP) - Why Coders Will Always Be Important was one of the conference's highlights. It was moderated by Kelly Shew, RHIA, CPC, CPCO, CDEO, CPB, CRC, CPMA, CPPM, AAPC Approved Instructor, and Jacob Swartzwelder, CPC, CRC, CIC, CEMC, AAPC Approved Instructor, SAFe 5 Practitioner. Shew and Swartzwelder contributed their knowledge and experience to the informal discussion, which focused on responding to questions from participants. They discussed the advantages and disadvantages of artificial intelligence (AI) and successfully applied it in a medical setting. The presenters began the conversation by sharing their early, more than ten years ago, experiences with natural language processing (NLP) software.
MedsDental is a renowned Dental Billing Company in the united states, equipped of the revenue cycle experts who are highly proficient in delivering fast and the error-free billing services to the dental practices by using the cutting edge technology.
Developing Pains
NLP is a branch of AI that aids machines in comprehending, interpreting, and using human languages. NLP gives computers the ability to read text, hear speech, and interpret it, enabling them to interact with people using human language.
Swartzwelder made light of the fact that his introduction to AI and NLP initiatives was rather "traumatic" for him. He claimed that the early software was difficult. A word (such as diabetes) would be searched for by software, and the word would appear with a code that you could accept or reject. This was done to save staff members from having to read through several paragraphs in pursuit of a particular word or phrase. The suggested code, however, was frequently erroneous because the programme frequently overlooked the word "not" or other negative notations when they were present.
The same thing happened to Shew, who acknowledged that "there was a lot of learning curve to it." She and her team continued to reject 50% of the codes that their software proposed after a 6-month trial. The software was inputting confusing information from patient histories or prescription lists. Shew observed that "we were getting some bizarre, weird stuff that was being offered."
However, the interface that programmers are likely to use at work nowadays is more interactive. According to Swartzwelder, modern NLP software has improved at comprehending complete words but can still be difficult, particularly when providers free-type and don't utilise capitalization or punctuation. According to Swartzwelder, "Now that natural processing language isn't natural because the sentence isn't natural."
Select a software
According to Swartzwelder, a variety of businesses are concentrating on a variety of different AI-related fields. You'll find both programmes created by pure programmers and those developed by physicians utilising a therapeutic approach. Before installing a programme, carefully weigh the benefits and drawbacks of each type.
Knowing who was present at the table during the product's creation is crucial, according to Swartzwelder. This is not to say that any particular product is good or bad; rather, it is to say that you should be aware of any holes and blind spots in the system you are utilising. To adapt a software to your employees, office size, and level of technology comfort, you may always overlay technologies and find remedies for flaws.
MedsDental billing and coding experts are highly proficient in processing claims foryourdental practice.Moreover,our Dental Billing Services are certified by HIPAA providing you an utmost experience in maintaining the privacy of the communication between you and your patients.
Shew stated that it takes work to get NLP programmes to operate with the electronic health record because they are not plug-and-play (EHR). Since all providers will be utilising the same manner of entry, she advised employing templates within the software. This might help the software understand the style of provider documentation. Otherwise, software may take longer to pick up on human quirks.
Put a Program in Place
In order to avoid having your production fall below acceptable levels, Swartzwelder advised implementing a new AI application gradually. Because "you can't just dump this product, and you will see your team's productivity drop," he said, "you must account for the learning curve in your project plan." It can be laborious to begin with: If a provider is dictating, the software needs to learn their speech patterns, according to Shew, and Swartzwelder added that you must send input to your vendor for improvements. In summary, to get everything functioning smoothly, a little tweaking and trial and error may be required.
In the same way that the electronic medical record (EMR) did not replace coders, according to Swartzwelder, AI will not do so either. In the foreseeable future, he remarked, "I don't envision any form of AI or natural language processing technology completely replacing coders." Shew concurred, saying that too much of what programmers do is complex and difficult to incorporate into software.
According to Swartzwelder, who compared using AAPC's Codify to receiving AI assistance, "the intent of these programmes is to reduce the complexity of complex coding scenarios for the coder or to assist them in being more efficient or to help the coder focus on things that cannot be done by the computer." Without Codify, I would be a less effective coder because it works similarly to Excel in locating NCCI modifications. He said, "I don't have time for that.
Accept technology
Technology comfort varies from person to person; some individuals adore it, others despise it, and the bulk lie somewhere in the middle. What most concerns programmers and why is AI technology intimidating?