In today’s world of ChatGPT and Generative AI, it was only a matter of time before AI made it’s way to the medical coding realm. While AI-generated coding will certainly be helpful, getting it off the ground can be a nauseating process for some. Here are some basic recommendations on choosing and implementing an AI-generated coding platform from a long-term coder’s perspective.
Choose a reputable platform vendor
The healthcare marketplace has been flooded recently with so many coding AI vendors each fighting for a place at the table, that deciding on the right one can seem like a daunting task, and it is important to do your due diligence in making your selection. There are a few important things to remember in the selection process:
- How was the platform created? Has the vendor taken the time to do a needs assessment on how the process should work? I’m not implying that vendors may not understand medical coding, but you want to ensure that the platform was developed to cover all the necessary components and ask who was involved in the development. Was a coder part of the team? Coders certainly understand coding rules so they at least should have had a place at the table in development.
- Does the vendor have an established clientele or are they new to the market? While all new vendors certainly deserve the chance to succeed in this market, those with an established clientele will more than likely have a better shot. Ask the vendor what clients they currently work with (who would know the platform better than those currently using it?) and reach out to those clients. Speak with rev cycle or coders. Ask what the overall experience has been, likes and any pain points. Has was the implementation process? How has the company addressed any technical issues, and have they been addressed in a timely manner? Don’t just take one clients word. If the vendor has several clients currently using the platform, contact as many as you can. One client’s experience may vary from another, so try to get both sides of coin.
Involve your coders in the process
Since the coding team will be one of the primary users of the AI coding platform, they should at least have a say in the process. Involve your coding team in live demos with each vendor allowing them to ask questions (you will be surprised at the questions coders will ask that you did not initially think of). Some things coders should look for are coding accuracy, potential missed coding opportunities, and overall efficiency. Ask the coders what they liked and didn’t like about the platform and strengths and weaknesses. Take any major concerns back to the vendor and have a detailed discussion. How does the vendor plan to address those concerns and could potential modifications be made to the platform? Many vendors are usually flexible in this area and can certainly modify the program to fit your specific needs, but make sure that your coder’s concerns are heard.
Address Privacy and Security Concerns
In a world where cyber-attacks and data breeches have now become common practice, privacy and security of patient data should be a top concern when choosing an AI-generated coding platform. Keep in mind that this is an outside vendor coming into your EMR and patient data is being put into private hands. Find out from the vendor what safeguards they have in place to protect your data. The vendor should be able to answer that question as privacy and security should be in front of their minds as well. Work with the vendor is needed to address the issue.
Calm coder anxiety
A recent report on what keeps coders up at night published in Becker’s Hospital CFO Report shows that” 94% of those surveyed said they are concerned about the accuracy and nuance of AI-generated codes and 97% believe critical human oversight may be lost. Additionally, 77% said they are concerned about AI’s potential to perpetuate biases in billing and access (Cass, 2024).” Another fear in the industry is that AI is completely replacing the need for coders in general. All of these concerns are certainly valid, and healthcare entities must do their part to help calm these fears amongst their coding staff. While AI-generated coding is certainly helpful, it is not a foolproof method and coder interaction is still needed to ensure accuracy and minimize denials and potential audit risks. Any healthcare entity that implements an AI-generated coding platform and relies on it solely for their coding without the human touch is opening a Pandora’s box, so make sure your coder’s are being seen and heard during this time.
I know these tips sound basic in general, but these are things you should keep in mind when selecting an AI-generated coding platform. Best of luck in your journey.
REFERENCE
Cass, A. (2024, October 23). What’s keeping medical coders up at night? Beckershospitalreview.com. https://www.beckershospitalreview.com/finance/whats-keeping-medical-coders-up-at-night.html
