Tech & Innovation in Healthcare

Find out if auditing could be a new career option.

Artificial intelligence (AI), natural language processing (NLP), and computer-assisted coding (CAC) are technologies that can make medical coding tasks more efficient and productive, but not everyone is enthusiastic about the platforms.

Kristen Viviano, MHA, CPC, CRC-I, AAPC Approved Instructor, senior product manager and eChart coder at Veradigm, examined the topic during her AAPC RISKCON 2024 session, “AI Coding World & Why Coders Are Still Needed.”

Read on to get answers to three frequently asked questions (FAQs) medical coders have about adding AI to their organization.

If AI Processes Data Faster Than Humans, Are Our Jobs on the Line?

AI has been front page news in recent years because the technology is evolving, transforming many industries, and can take on many personal and professional duties to save you time. However, not everyone is thrilled at the possibility of the technology joining the workforce. For example, some medical coders are worried that deploying an AI coding software in their practice might mean the end of their jobs.

The truth is, even though AI can process more records of varying complexity than a human coder can in the same amount of time, increased productivity doesn’t necessarily mean quantity, according to Viviano.

“What AI can do is not just increase your productivity to say you can look at more charts in a day, but the technology is looking at your team as a whole,” Viviano explained. If you’re part of a coding team and need to examine as many charts as possible, then AI can evaluate the simple cases or isolate key terms in complex cases to help you look at more records in a workday. “The technology can increase your productivity for the whole team, to say, if you only can budget for 10 coders, you'll be able to get more charts reviewed with those 10 coders with utilizing AI or computer assisted coding (CAC),” Viviano said.

Not to mention, human coders also possess the critical thinking skills needed to interpret coding guidelines, parse through documentation that’s not straightforward, and navigate gray areas in documentation and coding. There will always be medical cases that require human intervention to ensure the cases are coded correctly.

How is AI Maintaining PHI Security?

Securing your patients’ protected health information (PHI) is critical when using AI and similar technologies in your healthcare organization to ensure you’re complying with HIPAA requirements. Your IT and cybersecurity teams should take the necessary steps to shield patients’ PHI from falling into the wrong hands.

Encrypting the data protects the information from others viewing, overwriting, or stealing the data for decrypting the data later. This enables those who need to access the data to be the only ones to do so.

On the other hand, cybersecurity teams will also use red teaming to try to break into the system. “[Teams] have structured testing and you have somebody try to breach the security of the system, so you can find potential flaws or unforeseen behavior. It’s really putting it to the test as if you're trying to do something malicious,” Viviano stated.

Employing access controls ensures only authorized individuals have access to sensitive data. Role-based access controls are security protocols that provide access to data to certain employees whose position requires them to view specific parts of the patient’s information. For example, a front desk staff member may have access to billing, contact information, and scheduling, but cannot view the patient’s record.

This means your practice will need to establish minimum necessary access controls to allow staff members to have the minimal necessary access needed to perform their job duties. Deploying access controls helps reduce the risk of a data breach.

Additionally, standard cybersecurity practices, like implementing firewalls and continuous monitoring, will help AI platforms protect PHI security. Firewalls shield the AI technology from outside attacks, while continuous monitoring lets your cybersecurity team keep a constant eye on your practice’s network to detect and prevent security breaches.

How Can We Embrace AI in Our Organization?

Using AI in your medical coding and billing has several positives and disadvantages, but ultimately, it’s up to you to decide whether to use the technology. If you choose to deploy AI in your organization, you’ll need to help your staff embrace the platform to ensure its success.

Collaboration: One way your organization can embrace AI is by working with your IT team. They’ll want to hear what is working well and what could be improved to help ensure the technology is achieving its full potential within your practice.

“Make sure that your feedback will be heard, but make sure that the IT team has someone familiar with coding, who is working and looking at that tool to make sure that you know they're understanding that it’s not going to be a solution to replace coders,” Viviano stated.

Education: Next, you want to encourage your staff to learn as much as possible about the technology. Many medical coders are worried that AI’s speed and accuracy will end up putting them on the street, but having lunch-and-learn sessions, providing links to online videos, and offering articles on the technology can help ease their concerns. The nuances of medical records will always need a human element to verify the accuracy of the AI coding solution’s suggestions.

Stay involved: Your staff will be more receptive to AI when the correct checks and balances are put in place. For example, when your practice is evaluating different AI options, you’ll want someone familiar with the technology in the room. They can ask the pertinent questions relating to security, accuracy, and compliance, as well as how user friendly the interface is.

Plus, as mentioned above, humans will always need to be involved in medical coding to review codes suggested by the AI and ensure the recommended codes adhere to the guidelines. Coders who are credentialed in auditing, such as Certified Professional Medical Auditors (CPMAs), could carry greater weight and offer a new career role.

“Going forward, I can see that. That is really a good opportunity for coders to be auditing and checking the work of the machine learning and AI programs,” Viviano added.

Mike Shaughnessy, BA, CPC, Development Editor I, AAPC