By: Kendall Upton
Professional insights: Olga Lyubar
As artificial intelligence (AI) tools and software continue to advance at a breakneck pace, professionals across numerous industries wonder whether AI will eventually supplant certain human employees, leaving skilled workers scrambling to find other jobs. Many medical coders share these concerns and are wary of AI’s potential to do their job for them.
Though these concerns are valid and warrant further study, so far it appears that all is not lost for the medical coding profession. Human coders are still vital members of the healthcare cycle, and many professionals in the field believe that AI should augment, not eliminate, the need for skilled coders.
Medical coding overview
Medical coders are not like other coders—far from it. They play a crucial role in the billing of healthcare services by classifying the procedures and services a patient receives into standardized codes. Medical billers then use these codes to submit claims to the patients’ insurance provider for reimbursement. Some people perform both coding and billing together, while other employers keep these roles separate.
Medical coders typically utilize codes from one of the three main coding systems:
Medical coders review doctors’ notes and patient records to determine which codes to use. They must be able to understand these documents to know how to classify whatever procedures and services the patient received. Doing so requires a decent amount of medical knowledge, which is why doctors, nurses, nursing assistants, medical assistants and other allied health professionals can make great medical coders and billers.
In any case, accuracy is absolutely critical in the medical coding and billing field. If something is coded incorrectly, it could result in billing errors, denied insurance claims and even legal recourse.
AI in the medical coding field today
Medical coders and billers have manually processed patient documentation for years, but AI’s ability to quickly process large amounts of data and automate certain tasks presents new opportunities within the field.
“AI is being embedded in a lot of healthcare areas and fields. Medical coding actually would be one of the earlier fields that we started seeing more and more AI, especially with transcription [software],” said Olga Lyubar, Department Head for Health Information Management and Medical Billing and Coding at Clark College in Vancouver, Washington.
Transcription software, often in conjunction with natural language processing (NLP) tools, can help summarize and analyze doctors’ notes and patient medical records to make suggestions for which codes are necessary for billing purposes based on that data.
“We’ve seen things like natural language processing (NLP) that helps automate codes and assignment to unstructured patient information, or even structured patient information. With unstructured information, there’s maybe a note written and the AI is able to understand that information and then be able to process a code for that,” Lyubar said.
If AI tools are able to select procedural codes with a decent level of accuracy, the role of medical coders may have to pivot.
“The coder is going to have to adapt their positions with these tools that they’re using. The ultimate coder who’s going to be reviewing the document will have to be more of an auditor, instead of actually looking through the material and coding it initially. I think that’s kind of the biggest thing that’s happened [in the field],” Lyubar said.
Benefits of AI in medical coding
When used correctly, AI software has the potential to reap many benefits for the coding profession, including:
- Identify and correct coding errors
- Automate code selection by analyzing doctors’ notes and medical records
- Minimize claim denials and litigation due to billing errors
- Lower operational costs
- Identify and analyze patterns and trends in the billing process
However, the success of AI all depends on the type of software a coder is using, how well it works, whether the coder knows how to use it properly and the needs of any individual healthcare organization. Lyubar pointed out that there have been cases of hospitals implementing AI coding assistant software that produced a lot of errors and resulted in a lot of wasted time and money.
“If we learn how to use the tools properly, it’s only going to possibly benefit us. And we’re learning really quickly which tools work best for us and which ones don’t,” Lyubar said. “We also have a lot of tools that are actually pretty helpful, especially when we’re talking about billing and coding together, where they do predictive analytics and they’re able to look at multiple data [sources] and analyze trends and things like that.”
Limitations and challenges
In spite of its potential, the AI software of today (and most likely for the foreseeable future) is not flawless. Medical scenarios can be unusual and complex, meaning the correct code(s) may not be immediately obvious to a coder. Puzzling out these unusual scenarios requires critical thinking skills and medical context that AI just doesn’t have yet. AI software may be able to make code suggestions, but that is all they are—suggestions. Humans are still needed to check their work.
And since AI software in this context would be exposed to personal medical information, it’s imperative that any software used in the medical coding field be compliant with the Health Insurance Portability and Accountability Act (HIPAA). Ensuring that these tools follow the best data protection measures is an ongoing process.
Will AI replace medical coders?
So—how safe are medical coding jobs? Though it’s difficult to draw any conclusion with absolute certainty, it does not appear that AI-powered solutions will eliminate the demand for human coders. In fact, Lyubar suggests that medical coders who learn to work with AI instead of against it may be safest.
“The coder who doesn’t learn how to use AI will not have a job, but the coder who knows how to use AI will continue to evolve their position,” Lyubar said, repurposing a quote she heard elsewhere that resonated with her.
The coder who doesn’t learn how to use AI will not have a job, but the coder who knows how to use AI will continue to evolve their position.
The fact of the matter is that no matter how experienced a medical coder may be, human errors can happen. But the same is true for AI—it’s not perfect, and there will likely always be a need for human employees to double check their work.
By cross-referencing one’s work with AI—and creating more of a symbiotic rather than either/or relationship with AI software—medical coders can minimize the potential for coding errors that then lead to billing issues.
“I think it’s just learning the technology, learning to catch the errors. I think what’s going to happen with people who are employed in the coding field is they’re just going to have to make sure that they are more focused, making sure to catch those errors where there might be errors that a machine can’t necessarily understand because of the data that they have,” Lyubar said. “But again, we’ll also see AI tools get better over time because of how they learn and process the data that’s provided to them.”
Job growth projections for medical records specialists (of which medical coders are a part) also suggest that the demand for medical coders should remain strong. The U.S. Bureau of Labor Statistics points out, for example, that as the population gets older, people live longer and chronic conditions like diabetes and heart disease continue to rise, the demand for medical care—and medical coders as a result—should follow suit.
Final say
There’s no doubt that many fields and professions are in a state of flux as new AI technology redefines how we work. The medical coding field is no exception—many healthcare systems are integrating AI technology into medical coding and billing to process records faster and with heightened accuracy. Though this may mean that medical coders could morph into more of an auditing role, the need for skilled coders should remain strong for years to come. If you’re interested in starting a career as a medical coder or biller, hit our Find Schools button to search for medical coding programs today.
