top of page

Automated Medical Coding

Privacy Policy | Cookie Policy

©2023 aiHealth

©2023 aiHealth   Privacy Policy | Cookie Policy

  • LinkedIn

©2023 aiHealth. All rights reserved.

  • aiHealth

The Future of Medical Coding: Perspectives on Implementing Artificial Intelligence



People in a blue and back room full of lights

Coming from an extensive background in medical coding, auditing, compliance, and privacy, aiHealth's Director of Product & Implementation, Cynthia Sherman CCS-P, CHC, discusses the integration of Artificial Intelligence technology and medical coding practices. Cynthia lends insight into the complexities of organizational resistance, the benefits of collaborating with AI, and the future landscape of the healthcare industry.


 

Q: Please provide an overview of your professional journey and the pivotal experiences that led you to specialize in physician-based coding, auditing, and healthcare compliance. How have these areas of expertise shaped your career trajectory?


A: When I first started in healthcare I didn't know what my path was going to be so I did a bunch of different things to test the waters. Most of my experience in healthcare was from my time at the University of Iowa Hospitals and Clinics, where I worked in registration, scheduling, and insurance follow-up and then ended up in the Department of Otolaryngology as a receptionist. I had fallen into coding through a prior job, at an optometrist's office, and once the clinic discovered that their receptionist actually had some coding experience, they started throwing coding tasks my way. I eventually became a member of their coding team, working as an otolaryngology coder for roughly 15 years. I spent that time doing medical coding, provider education, resident education, and clinical documentation improvement-type activities.


I always knew I wasn't going to become a doctor or a nurse but I was intrigued with what they do and all the things in the medical field. Coding kind of became my passion and I fell into my career.


After I left the university in 2013, I joined a revenue cycle management company, MediRevv. At the time, MediRevv was managing all aspects of the revenue cycle, except for medical coding. So I joined as their first coder, kick-starting that division, and scaling up the team. I eventually shifted at the request of leadership into a compliance function where I served as the Privacy Officer and eventually the Interim Compliance Officer. In 2023, I joined aiHealth where I now serve as Director of Product & Implementation.



Q: In the earlier stages of your career, how did you witness the integration of coding practices and AI technology within the healthcare industry?


A: Early on in my career, coding was very manual and paper-based, doctors were still documenting on paper and drawing pictures in their charts, so we had to become masters of deciphering the scribble. As a coder, I would have to do a hundred percent review of all the documentation and handwrite the codes on a slip of paper, handing it over to my partner who would enter the charges into the system. Later, a lot of the documentation became electronic, allowing for better workflows and pre-populated information based on scheduled appointments.


My first exposure to AI, not actually AI but automation, was back in 2013/2014 when one of my clients was using a computer-assisted coding product for their radiology charges. The product was using natural language processing to go through a document, to assign CPT and ICD-9 codes. While we found that the CPT code could usually be accurately assigned by the computer-assisted coding product, the ICD-9 code (the diagnosis code) was never correct, and we would always have to manually review it.


Since around 2009, one of the biggest challenges that I’ve seen in our industry is the shortage of medical coders - it's really never been alleviated, and it just feels like every few years another challenge comes up that is threatening our industry.

The average age of a medical coder is 45, we are not getting any younger and heading closer to retirement. Will there be a new round of young coders willing to learn this complex field? As the healthcare industry continues to rapidly grow, we are facing a continual shortage of coders. It is really important that we start to look at AI integrations that can help alleviate this problem. I really think anyone who hasn’t yet implemented autonomous coding should have it on their roadmap.



Q: Have you noticed any shifts in the perspectives or attitudes of healthcare professionals towards the utilization of AI and coding?


A: I think there are a lot of people who are fearful of using AI for medical coding because they’re afraid that they’re going to lose their jobs. What they need to start seeing is that autonomous coding is really just a helpful coworker, taking away mundane routine procedures. This means that coders get to focus on the more complex cases, working at the top of their coding certification.


There needs to be a shift in focus from the churn and burn of a coder trying to get out as many charges as possible per day. Unrealistic high productivity standards ultimately sacrifice quality in the long run. Coders should be allowed to have their productivity expectations lessened, so quality is once again at the forefront.

The overall goal is to code compliantly with a high level of quality, and AI is just another way to drive that quality forward by allowing coders to spend time on those complex cases


I think there is always going to be a need for medical coding professionals, especially when it comes to reviewing complex cases, auditing, and identifying trends in documentation. In addition, as AI models learn from ingested data, a human is going to need to review to ensure that all codes are appropriately applied.



Q: As someone familiar with the intersection of healthcare compliance and technology, what are the benefits of the collaboration between AI and healthcare professionals?


A: In the medical coding process, there is a really big focus on compliance and following coding standards. Using AI helps to ensure adherence to coding guidelines since the models will continuously be updated based on the latest guidelines and code set changes. Models are continuously improving and learning as they ingest new data. All of the models and algorithms in the background are based on coding standards and regulatory changes and all of that information is factored in, which reduces the risk of non-compliance.


Another benefit of AI is the improved coding consistency. Due to humans having a tendency to disagree with one another, medical documentation can be interpreted differently, which can lead to inconsistencies in coding practices. Artificial intelligence eliminates these inconsistencies by always interpreting documents in the same fashion.


Autonomous medical coding streamlines coding processes, decreases charge lag by 2-3 days, and accelerates cash flow. Organizations have the ability to implement thorough audits, and a real-time feedback loop with their providers. When ICD-10 and CPT codes are entered accurately and timely we see a direct correlation between improved clinical documentation integrity and clinical quality.

In addition to reducing manual coding reviews by upwards of 60+%, another key benefit of our platform (aiH.Automate™) is its ability to change sampling rates, whether it's based on a particular coder, provider, or organization. When bringing on new coders or even providers, it is quite common to want to sample 100% of their work for a period of time, ensuring that they are complying with organizational policies and guidelines. When onboarding new providers, organizations can identify and address documentation issues or concerns at the outset of the relationship.



Q: In general, how do you foresee the future landscape of healthcare and AI technology? How might these advancements influence patient care and optimize healthcare processes?


A: In the future, I think there are going to be really big improvements in the patient care arena. I think AI will be able to help in very complex patient cases where doctors are unsure of what path they're going to take in a treatment plan and assist in diagnosing those complex cases. From an autonomous coding perspective, I think that future models will require less data to train and learn, adapting quickly to new or revised codes. I don’t see the role of the coder going away, but perhaps their relationship with autonomous coding will become even more of a partnership.


 

About Cynthia Sherman:


Cynthia Sherman CCS-P, CHC is the Director of Product and Implementation at aiHealth. With more than 25 years of experience in healthcare, Cynthia has provided leadership and guidance in a wide range of healthcare settings from academic medicine to private sector revenue cycle management companies. Cynthia has experience building and developing highly engaged, successful teams able to address the ever-changing healthcare industry. She has considerable knowledge and experience in the areas of medical coding, auditing, regulatory compliance, privacy, and information security. Cynthia holds certifications and has active memberships with the American Health Information Management Association (AHIMA) and Health Care Compliance Association (HCCA).



A graphic with text saying "the future of medical coding: perspectives on implementing artificial intelligence", picture of Cynthia Sherman

Comments


Automated Medical Coding

©2024 aiHealth. All rights reserved.

©2023 aiHealth   Privacy Policy | Cookie Policy

  • LinkedIn

©2023 aiHealth. All rights reserved.

©2023 aiHealth. All rights reserved

bottom of page