Home  |   Subscribe  |   Resources  |   Reprints  |   Writers' Guidelines

Summer 2023

Coding Corner: Medical Coding Is Going Hybrid
By Taylor Ross, CPC, CCS
For The Record
Vol. 35 No. 3 P. 8

How Innovative Coding Teams Integrate Automation

Medical coding is evolving. While human coders have done the job for decades, automation based on artificial intelligence (AI) is now reinventing the process in a way few imagined.

Innovative coding teams are embracing AI’s benefits as part of a new operating model. This hybrid approach combines the best of human coders and technology: Coders bring vital experience and intelligence, while advanced technology and AI allows them to unlock their true potential and reach new levels of efficiency. For the broader organization, this hybrid approach also drives improvements to patient care as it helps reduce the administrative burden for physicians and free their time for patients.

What Do These Hybrid Models Look Like?
A hybrid coding operation applies the capabilities of coders and technologies in this manner.

#1: AI Tackles the Bulk of Encounters
The shortage of medical coders has created a serious problem for health systems and physician practices. Often, coding teams are overloaded with work. This is especially true for simple encounters such as routine checkups, blood tests, and health screenings. These high-volume encounters can pile up over time and create a bottleneck, resulting in chart backlogs that delay reimbursement and payment.

Automating high-volume encounters is the beginning of the journey to building your hybrid coding team. These encounters comprise the bulk of the workload and are typically the least interesting for coders.

Medical coding automation is well suited for these routine, simple encounters. This is why innovative coding teams start their hybrid model here. The AI should be performing more than 80% of the coding volume and sending it straight to bill without any human intervention—typically in fewer than 90 minutes. Automating this work frees coders to focus on higher-value activities.

For instance, one national physician group brought on coding automation for its radiology volumes in early 2022. The AI vendor was able to ramp up to a 97% automation rate, leaving just 3% of encounters for the coding team to handle manually and thus freeing significant capacity for other revenue cycle tasks.

#2: Coders Complete the Complex and Interesting Encounters
While medical coding automation works well for most encounters, the technology has its limits. Some fraction of encounters might be too uncommon or complex for it to handle well, such as cases that involve patients with rare diseases, extensive medical histories, or multiple chronic conditions.

As a rule of thumb, complex and unusual encounters—and those requiring a detailed review of medical records—may be beyond the scope of some technologies’ ability to code confidently today (although these tools are constantly improving). This is where coders’ experience and specialized knowledge take over.

Technology allows coders to give more detailed attention toward completing complex and unusual encounters, leaving the highest-volume and routine cases to automation. These complex encounters are often more stimulating and interesting for coders and allow them to use their expertise to ensure proper coding and reimbursement for what are often higher-dollar cases.

For example, one regional health system that implemented AI for emergency department coding in the fall of 2022 was able to reduce its preventable denials rate by 9%. With the AI handling the vast majority of patient volumes, the coders reallocated some of their time toward managing high-value claims and other valuable revenue cycle activities, such as responding to potential denials.

#3: Coders Supervise the AI’s Results
While providing coders the time to focus on more interesting encounters is a plus, this next area is where the real fun begins. This is where innovative medical teams truly begin to look toward the future of the industry.

Imagine a command center similar to what you may see in the movies. Rows of computers and high-resolution monitors fill the room, and coders’ eyes are fixed on screens that display streams of coding results coming in from automated systems. The coders are accomplishing a very important job: they’re supervising the quality of the AI’s results.

Coders can easily audit to ensure quality, communicate with each other, and provide feedback to the automated system. And when a more complex encounter is spotted, it’s delegated to the team member best suited for it. Not only does this setup provide coders with a strong sense of collaboration but it also ensures high-quality, fast, and efficient work that many health care facilities have never imagined possible.

In this hybrid model, coders play an integral role in the coding process. Though automation handles the bulk of the work, the skills of coders are needed to quality-check results, handle the most complex encounters, and ensure the process moves smoothly. This is an exciting atmosphere for coding staff. They’re harnessing the most advanced technology available and leveraging it to improve their teams.

The Result: Coders Spend More Time on High-Value Activities
With more free time, coders can focus on high-value activities that can improve overall revenue cycle health. These include the following:

• Preventing claims denials: While claims denials are a part of any health care organization, the rate can often be reduced. Coders can play a key role in this process by conducting routine audits to identify trends in denials. If they detect recurring coding or billing issues, they can prevent future denials by communicating with payers and providers to resolve them.

• Improving upstream documentation: By improving upstream documentation, coders can improve the efficiency of the reimbursement process. By responding to patterns in coding and claims results, coders can suggest improvements to documentation, ensure documentation is accurate and complete, and streamline the process.

• Educating coding teams and providers: As the team members work closely together to oversee quality and review higher-complexity cases, they’ll be able to refine their processes, upskill other coders, and improve automation training. Coders also can educate providers and doctors at the point of care. For example, coders can teach physicians to be more explicit and clear with their language to represent each case correctly. By educating providers and other coders, the efficiency of the revenue cycle improves as a whole.

A New Day Has Dawned for Medical Coding
With automation, the future is bright in the medical coding world. A hybrid approach gives coders the space to take on more challenging tasks and harness the power of advanced technology. Health care systems and physician practices will reap the benefits of lower costs, a faster revenue cycle, and improved patient care.

While this vision may sound like it’s years or even decades away, the future is now. Forward-thinking health systems and physician practices are already implementing hybrid coding models. Doing so has enabled leaders to plan their organizations differently, improving overall efficiency and allowing them to better manage workflows, reduce backlogs and denials, and accelerate reimbursement cycles. The result? These organizations have become more productive and profitable while improving patient care.

For your organization to become a successful hybrid model, automation technology and coders must work hand-in-hand. Innovation in revenue cycle management won’t flourish without both. n

— Taylor Ross, CPC, CCS, is the strategy and operations lead at Fathom (www.fathomhealth.com), a health technology company that uses deep learning artificial intelligence to automate medical coding for a wide range of specialties and practices. There, she leads coding quality and verification, working cross-functionally with customer success, engineering, and product development to support health care organizations through onboarding, production, and ongoing quality assurance. Ross also manages strategic analysis, client analytics, and reporting. Prior to joining Fathom, she was a consultant with Berkeley Research Group, specializing in coding and compliance consulting for various health care organizations and physician practices. Ross graduated from the University of Pennsylvania with a bachelor’s degree in economics and minors in mathematics, health care services management, and the biological basis of behavior.