artificial intelligence, brain, thinking-7369048.jpg

Computer-Assisted Coding: A Coder’s Perspective

If you’ve been a Medical Coder for long, you’ll remember a time before CAC or computer assisted coding was implemented. Back when code books were the tool a coder had to use. Most coders would look forward to the new code books to print in October every year (I still do) that would highlight all the new codes and any changes to the guidelines. The sheer joy of reviewing our new books while creating tabs (yes, some of us made our own tabs) for the Tabular list as well as the Index, Guidelines, and the Appendices, brought excitement and curiosity every year. My personal favorite code books were printed by Channel Publishing. Those were giant books that were around in ICD-9 days as well. Fast forward to 2022 where CAC is the coder’s primary tool for productivity and accuracy on every record which is driven by AI (artificial intelligence). CAC was implemented to bolster productivity, maintain ICD-10 codes and increase accuracy through code capture of auto-suggested codes. Ideally, a coder should utilize the CAC to quickly accept codes that are auto-suggested that are either based on logic or knowledge, depending on facility options. These auto-suggested codes are derived from physician documentation throughout the medical record. Sounds like a really amazing way to help the coder “speed up” without having to look up every diagnosis code in their code books right? I mean code books are antiquated now aren’t they? In this age we want everything fast and digital, instant gratification!

Coders that haven’t used code books in the past may not experience any apprehension that some more senior level coders can experience with CAC. There are many reasons “old school” coders still purchase updated code books every year along with many other resources available. While CAC is very effective with proper training and use, there can be a few drawbacks. The benefits do outweigh the drawbacks however, too much emphasis on any one resource could lead to errors, especially if the coder is not trained properly. Just as coders shouldn’t solely rely on code books as the only resource, CAC shouldn’t be the only resource in a coder’s toolbox either.

Pro’s and Con’s

Utilizing the CAC to validate auto-suggested codes is often a required step in the process of code capture within the medical record. Now with CAC, code capture is as simple as a click to either accept the auto-suggested code or to deny it. Most CAC tools also have a function that the coder can use to observe where the suggested diagnosis has been documented in the medical record. This step is a must to not only view the documentation but to also ensure it is a reportable diagnosis. This brings me to an example of a potential drawback for some “auto-suggested” codes. Because the tool is knowledge or logic based, it can suggest conditions or diagnoses that are found throughout the record in areas of documentation that the coder is not allowed to abstract diagnoses from, such as nursing documentation that isn’t signed by the physician or from family history noted in the record. It’s absolutely crucial for the coder to actually review each suggested diagnosis rather than simply accept each one. Accepting the codes suggested by the CAC software without validating can lead to costly errors as well as duplicate codes or redundancies.

One of the many benefits of utilizing the CAC software is the availability of additional resources like “AHA Coding Clinic”or “CPT Assistant” that are readily available to provide expert advice with a few clicks while a coder is performing chart review. Depending on what the facility has included in their CAC software there are many additional resources as well for the coders to use to increase accuracy. Another benefit is the search function that is found in many CAC tools where the coder can search the record for a particular word if needed. For example, while validating a particular code, a quick search can be used to locate treatment modalities to help the coder determine if the code is applicable and reportable before reading the rest of the record. It’s like giving the coder a head start in the actual coding process. The auto-suggested codes can also be accurate and helpful with a diagnosis or condition the coder could potential miss during the review process had it not been auto-suggested. This can prevent errors such as missed conditions that could impact the DRG and missed procedures that also affect the DRG.

Even with all of the benefits the CAC has, there are still several good reasons why facilities still encourage their coders to utilize code books when applying medical codes. Because CAC is a tool for coders to utilize, auto-suggested codes shouldn’t be thought of as complete or accurate without validation. Auto-suggested codes shouldn’t be considered as the only way to capture every condition either. In most cases, auto-suggested codes lack specificity that is often found by completing a thorough review of the patient’s record rather than simply accepting the auto-suggested codes. The chart might be coded “faster” which can enhance productivity however, the process of correcting potential errors and re-billing can be costly as well as time consuming. Preventable errors like this can ultimately bring down the overall quality for the facility and the coder. The entire record must be reviewed for accuracy using a systematic method to ensure accurate code capture.

There are many times when auditors and coders alike go back to the code books to follow the code path, instructional notes, conventions to compare or validate the code capture within the CAC. This isn’t to say that CAC doesn’t provided instructional notes or conventions because all of these items are available within the software. Maybe its the method of following the Index then reviewing in the Tabular, where further explanation, eponyms and similar conditions are found that can assist the coder in knowing they are on the right code path for those new or uncommon conditions that arise. This method has been the foundation of Medical Coding and has provided a “tried and true” course of action for the coder seeking accurate code capture. It also could be the physical act of looking up a particular condition for coders and auditors while utilizing the research skills, that makes the difference. In some ways, we may have forgotten why CAC was to be used in the first place, which was to “help” the coder with productivity due to ICD-10-CM/PCS implementation. The expert level coders were able to navigate the “encoder” efficiently because of their experience level with applying coding guidelines and conventions, they were able to quickly identify, for example, when a condition required a combination code rather than an auto-suggested code. This seems to hold true even now and can be corrected through additional education and training for entry-level coders.

Conclusion

Medical coders are required to utilize resources available to them when reporting diagnosis codes using critical thinking to apply “Official Coding Guidelines” to every condition that is supported through physician documentation by either validating the auto-suggested codes from CAC and/or through record review. CAC provides many valuable resources for the coder during this process, if used properly. Training and education for the coder on the proper use of CAC software is paramount for the coder to reap the benefits of accuracy and productivity. The code books are still extremely helpful and reliable for the coders as well. The more resources the merrier in this present age of continuous auditing.

For additional tips and resources be sure to subscribe to our quarterly newsletter where we bring news that’s relevant to Health Information, upcoming events and discounts for our online courses.

Copyright © 2021 HIM Relevant Insight All Rights Reserved

Scroll to Top