Is Computer-Assisted Coding Worth the Investment?

As healthcare takes more steps toward value-based care, provider organizations need to be capable of doing more accurate and complete coding. When providers do not get all the right codes into a claim, they can end up not getting the correct reimbursement because the patient’s situation will not seem complex enough to the payer.

This is where clinical documentation improvement (CDI) and computer-assisted coding (CAC) comes in. CDI and CAC solutions work hand in hand to make the coding process easier. A CDI solution look at the clinical documentation from the doctor and makes sure that it checks all the necessary boxes. Then a CAC solution looks at all the documentation of codes for what the physician did and makes sure that the claim gets coded correctly. This process ensures that when the insurance company receives the claim, the payer can actually see the holistic picture of how sick the patient is and what the provider did for him or her.

CAC also helps reduce the number of claims denials because it catches potential issues before the claims are even sent out. We as humans are awesome, but there are too many codes to remember. Instead, the CAC system goes through the documentation digitally to suggest what codes it thinks should be used. This doesn’t relieve the entire burden; the coder still has to make the final decision. But overall, the process saves coders time because they don’t have to look for all the codes manually every single time.

KLAS last published a CAC report in 2016 around the time that ICD-10 came out. At that time, providers mainly needed to have a solution to help with implementing ICD-10. Now that things have settled after those changes, we have shifted our focus to finding out whether CAC solutions are worth the investment and whether they provide outcomes for the facilities using it.

All Vendors Improving 

I was actually surprised to see how all of the vendors had improved so much in the last few years. 94% of respondents told us they would purchase their CAC solution again. Even the big players, 3M and Optum, who were rated lower previously, both had a significant increase in their scores. This widespread improvement shows me that these vendors have started figuring out how to provide better customer service and improve their technology across a broader platform.

It made me happy that all of the vendors have made improvements, including Dolbey and ezDI, who were doing well in our previous report. Our goal at KLAS is to help vendors do better. It seems like this group of vendors is pushing each other and lifting each other up.

The Specifics

Even though most vendors in this space are doing well, there are still areas that each vendor performs better in. The report shows pretty clearly who the major players in the space are. Readers will get a feel for which vendors are delivering a better relationship and customer service experience and which vendors have absolutely nailed the technology but are so big that they find it hard to support 700+ customer sites with a consistent level of service.

I recommend reading the report for a closer look at what sets each vendor apart.

Future Improvements in CAC

Machine learning and natural language processing (NLP) capabilities will start showing up more within CAC solutions. All the vendors are talking about these two things. And the more CAC users implement NLP and machine learning, the more we'll start learning and getting better. If NLP gets really good, a doctor could, for example, mention that a patient’s spouse had pneumonia. And the computer would be smart enough to know that though the word pneumonia showed up, it's not actually referring to the patient. Coders would not have to make as many decisions as a result.

One potential improvement that came up in the report is that customers really want to more benchmarking capabilities so they can compare how they are doing with other facilities. A facility might already see that they are doing better than they were. But these providers want to know whether they are doing as good as the top-performing clinics. They want to benchmark coding accuracy and number of claims denials.




     Photo cred: Adobe Stock, Ivan Traimak