By Rose T. Dunn, MBA, RHIA, CPA, FACHE, FHFMA
Healthcare organizations utilize CAC applications for a variety of purposes, some successfully, others less so.
When coding professionals can assign codes more accurately AND in less time than it takes to validate the condition codes suggested by a CAC application, one should not be surprised that many coding professionals believe it is a waste of time to validate the CAC’s suggested codes.
Coding professionals working in all venues, including provider and HCC environments, need to recognize that CAC applications are yet another tool for doing their job. Set up right, with realistic expectations, the CAC application can be a friend to coders. The CAC may even provide an opportunity for coders to once again demonstrate their value to the organization.
All coding professionals have an advantage over CAC applications—a brain! Yes, CAC applications are much faster at reading electronic text and picking out words and phrases that represent health conditions. This speed is particularly beneficial for long-stays and where there is excessive use of copy forward text.
CAC applications are dependent upon the natural language processing (NLP) engines and rules programmed in the application. CAC applications are equally dependent upon the “learning” process that is driven by the coders. When a coder rejects a CAC suggested code and enters a replacement code or when the coder modifies a CAC suggested code, the application “learns.”
Although the CAC learns over time, someone—with a brain—must validate the CAC’s code selection.
CAC applications work effectively with limited vocabulary encounters, such as radiology exams, and often their fine-tuned learning allows the application to code independently of a coder for much of the caseload assigned.
CAC applications work less well with unlimited vocabulary encounters such as those of a medical or surgical provider or a provider whose English is a second-language. These cases pose a greater challenge for the CAC application and require more extensive oversight by coding professionals.
CACs may improve overall productivity and code capture by identifying conditions buried among the pages and pages of copied forward text. But coders are still essential to apply the coding guidelines, combine codes when appropriate, and read the relevant verbiage around the CAC identified condition. The coder is needed to ensure verbiage is not present that would jeopardize the legitimacy of a condition for coding, such as “ruled out,” “history of,” “resolved,” and “not present.”
Explaining the pros and cons of these CAC applications and (to use a term that has been popular during the pandemic) tracking the “false positives” suggested by the application, will help coding professionals demonstrate that computers are not perfect. Demonstrating how your coding skills enhanced what the CAC application suggested, is worthy of touting as well. However, try not to by-pass the teaching of the CAC.
Often the NLP within the CAC can be useful for other purposes. One such purpose may be real-time identification of risk-related issues such as ordered medications that may be contraindicated based on the medication reconciliation in the record.
The firm that supplies your CAC application will be interested in purposes you might identify that would enhance the product’s uses and benefit your organization. Your identification of these opportunities to leverage the use and cost of these applications for other purposes further heightens your visibility and value.
HCC Problem Lists: Challenges and Opportunities
Are your problem lists a problem? An up-to-date and accurate problem list can facilitate patient safety and care as well as being a key aspect of some risk adjustment contracts. HIM and CDI specialists have a role to play. Learn more here.