Many organizations are incorporating Computer Assisted Coding Programs (CAC) into their abstracting systems with the goal of increasing Charts Per Hour (CPH) in their teams and improving the organization’s cash flow with no added manpower cost.  It’s true, a coder’s performance may increase using a CAC if documentation is thorough and in one place, workflows are efficient, and the coder has sufficient technical training.  The natural language processor (NLP) learns and adjusts to suggest correct codes based on the provider’s documentation.  The coders can be more productive because they only have to validate the code suggestions and complete the records.  Sounds great, right?

The reality is not so simple.  It takes time for an NLP engine to learn from usage over time what it should and should not suggest as valid diagnosis and procedure codes.  The NLP is only as skilled as the coder that it learns from, so in many cases the CAC pulls in codes that are not valid or do not follow coding guidelines, adding a lot of codes for the coder to read through and validate.  More codes means it take more time to complete a record and no CPH improvement.  Each organization must build documents to flow into the CAC program so that everything a coder needs is in one place.  Efficiency can be lost if coders are missing documentation in the CAC and have to go back to the EHR to find the other necessary data to code the encounter.  Rather than streamlining the process, the CAC can become an additional step if not set up well.  Solely adding a CAC should not give the false hope that coding productivity will automatically increase for all chart types. 

In addition to efficient CAC set up at their facility, managers must remember that length of stay, dollar value, missing documentation, and incomplete records also play a part in the coder’s ability to be productive. A CAC system will not overcome the other these other factors that can impact a coder’s CPH.  When creating production goals managers must take into consideration the full work flow the coder has to complete in order to finish coding a record.

Jayme Greer, RHIT, CCS-P