If you have been hearing about computer-assisted coding - or CAC - from your Health Information Management (HIM) department, there's a good reason. At most hospitals, inpatient coders spend at least 70 percent of their time reading, navigating and assessing patient documentation to determine appropriate codes for billing and reimbursement. With the implementation of MS-DRGs for Medicare payment and the adoption of ICD-10, coders are expected to code with ever greater precision. At the same time, revenue pressures demand faster coding turnaround time as organizations work to streamline the claims process, reduce DNFB days and improve cash flow.

Given such pressures, HIM directors are actively searching for ways to boost coder productivity without sacrificing accuracy and compliance. One solution is through technology - specifically, CAC.

Understand your environment

Outpatient and inpatient coding environments have different demands. In an outpatient facility, the documents in a medical record are less complex, with a narrower range of codes and fewer codes per record than typical inpatient records. In this environment, codes can be generated automatically with little risk to their accuracy.

In an inpatient environment, however, the reverse tends to be true. The documentation is more complex to reflect a broader scope of treatment, which means a wider range and greater number of possible codes. Also, reimbursement values tend to be higher, making completeness and compliance more critical. In such an environment, look for CAC applications that reduce the complexity of the documentation and speed the assignment of codes.

These systems offer three different approaches to CAC, each appropriate to a particular need or environment.

One option is an auto-coding system that parses medical records using natural language processing (NLP) to capture the medical concepts in the text and assign codes. The intent of an autocoder is to eliminate the need for any human validation of the codes, making it appropriate for the processes typically found in some outpatient and physician coding environments.

An alternative CAC approach identifies terms and then displays or assigns codes, which are edited and validated according to coding logic and groupings. Such systems can be used to code concurrently at the point of care as reports are entered into the hospital information system. This approach offers efficiency gains, especially when used to code structured documents in an outpatient environment. In treatment settings with complex medical records, however, the time spent in validating the results can offset earlier efficiency gains.

A third option integrates the coding process with the coder's review of the documentation. Using NLP and expert logic, the system identifies terms and phrases in the documentation and automatically links those terms to coding paths in the coding software. On one screen, the coder can review key terms, view links to the documentation, approve proposed codes, and investigate related terms and/or secondary diagnoses. Integrating documentation review with the coding process offers productivity gains for coders, making it well-suited to the inpatient coding environment.

Handling interfaces

CAC systems should be able to annotate and parse text-based documents, as well as view images, including physician notes. This requires interfaces between the CAC system and other software applications that contribute to the documentation and coding data stream.

There are three options for accessing electronic documents from a CAC system: using existing interfaces, modifying what you have or creating new ones. Using the same vendor for both the CAC system and the encoding software helps minimize the number of interfaces required, resulting in reduced installation and maintenance costs.

Bridging people, processes

Careful consideration of staff and process issues will help your organization achieve efficiency gains while still maintaining coding accuracy, even as regulations change and new coding standards such as ICD-10 are implemented.

Mr. Mark is vice president of development at 3M Health Information Systems.