The fact that more than 80 percent of organizations staff clinical documentation improvement (CDI) professionals, despite it being a supplemental business function, is an indication of the undeniable benefits of CDI programs. Organizations have been slow, however, to invest in technology developed especially for CDI. But technology can increase coverage through the automation of manual tasks, such as worklist automation, case prioritization, and analytics, and also increase efficiency by identifying cases most likely to have significant documentation improvement opportunities.
As the Centers for Medicare and Medicaid Services (CMS) transforms its payment methodologies from a system that rewards the volume of care to one that risk-adjusts comparative outcome data and rewards quality and efficiency of care, and CDI efforts are expanding, technology can maximize existing staff's impact by eliminating inefficiencies in workflow and data management. In addition, leveraging technology to produce meaningful clinical data allows CDI professionals to focus their efforts on creating a culture where accurate documentation is considered a priority through educate and encouraging changes in behavior.
One type of technology that can assist CDI efforts is worklist automation that leverages an existing admit, discharge, and transfer (ADT) interface in conjunction with natural language processing (NLP). This can help CDI professionals quickly identify the target population as well as prioritize cases based on the potential impact of identified documentation opportunities. Identifying query opportunities based on unspecified codes demonstrates the basic power of an NLP engine. For example, if congestive heart failure (CHF) is documented and recognized by the NLP engine, a query for CHF specificity will be suggested. Depending on the sophistication of the CDI solution, the tool may also be able to "reason" or examine the totality of the health record to identify clinical indicators that support a query opportunity for a missing diagnosis, or to assist with clinical validation for a diagnosis that is already documented. In other words, the tool may be able to data mine indicators that allow it to infer an associated diagnosis like sepsis, even when a code associated with sepsis was not suggested. This approach is possible because the NLP can be trained to identify the relationship between discrete data elements that might otherwise be missed.
The computer assisted coding (CAC) element associated with some CDI solutions can support CDI and coding collaboration, as the focus shifts from identifying codes to validating diagnoses, their sequencing, and clinical validity. This is especially true if the solution links the associated documentation to each suggested code and allows both the CDI and coding professional to add or remove the associated documentation, creating transparency, and alignment. The relationship between the documentation and its associated codes also minimizes DRG differences, allowing both CDI and coding professionals to focus on accurately capturing the clinical indicators in the suggested codes.
In addition, another valuable outcome of CDI technology is improved traceability of the CDI process in line with the coding process. Overall, technology allows CDI professionals to focus on improving documentation and integrity of the medical record, while making the CDI process more transparent to those entrusted with coding the final encounter.