Do health information management (HIM) and clinical documentation integrity (CDI) professionals need a standard operational definition for “clinical documentation?” Put another way, should everything within a physician note, for example, be considered clinical documentation?
Back in the day of paper records, it was easy to distinguish between a template header, a prompt, and physician documentation, because the template was preprinted and the physician documentation was handwritten or transcribed. With movement towards the electronic medical record (EMR), it is more difficult to differentiate what the provider entered into the record from what is part of a template, from what was “pulled forward” by the provider, from what was auto-populated by the EMR, etc.
Why does this matter? Well, I review a lot of records in my role, and I am seeing a lot of cases in which words or phrases within the health record are being used to report an associated diagnosis; however, Official Coding Guidelines for reporting “other diagnoses” are not met. Specifically, the Coding Guidelines state:
For reporting purposes, the definition for “other diagnoses” is interpreted as additional conditions that affect patient care in terms of requiring:
- Clinical evaluation;
- Therapeutic treatment;
- Diagnostic procedures;
- Extended length of hospital stay; or
- Increased nursing care and/or monitoring.
Not only is technology changing how the medical record is formatted, but it is also changing how CDI and coding professionals perform their duties. A 2015 article published by the American Health Information Management Association (AHIMA, Weinberg, J, et. al) defined computer-assisted coding (CAC) as the use of computer software that automatically generates a set of medical codes for review and validation, based upon the clinical documentation of healthcare practitioners. Furthermore, “CAC includes a variety of computer-based approaches that do not require human interaction to transform narrative text in clinical records into structured text, which may include assignment of codes from standard terminologies such as ICD-9-CM, ICD-10-CM/PCS, CPT/HCPCS, and SNOMED CT.”. However, the article also warns,
“CAC requires a very high level of data integrity. This is due to the inherent nature of natural language processing (NLP) engines. These engines utilize a lexicon to determine if documentation meets criteria to be assigned a final code. If the CAC engine cannot understand a term, concepts are not completely documented, or terms are spelled incorrectly, then the engine may not recognize the term and assign a code accordingly.”
As someone who works in the technology field, I must say that my personal experience is that few CAC tools are robust enough to actually consider the context of the documentation to see if it meets criteria for code assignment. Yes, CAC tools can identify when a term like “shock” appears, for example, and map it to the associated unspecified code for shock, but not all of them are able to realize that the word was highlighted as part of the phrase “shock index,” which is a header within the template to support clinical assessment. In other words, in this context “shock” isn’t even a documented diagnosis. It is simply a header within the health record prompting the provider to complete a comprehensive patient assessment. Depending on the sophistication of the NLP engine, some are able to determine the context of a word as a positive or a negative mention. For example, “no heart failure” would be a negative mention, but what about when the assessment of heart failure is part of a template, so it appears as “heart failure: negative” or “heart failure: absent,” or any other number of variations? Not all NLP engines are able to process terms like a historical mention of a condition, or when it references a family member (or when it is uncertain).
There are also many EMRs that can import ICD-10-CM Codes or SNOMED CT codes with an associated code title as the physician enters data. In some records, the only reference to a particular diagnosis may be the code title. Is this really clinical documentation? I know there has been and continues to be much debate about problem lists, but what about code titles? Keep in mind that the Official Coding Guidelines state, “the assignment of a diagnosis code is based on the provider’s diagnostic statement that the condition exists.” A code title is not the same as a diagnostic statement, and it doesn’t support the condition as reportable.
Another situation to consider is related to the ability to copy notes within the EMR. I have reviewed many records in which documentation from the history and physical (H&P) is copied forward into the discharge summary. Although this may be a time-saver for the provider, it is problematic when the H&P states that a patient is admitted for “possible pneumonia” or “suspected sepsis” – or any other condition that is, understandably, uncertain at the time of admission. However, when this documentation is copied into the discharge summary, many coders erroneously invoke the Official Coding Guidelines regarding the reporting of uncertain diagnoses:
Yes, the discharge summary contains the words “possible pneumonia;” however, the totality of the record usually demonstrates that the pneumonia (or other possible conditions) was ruled out. Again, the diagnosis likely does not meet reporting guidelines. I find that these cases are particularly prevalent within the newborn population, as many within this patient population are admitted for a suspected condition or in order to rule out a condition.
So, let’s get back to basics, and make sure our CDI and coding teams are taking the time to validate terms within the health record to confirm them as clinical documentation that reflects a “diagnostic statement.” Let’s also reinforce the need to meet Official Coding Guidelines for reporting “other diagnoses.”
Not all words that can be mapped within a health record to a diagnosis code are clinical documentation or reportable.