• Home
    • » e-Newsletters

Correctly code encephalopathy with seizures/CVA

JustCoding News: Inpatient, March 25, 2015

Want to receive articles like this one in your inbox? Subscribe to JustCoding News: Inpatient!

By James S. Kennedy, MD, CCS, CDIP

The physician documented “encephalopathy” in the progress note of a patient who was admitted with a cerebrovascular accident (CVA) and/or possible seizures. The patient was confused but has returned to baseline.   

The coding and clinical documentation improvement (CDI) teams are conflicted about how to report this. Is the documented encephalopathy integral to the CVA or seizure and thus not to be coded? Is the encephalopathy not integral to these conditions and so can be coded on hospital admissions and reported as an MCC with MS-DRGs? Is it okay to code the documented term “encephalopathy” without a query for more specificity or to determine its underlying cause?

Before addressing whether a documented diagnosis of encephalopathy can be coded (and thus affect MS-DRG or APR-DRG assignment), let’s first clarify several concepts related to ICD-9-CM or ICD-10-CM code assignment and how the CDI process should work.

  • Principal diagnosis: As you know, according to the ICD-9-CM or ICD-10-CM guidelines and the Uniform Hospital Discharge Data Set (UHDDS), the principal diagnosis is defined as:

That condition established after study (emphasis added) to be chiefly responsible (emphasis added) for occasioning the admission of the patient to the hospital for care.

This determination is based on information available at the time of admission, which is when the physician writes an inpatient order, and is governed by circumstances of admission, the diagnostic approach, the treatment rendered, and ICD-9-CM/ICD-10-CM coding conventions.   The ICD-9-CM and ICD-10-CM Official Guidelines for Coding and Reporting have clarifying rules governing principal diagnosis assignment which, incidentally, are not the same. Both sets of the Official Guidelines for Coding and Reporting emphasize the importance of consistent, complete documentation in the medical record.  Without such documentation, the application of all coding guidelines is a difficult, if not impossible, task.

  • Secondary, additional, or other diagnoses: According to the ICD-9-CM and ICD-10-CM Official Guidelines for Coding and Reporting, assigning a code for additional conditions that affect patient care require that documented condition require at least one of the following:
    • Clinical evaluation
    • Therapeutic treatment
    • Diagnostic procedures
    • Extended length of hospital stay
    • Increased nursing care and/or monitoring

The UHDDS item 11-b defines other diagnoses as:

All conditions that coexist at the time of admission, that develop subsequently, or that affect the treatment received and/or the length of stay.

  • CDI: While many definitions exist, CDI is usually defined as the policy, process, and procedure that assures the clinical integrity of ICD-9-CM or ICD-10-CM code assignment based on provider documentation, usually executed by the rendering of a provider query. CDI requires rigorous adherence to the ICD-9-CM or ICD-10-CM conventions, Official Guidelines for Coding and Reporting, official advice from Coding Clinic, and is exercised when provider documentation:
    •  Is conflicting, imprecise, incomplete, illegible, ambiguous, or inconsistent
    • Describes (or is associated with) clinical indicators without a definitive relationship to an underlying diagnosis
    • Includes clinical indicators, diagnostic evaluation, and/or treatment not related to a specific condition or procedure
    • Provides a diagnosis without underlying clinical validation
    •  Is unclear for present on admission indicator assignment

In discussing the coding of encephalopathy, we need to clear up five questions.

1. Note that in the clinical circumstance described above, there is no adjective before the word “encephalopathy,” nor any documented indication of its underlying cause, given that the physician did not link the word “encephalopathy” with any of the other documented conditions.

If we look at the ICD-9-CM or ICD-10-CM Index to Diseases, there are approximately 100 adjectives or statements of the underlying cause of encephalopathy that require documentation if we are to code to the highest level of specificity. Some of these affect the DRG assignment, such as toxic or metabolic encephalopathy or encephalopathy due to drugs; some of these do not. Ideally, the unspecified encephalopathy code should be used only when the physician does not know the etiology of the patient’s encephalopathy.

Therefore, the first thing to do is to query the physician (in a non-leading way) to determine the underlying cause of or the nature of the encephalopathy. Refer to the 2013 ACDIS/AHIMA query practice brief Guidelines for Achieving a Compliant Query Practice for an industry standard in how to accomplish this. 

2.  Is the term “encephalopathy” consistently documented and to what extent did it persist, worsen, or resolve? While there are no published standards, the more often terms are documented (e.g., three times or more, especially with comments regarding its improvement or resolution), the less likely a Recovery Auditor or accountability agent is able to remove an ICD-9-CM or ICD-10-CM code from the bill, especially if it affects reimbursement. 

3.  Should the condition “possible seizure” be coded? In the opening case, we don’t have enough   information related to whether the physician documented the term “possible seizure” in the discharge summary.

Many coders and CDI specialists are unclear in how the ICD-9-CM or ICD-10-CM guidelines govern the coding of uncertain diagnoses. Many do not know that the ICD-9-CM and ICD-10-CM guidelines explicitly state that “If the diagnosis documented at the time of discharge (emphasis added) is qualified as ‘probable,’ ‘suspected,’ ‘likely,’ ‘questionable,’ ‘possible,’ or ‘still to be ruled out,’ ” or other similar terms indicating uncertainty, code the condition as if it existed or was established” for inpatient facility admissions, not physician or outpatient or observation services. The basis for this guideline is the diagnostic workup, arrangements for further workup or observation, and initial therapeutic approach that correspond most closely with the established diagnosis. Coding Clinic for ICD-9-CM, Third Quarter 2005, p. 22, emphasizes this same principle by stating that this guideline does not apply to admitting or interim diagnoses. 

As such, if it's unclear whether “possible seizure” was documented at the time of discharge, a query would be necessary to determine if the diagnosis was still valid “at the time of discharge.” If the provider affirms the presence of the “possible seizure” at the time of discharge, we also need to know what the underlying cause of the seizure is (i.e., the current CVA, the late effect of an old CVA, or another cause), whether it is part of a recurrent seizure disorder (i.e., epilepsy), and, if clinically indicated, if it is status epilepticus or part of an intractable or poorly controlled seizure disorder.

4. What is the principal diagnosis for this admission?

The ICD-9-CM Official Guidelines for Coding and Reporting state:

When two or more diagnoses equally meet the criteria for principal diagnosis as determined by the circumstances of admission, diagnostic workup and/or therapy provided, and the Alphabetic Index, Tabular List, or another coding guidelines does not provide sequencing direction, any one of the diagnoses may be sequenced first.

As such, we have a seizure (which is a symptom code in ICD-9-CM or ICD-10-CM), a CVA (which is a diagnosis code), and the yet-to-be-determined encephalopathy as potential candidates. My hunch is that the CVA (stroke) will likely be the principal diagnosis, given that this is probably the underlying cause of the “possible seizure” and more than likely required the brunt of the diagnostic workup and treatment rendered. That’s not to say that provider documentation could not provide alternatives; however explicit documentation at the time of discharge would be necessary to amend this impression. 

5. Let’s say that the physician documented at the time of discharge that the “possible seizure” was due to the CVA but stated that the encephalopathy was due to the seizure and due to the stroke. We’ve determined that the CVA is the principal diagnosis. How should a coder or CDI specialist handle the documentation of the term “encephalopathy” in this situation?

This scenario raises a number of concerns.

  • Coding of encephalopathy due to a seizure: Fortunately, we have clear official advice on this issue from Coding Clinic for ICD-9-CM, Fourth Quarter 2013, which states:

“On admission the patient had mental status changes, which subsequently resolved. Consequently, we determined the patient had encephalopathy secondary to postictal state. Should encephalopathy be reported as an additional diagnosis with seizure when it is due to a postictal state? Would encephalopathy be considered inherent to the seizure or can it be reported separately?

Their answer was:

Encephalopathy due to postictal state is not coded separately since it is integral to the condition...The postictal state is a transient deficit, occurring between the end of an epileptic seizure and the patient’s return to baseline. This period of decreased functioning in the postictal period usually last less than 48 hours.

With this in mind, we do not add an additional code for encephalopathy because it is due to a seizure.

My personal belief is that a stroke is a focal brain disease and, as such, the term “encephalopathy” is integral to stroke. As such, I would not code it or, alternatively, I would discuss with the documenting provider as to how he or she believes this should be coded and base my coding upon the documented answer.

I also believe that the term “encephalopathy” is integral to other defined brain conditions besides seizures, such as Alzheimer’s disease, Lewy body dementia, Parkinson’s disease, and the like, and thus a non-specified encephalopathy should not be coded if documented to be due to the underlying brain disease. That’s not to say that patients with these underlying brain diseases cannot have another defined encephalopathy (e.g., toxic, metabolic, anoxic, hepatic, hypertensive) supposed on their pre-existing brain condition; however that would have to be clearly documented by the provider to be coded as such.   If the coder or CDI specialist is unclear, then a query is warranted. 

In summary, CDI specialists and coders need to consider how documented conditions are defined, identify the essential components of the documented conditions, and how ICD-9-CM or ICD-10-CM coding conventions, Official Guidelines for Coding and Reporting, or official Coding Clinic advice address these conditions. Sadly, the ICD-9-CM (soon-to-be ICD-10-CM/PCS) conventions do not always define what conditions are “integral” within the code set nor govern how Recovery Auditors may view the issue. Wisdom, logical thinking, and outside support are needed to navigate these treacherous waters. 

When considering if a diagnosis is integral to a condition, I typically ask myself whether most patients with this condition experience this symptom or diagnosis or what the literature may say about it. No matter what I think, however, the documenting provider has to make the call and then I have to determine if I can defend his or her answer if challenged by an auditor or other accountability agent. 

Bottom line: When in doubt, query the provider to further clarify the situation. As such, for this circumstance, I would query the following issues:

  • Whether the uncertain diagnosis of “possible seizure” should be coded
  • What the nature, severity, repetitiveness, and underlying cause of the seizure is likely to be
  • What adjective best describes the documented encephalopathy or what the underlying cause of the encephalopathy is
  • Whether or not the documented encephalopathy is integral to the CVA or seizure

The final coding of this record depends upon provider documentation and the coder’s comfort in assigning the code, which is an entirely different discussion we need to have, since it is the coder’s initials that go on the coding sheet, not the CDI specialist’s.

Editor’s note: James S. Kennedy, MD, CCS, CDIP, is president and chief medical officer of CDIMD -Physician Champions in Smyrna, Tennessee. Contact him at JKennedyMD@cdimd.com.This article arose from a Q&A on the ACDIS website.


Want to receive articles like this one in your inbox? Subscribe to JustCoding News: Inpatient!