Acute Encephalopathy and the Coma Scale

This week, as we continue to explore querying for acute (metabolic or toxic) encephalopathy, I want to examine the Glasgow Coma Scale (GCS) as a clinical indicator for encephalopathy.

I know many consultants encourage this approach, but I hope to convince you that this is an outdated practice, and could contribute to clinical validation denials. As mentioned last week, there is not a widely accepted definition for metabolic or toxic encephalopathy. However, of all the definitions that I reviewed for my article, as confirmed by an artificial intelligence (AI)-driven search, I did not find any that incorporated the GCS.

Let’s begin with a discussion about the GSC. It is a universal tool used to quantify a patient’s level of consciousness on a scale from 3 to 15, wherein 15 represents normal consciousness and anything less than 8 indicates severe impairment, such as a coma or the inability to protect one’s airway. It is primarily used in patients who have suffered a traumatic brain injury. The GCS was designed to assess a patient’s level of consciousness, which is different from assessing if they are experiencing altered mental status, a component of acute encephalopathy.

The GCS is composed of three subscales: eye opening, verbal responsiveness, and motor responsiveness. A score of 13 to 15 indicates mild brain injury. Patients with a lower GCS score have an increased risk of mortality after a head injury.

• Eye opening is often the first sign of arousal. The scale includes the choices of eyes open never, to pain, to sound, or spontaneously. Most encephalopathic patients will have spontaneous eye openings or will open their eyes to sound. A lack of eye opening may occur in severe encephalopathy because it indicates coma or brainstem dysfunction, a very late sign of acute encephalopathy.

• The motor response subsection evaluates purposeful movement and brain-spinal cord integration. The scale includes choices of none (no motor response), extension, abnormal flexion, flexion withdrawal, localized pain, and obeys commands. It is the most predictive component for outcomes in patients with brain injury. Most encephalopathic patients will be able to obey commands or will exhibit spontaneous movement. Motor response, like eye opening, is an indicator of severe encephalopathy.

• Verbal response is the most sensitive to identifying acute encephalopathy without coma. It is used to assess cortical functions associated with language and cognition. It is assessed on a five-point scale ranging from inability to speak to being oriented. 

Assessing the best verbal response should be determined through conversation with the patient. The assessor should avoid questions that evaluate the patient’s long-term memory by asking personal questions like their name or date of birth. More appropriate questions to ask include the following:

• Can you tell me where you are?
• Can you tell me what season it is?
• Can you tell me why you are here?

A person who answers questions correctly but is confused in the conversation is not oriented. Additionally, the person must be able to respond appropriately in complete sentences to be oriented. Many studies report subjectivity in the scoring of the coma scale due to inadequate training on how to properly administer and score the assessment.

Acute encephalopathy is likely to only be reflected in the verbal subscale, resulting in a GCS score in the range of 13-14. Because this score is so close to 15, which represents normal consciousness, payors may challenge the clinical validity of acute encephalopathy if they lack understanding of its limitation for use in patients without a head injury. Due to this lack of sensitivity, including the coma score or the verbal subscale is unlikely to persuade an auditor that acute encephalopathy is clinically supported.

Between fiscal years 2016 and 2020, coding guidelines allowed the GCS to “be used in conjunction with traumatic brain injury codes, acute cerebrovascular disease or sequelae of cerebrovascular disease codes.” The GCS may also be used to assess the status of the central nervous system for other non-trauma conditions, such as monitoring patients in the intensive care unit (ICU) regardless of medical condition. This guidance likely contributed to its use as clinical evidence for encephalopathy.

Another contributor to the use of GCS as an indicator of acute encephalopathy is its inclusion within Sequential Organ Failure Assessment (SOFA) criteria for identifying severe sepsis/sepsis with organ failure. The GCS is included as a measure of altered mental status. There is such a high correlation between sepsis and encephalopathy that many AI-driven clinical documentation integrity (CDI) tools use clinical indicators of sepsis to predict encephalopathy.

A 2013 article, Understanding Brain Dysfunction in Sepsis (Sonneville et. al), states that “sepsis-associated encephalopathy is characterized by acute changes in mental status, cognition, alteration of sleep/wake cycle, disorientation, impaired attention, and/or disorganized thinking. Sometimes exaggerated motor activity with agitation, and/or hallucinations can be observed, and agitation and somnolence can occur alternatively.”

This description further explains why the GCS is of limited value even in the encephalopathic septic patient, since only those who are somnolent are likely to have a score below 14.  

The value of the GCS is in identifying and assessing the severity of acute encephalopathy, once it affects alertness and consciousness. Although earlier studies found one-third of patients with sepsis had a GCS of < 12, a 2024 article, SOFA in sepsis: with or without GCS (Wang, L. et al.), found that the use of SOFA without GCS did not affect the predictability of ICU mortality associated with sepsis.

The bottom line is that a GCS score of 15 does not exclude the diagnosis of acute encephalopathy. A patient can be awake with a high coma score, but still be profoundly confused (or in terms of coding, encephalopathic).Therefore, use of the GCS can be an adjunct to help monitor and communicate patient status, but it should not be considered a definitive clinical indicator.