Paranoid, agitated, and manipulative

Psychiatry recommends withholding psychotropics, stopping CIWA, and using vital sign parameters along with objective signs of diaphoresis and tremors as indicators of alcohol withdrawal for lorazepam administration. Mrs. M receives IV haloperidol, 1 mg, once during her second day in the hospital for severe agitation, but this medication is discontinued because of concern about lowering her seizure threshold.⁷ After treatment with lactulose, her ammonia levels trend down to 33 μg/dL, but her altered mental state persists with significant deficits in attention and orientation.

The neurology service performs an EEG that shows no slow-wave, triphasic waves, or epileptiform activity, which likely would be present in delirium or seizures. See Figure 1 for an example of triphasic waves on an EEG and Figure 2 for Mrs. M's EEG results. Subsequent lumbar puncture, MRI, and a second EEG are unremarkable. By the fifth hospital day, Mrs. M is calm and her paranoia has subsided, but she still is confused and disoriented. Psychiatry orders a third EEG while she is in this confused state; it shows no pathologic process. Based on these examinations, neurology posits that Mrs. M is not encephalopathic.

Figure 1: Representative sample of triphasic waves

This EEG tracing is from a 54-year-old woman who underwent prolonged abdominal surgery for lysis of adhesions during which she suffered an intraoperative left subinsular stroke followed by nonconvulsive status epilepticus. The tracing demonstrates typical morphology with the positive sharp transient preceded and followed by smaller amplitude negative deflections. Symmetric, frontal predominance of findings seen is this tracing is common

Figure 2: Mrs. M’s EEG results

This is a representative tracing of Mrs. M’s 3 EEGs revealing an 8.5 to 9 Hz dominant alpha rhythm. There is superimposed frontally dominant beta fast activity, which is consistent with known administration of benzodiazepines

The authors’ observations

Mrs. M had repeated admissions for alcohol dependence and subsequent liver failure. Her recent hospitalization was complicated by a TIPS procedure done 1 month ago. The incidence of hepatic encephalopathy in patients undergoing TIPS is >30%, especially in the first month post-procedure, which raised suspicion that hepatic encephalopathy played a significant role in Mrs. M’s delirium.⁸

Because of frequent hospitalization, Mrs. M was well known to the internal medicine, neurology, and psychiatry teams, and each used different terms to describe her mental state. Internal medicine used the phrase “acute mental status change,” which covers a broad differential. Neurology used “encephalopathy,” which also is a general term. Psychiatry used “delirium,” which has narrower and more specific diagnostic criteria. Engel et al⁹ described the delirious patient as having “cerebral insufficiency” with universally abnormal EEG. Regardless of terminology, based on Mrs. M’s acute confusion, one would expect an abnormal EEG, but repeat EEGs were unremarkable.

Interpreting EEG

EEG is one of the few tools available for measuring acute changes in cerebral function, and an EEG slowing remains a hallmark in encephalopathic processes.^10,11 Initially, the 3 specialties agreed that Mrs. M’s presentation likely was caused by underlying medical issues or substances (alcohol or others). EEG can help recognize delirium, and, in some cases, elucidate the underlying cause.^10,12 It was surprising that Mrs. M’s EEGs were normal despite a clinical presentation of delirium. Because of the normal EEG findings, neurology leaned toward a primary psychiatric (“functional”) etiology as the cause of her delirium vs a general medical condition or alcohol withdrawal (“organic”).

A literature search in regards to sensitivity of EEG in delirium revealed conflicting statements and data. A standard textbook in neurology and psychiatry states that “a normal EEG virtually excludes a toxic-metabolic encephalopathy.”¹³ The American Psychiatric Association’s (APA) practice guidelines for delirium states: “The presence of EEG abnormalities has fairly good sensitivities for delirium (in one study, the sensitivity was found to be 75%), but the absence does not rule out the diagnosis; thus the EEG is no substitute for careful clinical observation.”⁶

At the beginning of Mrs. M’s care, in discussion with the neurology and internal medicine teams, we argued that Mrs. M was experiencing delirium despite her initial normal EEG. We did not expect that 2 subsequent EEGs would be normal, especially because the teams witnessed the final EEG being performed while Mrs. M was clinically evaluated and observed to be in a state of delirium.

OUTCOME: Cause still unknown

By the 6th day of hospitalization, Mrs. M’s vitals are normal and she remains hemodynamically stable. Differential diagnosis remains wide and unclear. The psychiatry team feels she could have atypical catatonia due to an underlying mood disorder. One hour after a trial of IV lorazepam, 1 mg, Mrs. M is more lucid and fully oriented, with MMSE of 28/30 (recall was 1/3), indicating normal cognition. During the exam, a psychiatry resident notes Mrs. M winks and feigns a yawn at the medical students and nurses in the room, displaying her boredom with the interview and simplicity of the mental status exam questions. Later that evening, Mrs. M exhibits bizarre sexual gestures toward male hospital staff, including licking a male nursing staff member’s hand.