COVID-19-Associated Persistent Psychosis

Over the past few months, I have come across case studies that have described psychosis and delirium as potential symptoms of COVID-19, but none have described persistent psychosis after infection with COVID-19. This week, a report in The British Journal of Psychiatry describes such a case in the UK (Lim, et al., 2020).

The patient was a 55-year-old white female. She had a 14 day history of fever, myalgia, breathlessness, cough, anosmia, ageusia, and headache. Her medical history had nothing abnormal besides resolved renal calculi (kidney stones) and she was not on any medications. She had no history of mental or neurological illness and there was also no family history of such illness. On admission, chest x-ray revealed bilateral lower zone and peripheral pulmonary infiltrates. Chest CT revealed bilateral subpleural ground-glass opacities and interlobular septal thickening consistent with COVID-19. After receiving IV fluids and oxygen via nasal cannula, she was discharged two days after admission. The next day, she was “confused with odd behavior” and was admitted to the hospital via ambulance. Upon admission, she had visual hallucinations that involved seeing animals that were not there. She had automatisms and tremors, and was suspected to have possible seizures. Her oxygen saturation was 98% on room air. She then became acutely agitated and aggressive towards the staff, “spitting and throwing items of clothing and swearing.” After being given IM lorazepam, she was admitted to a medical ward, where she was suspicious and anxious around the new staff. 

“She described paranoid delusions involving colour symbolism whereby she attributed the colour red to people who were trying to kill her. She believed that the nursing staff were ‘devils’, trying to harm hyer and a family member. She exhibited bizarre behaviour including washing her phone in the sink, and repetitively brushing her teeth with soapy water.”

Her agitation was managed by lorazepam. Her levels of confusion and agitation fluctuated during these first few days. Her disorientation was punctuated by moments of lucidity where she was able to communicate with her spouse over the phone. On day 5, her agitation worsened and she developed “persecutory delusions of a family member being murdered by staff.” She was started on 0.5 mg haloperidol one week into admission. Her sleep improved and she became less agitated. By day 10, she was admitted to a neuropsychiatric ward and had noticeably improved, but was still distracted, muttering to herself, and suspicious of stuff. She demonstrated a Capgras-like delusion (a delusion where a person believes that an identical imposter has replaced someone they know). She also developed auditory hallucinations.

Her early lab results revealed a positive SARS-CoV-2 swab test, elevated CRP, ferritin, D-dimer, and TNF-ɑ. During her psychosis, her TNF-ɑ remained elevated independent of any bacterial or viral infection. Head MRI revealed no abnormalities. EEG revealed no evidence of encephalopathy or seizures. She was switched from haloperidol to 0.5 mg risperidone on day 14. On day 20, she was discharged at her own request and her symptoms were improving. During a follow-up on day 29, she revealed that she had stopped taking risperidone on day 27. On the final follow-up on day 52, she had no cognitive impairment and her sleep cycle normalized. The only abnormalities her family noticed were increased anxiety around strangers outside the house. She improved after this point and had no psychotic symptoms.

The authors believe that the psychosis was related to COVID-19 because her psychotic symptoms were correlated with an inflammatory state (elevated TNF-ɑ). They also do not rule out that her psychosis may be antibody-mediated. However, antibodies typically associated with post-infection mental changes such as NMDAR, CASPR2 and LGI1 were negative. There are no clear answers in this case, but it adds a fascinating yet terrifying possibility to the many ways that COVID-19 manifests psychiatrically. Further research into how TNF-ɑ may mediate psychotic episodes is warranted, as IL-6 has been associated with altered mental states. Additionally, more research is needed on how viruses may be associated with antibody-mediated encephalopathy.