Delusions and Hallucinations Are Associated With Greater Severity of Delirium

Background: The 3 core domains of delirium (cognitive, higher level thinking, circadian) do not include the less common noncore psychotic symptoms. However, psychosis might inform about perturbations of neural circuitry, outcomes, or suggest tailored clinical management. Objective: We assessed for the first time the relationships between psychosis and other characteristics of delirium in patients without confounders for delirium phenotype, such as dementia or antipsychotics treatment. Methods: Cross-sectional analysis of 366 adults with delirium per the Delirium Rating Scale Revised-98, whose items distinguish hallucinations and delusions from other types of misperceptions and abnormal thought content, assessed during the preceding 24 hours to capture symptom severity fluctuation. The relationship of psychosis with other delirium characteristics was assessed using bivariate comparisons and analysis of variance as appropriate for groups with no psychosis and any psychosis (hallucinations and/or delusions), and subgroups with only hallucinations, only delusions, or both. A discriminant logistic model assessed variables associated with presence of any psychotic features versus none. Results: Delirium with any psychotic features occurred in 44.5% (163 of 366). Of the 366, 119 (32.5%) had only hallucinations (Hall), 14 (3.8%) had only delusions (Del), and 30 (8.2%) had both (Both). In the psychotic group (n = 163), 73.0% were Hall, 8.6% Del, and 18.4% Both. All psychotic patient groupings had significantly greater delirium severity on the Delirium Rating Scale Revised-98. Delusions and hallucinations were discordant for occurring together. The discriminant model found increased odds of having psychosis as 3 symptom severities increased (visuospatial ability, thought process, and sleep-wake cycle) where these each represented a delirium core domain. The noncore symptom of lability of affect had high odds ratio for psychosis, while motor retardation reduced odds of psychosis in this model. Conclusions: Consistent with prior reports, psychosis occurred in less than half of delirious patients with delusions being infrequent, and an association with affective lability was found. We are the first to report that psychotic features are a marker for more severe delirium affecting all core domains. Given that previous functional magnetic resonance imaging research found a correlation between neural network dysconnectivity with greater severity of delirium, psychotic symptoms might be a clinical marker for greater underlying cerebral cortical neural circuitry dysfunction.