(Q) Utilization of Antipsychotics for Delirium After Hospitalization for COVID-19 and Other Infections

Publication Number: 641

Background: Delirium, characterized by acute onset of disturbance of consciousness and cognition, is common for elderly patients hospitalized for infections. Antipsychotic medication (APM) is the most commonly used medication to treat delirium. It is recommended to deprescribe APM as soon as the delirium, typically a transient and short episode, resolve, but the utilization of APM following delirium due to infection-related hospitalizations remains unclear.

Objectives: To investigate the utilization of APM following a hospitalization for patients with COVID-19 and other infections.

Methods: This study used all available data from Optum® Clinformatics® Data Mart and included patients aged 65 and older with a new dispensing of an APM in Optum within 30 days of hospital discharge with an infection diagnosis. We excluded patients with chronic indications for APM (i.e., Schizophrenia and other psychotic disorders) in the preceding 180 days and used new initiation of APM as the proxy for having delirium based on a previous validation study. We assessed the median duration of use of the APM and cumulative incidence of discontinuation from Kaplan-Meier estimators and Cox Proportional Hazard survival model, stratified by typical vs. atypical APM. The discontinuation was defined by having a gap between dispensing greater than 15 days with this gap threshold as 7, 30, 45, 60 days in sensitivity analyses.

Results: Our study cohort consisted of 21,028 patients (mean age 78.74 years and 54.6% female, 596 with COVID-19, 7984 with non-COVID respiratory infections, and 20432 patients without COVID-19 infections). In the overall population, the median duration of use of APM was 14 days for typical APM, and 31 days for atypical APM. At 1, 3, 6 months, cumulative incidence of discontinuation was 88.2%, 94.5%, and 97.2% among typical APM users and 42.9%, 61.2%, and 73.0% among atypical APM users. The adjusted hazard ratio for discontinuing APMs comparing patients with COVID vs. patients without COVID was 0.89 (95%CI: 0.81-0.98, p = 0.0198). Significant positive predictors of discontinuation included younger age, non-black race , antibiotics, anticoagulants, cancer, anemia, liver disease, etc. Significant negative predictors included heart disease, falls, stroke, etc. Sensitivity analyses shows the relation of gap definitions to the results from Cox model.

Conclusions: The mean duration of APM for delirium following infection-related hospitalization is longer than expected considering the typical clinical course of delirium. Patients with a new prescription of typical APM had a significant higher discontinuation rate compared to atypical APM. The hazard of APM discontinuation following COVID-19 related hospitalization was lower comparing to hospitalizations due to other infections.