(C-007) Long-Term Incidence of Psychoactive Medication Prescribing and Psychiatric Condition Occurrence Following ICU Admission with Mechanical Ventilation

Background: The diagnosis of new psychiatric conditions (PCs) and the prescription of new psychoactive medications (PMs) are common but remain poorly characterized. Existing post-ICU population studies are limited by the number of PCs and PMs evaluated, a failure to consider pre-ICU histories of PC occurrence or PM prescribing, and a focus on non-U.S. populations.

Objectives: To characterize the cumulative incidence of PM prescribing and PC occurrence in the 2 years following an ICU hospitalization in critically ill Americans.

Methods: In this retrospective cohort study using the PharMetrics Plus for Academics claims database, we included U.S. adults with an ICU admission requiring mechanical ventilation (i.e., index hospitalization) from 1/2018 to 12/2019. We excluded patients with an ICU or long-term care admission in the prior year, whose insurance ended during the index hospitalization, or who were discharged to hospice. We followed patients from the end of index hospitalization to the earliest of either insurance disenrollment or 2 years. We evaluated the cumulative incidence of post-ICU PM prescribing (anxiolytics, antidepressants, antipsychotics or sleep aids) and PC diagnosis (anxiety, depression, stress-related disorders or sleep disorders) over up to 2 years. We also evaluated the cumulative incidence of individual PM and PC, time to first incident PM and PC, and PM persistence. Descriptive analyses were used for all other outcomes. Cumulative incidence was described using Kaplan-Meier curves.

Results: Among the 11,822 patients (median [IQR] age 70 [59–79], Charlson Comorbidity Index 5 [3–7], Hospital Frailty Risk Score 9 [6–14], medical admission 76%), the cumulative incidence of post-ICU PM prescribing at 2 years was 30% (95% confidence interval [CI] 29–32%)—anxiolytics 17% (16–18%), antidepressants 13% (12–14%), antipsychotics 8% (7–8%), and sleep aids 2% (2–3%). The cumulative incidence of a PC diagnosis was 42% (95%CI 40–43%)—depression 20% (19–21%), anxiety 19% (18–20%), sleep disorder 14% (13–15%), and stress-related disorder 10% (10–11%). Among patients who had any PM or PC, a quarter (25%) of patients were prescribed ≥2 different PMs, and more than one-third (35%) were diagnosed with ≥2 PCs. Among the PMs and PCs considered, second-generation antipsychotics (median [IQR] 43 [8–165] days) and anxiety (50 [9–201] days) were prescribed or diagnosed the soonest after the index ICU hospitalization. Antidepressants were prescribed for the greatest proportion of post-ICU days (30% [95%CI 12–66%]).

Conclusions: American critically ill adults experience a high incidence of post-ICU PM prescribing and PC occurrence, highlighting the importance of early and regular patient assessment in the months after ICU discharge.