Impact of ARDS on Post-ICU Psychoactive Medication Prescribing Without a Justifying Diagnosis

Introduction: Psychoactive medication prescribing (i.e., anxiolytics, antidepressants, antipsychotics, or sleep aids) to ICU survivors in the absence of a diagnosis justifying prescribing may be greater in acute respiratory distress syndrome (ARDS) patients. We compared psychoactive medication prescribing in the absence of a justifying diagnosis up to two years after discharge from an ICU admission requiring mechanical ventilation (MV) between ARDS and non-ARDS groups.


Methods: This retrospective cohort study analyzed data from the U.S. claims database PharMetrics Plus for Academics for adults admitted to an ICU requiring MV between 1/1/18 and 12/31/19. We excluded patients with a psychoactive medication prescription, a justifying diagnosis for one, or another ICU admission in the year prior. Patients with no observation after the hospitalization end-date were also excluded. The cohort was categorized into ARDS and non-ARDS groups if the ICD-10 code J80 was documented during hospitalization. We applied two different propensity models with a 1:4 matching ratio to balance ARDS and non-ARDS group baseline characteristics: a large-scale propensity score (LSPS) model incorporating >50,000 covariates and a conventional propensity score (PS) model using 7 covariates. Covariates were selected a-priori. Univariate Cox proportional hazards models, conditioned on the matched sets, were used to compute the hazard ratios (HR) for post-hospital psychoactive medication prescribing in the absence of a justifying diagnosis between the two groups. Subgroup analyses were conducted using LSPS models.


Results: ARDS was diagnosed in 404 (5.1%) of the 7,898 eligible patients. No significant difference was found between the ARDS (versus non-ARDS) groups for psychoactive medication prescribing in the absence of a justifying diagnosis using either the LSPS (HR 0.72 (95% CI 0.42-1.17) or PS (HR 0.85 (95% CI 0.51-1.42) models. Patients with an acute neurological condition had a lower risk for psychoactive medication prescribing in the absence of a justifying diagnosis [HR 0.19 (95% CI 0.01-0.91)].


Conclusions: The rate of psychoactive medication prescribing in the absence of a justifying diagnosis after an ICU hospitalization is not significantly different between ARDS and non-ARDS patients.