(1002) Analgosedative Comparative Use in COVID Versus Non-COVID-Associated ARDS Pneumonia

Introduction: Medications for analgesia and sedation are used in critically ill mechanically ventilated (MV) patients with acute respiratory distress syndrome (ARDS) for comfort. Some studies have shown increased analgosedative use in COVID vs non-COVID ARDS. This study aims to compare analgosedative use in critically ill patients with COVID vs non-COVID ARDS pneumonia and its association on outcomes.

Methods: This was a retrospective cohort study of adults requiring MV for ARDS pneumonia admitted to 22 hospitals within Intermountain Health between 1/1/2019 and 12/31/2021. Excluded patients were less than 18 years, pregnant, admitted for trauma, or had seizures or alcohol withdrawal during hospitalization. We compared the use of propofol, opioids (fentanyl mcg equivalents), benzodiazepines (BZD) (lorazepam mg equivalents), dexmedetomidine, and ketamine between COVID vs non-COVID ARDS. Using multivariable logistic regression, we evaluated the association of these on mortality, ICU length of stay (LOS), delirium-free days, and MV-free days.

Results: 3,476 patients were analyzed, (1,097 COVID, 2,379 non-COVID ARDS). Patients were 61.2 years (SD 15.5), had mean APACHE III score of 71.6 (33.1), initial worst PaO2/FiO2 of 102.6 (IQR 90.7), with median MV duration and ICU LOS of 6.7 (11.3) and 7.8 (10.8) days, respectively. Non-COVID patients had significantly higher APACHE III and Elixhauser scores and were older (all P < 0.01). In-hospital mortality was 33.1% (42.8% COVID vs 28.7% non-COVID; P < 0.01). Upon multivariable logistic regression, COVID and propofol were associated with increased mortality (P < 0.01); opioids, propofol, and dexmedetomidine (all P< 0.01), but not COVID (P=0.13) were associated with increased ICU LOS; opioids, propofol, and dexmedetomidine (all P< 0.01), but not COVID (P=0.23), were associated with fewer delirium-free days; and propofol (P < 0.01), but not COVID (P=0.36) was associated with fewer MV-free days.

Conclusions: Some analgosedatives were associated with increased mortality and ICU LOS and fewer delirium- and MV-free days, whereas COVID ARDS was associated only with mortality. Future studies should evaluate the relative contribution of underlying disease vs analgosedative use on patient important outcomes as well as optimal analgosedative strategies in patients with COVID ARDS.