How much is too much? A medication use evaluation of VA ICU sedation practice during the COVID-19 pandemic

by Ian C. Murphy, Kelly Bryan, Muriel Burk, Rong Jiang, Francesca Cunningham, Sarah Providence, Elizabeth Rightnour, Sarah Zavala, Kathleen Morneau, Trisha Exline, Stacey Rice, Travis Schmitt, Kelly Drumright, Jennifer Lee, BreAnna Davids, Tram Guilbeault, Brooke Klenosky, Ann-Marie Sutherland, Abbie Rosen, Lauren Ratliff, Kenneth Bukowski, Margaret A. Pisani, Andrew Franck, Mark Wong, Preston Witcher, Kathleen M. Akgün

OBJECTIVES

Early data suggested higher sedative requirements for ventilated COVID+ patients, deviating from established guidelines. We assessed the relationship between sedative use and outcomes in mechanically ventilated Veterans during the COVID-19 pandemic.

Design

Retrospective Medication Use Evaluation

Setting

National Sample of 13 Distinct VA Medical Center Intensive Care Units

Patients

Critically ill Veteran patients requiring mechanically ventilation for ≥2 days

Interventions

None.

Measurements and main results

The proportion of patients receiving fentanyl, midazolam and propofol was higher during COVID years. Compared with pre-COVID, median fentanyl dose was higher during Years 1 and 2 (1575mcg [(IQR) 1000–1650] vs. 1900 [1250–3000] vs. 1910 [1150–3500]). Adjuvant antipsychotics use was relatively low but tended to increase over time (pre = 10.5% vs. Year 1 = 12.3% vs. Year 2 = 14.1%). Most patients started on antipsychotics in the ICU were continued on the drug after extubation. Mortality was higher during COVID years (pre = 26.9% vs. 1 = 36.8% and 2 = 35.9%). In stratified analyses by COVID status years 1–2 (n = 79, 27%), a higher proportion of COVID+ patients received fentanyl (96% vs. 84%) and propofol (90% vs. 77%) and at higher doses (fentanyl = 1650mcg vs. 2688mcg median cumulative dose; propofol maximum infusion rate = 30 mc/kg/min (20–50) vs. 40 (25–50)). Sedative doses were similar to pre-COVID among non-COVID patients. Anti-psychotics were more frequently continued post extubation among COVID+ (34.6% vs. non-COVID+=14.9%). COVID+ patients were also less likely to have awakening and breathing trials at 48 hours after intubation (18% vs. 46%).

Conclusions

Sedative use and dosing increased during the first two years of COVID compared to pre-COVID, especially for COVID+ patients. The sustained elevated levels of fentanyl use in Year 2 suggests possible ‘therapeutic creep’ away from guideline-concordant practices for COVID+ patients. Antipsychotic prescription during intubation and following extubation was also more common among COVID + . These findings could inform development and implementation of safer sedation practices across VA ICUs during respiratory pandemics.