Abstract Rationale The use of continuous neuromuscular blockade during invasive mechanical ventilation (IMV) is associated with prolonged IMV time and increased in-hospital mortality. Language barriers may impair assessment and management of agitation and delirium, which can cause ventilator dyssynchrony and lead to increased paralytic use. However, the relationship between language preference and paralytic use remains unknown. Methods We analyzed data from the Common Longitudinal ICU data Format (CLIF) consortium where each participating health system analyzed local electronic health record (EHR) data and shared aggregated results. We identified adult patients admitted to an intensive care unit (ICU) who received at least 24 hours of IMV between July 1, 2017 and July 30, 2025, excluding patients with a language preference other than English or Spanish. The primary outcome was initiation of a continuous infusion of paralytic medication during IMV. Secondary outcomes included total IMV time and thirty-day mortality. The exposure was the patient’s EHR-documented preferred language. Site-specific logistic regression models estimated the association between language preference and outcomes, adjusting for age, sex, race, BMI, LAPS-2, Elixhauser comorbidity score, Social Vulnerability Metric, and fixed effects for hospital and ICU type. A sensitivity analysis was conducted to adjust for COVID status. Random effects meta-analysis was conducted to aggregate results across health systems. Results The cohort included 66,099 mechanically ventilated adults (mean age 60.8 years SD 16.2, 27,970 42% female, 20,954 32% Black, and 4,492 7% Hispanic of any race) across 9 health systems and 40 hospitals. There were 63,810 (96%) patients with an EHR-documented English-language preference and 2,289 (4%) with a Spanish-language preference. Patients with Spanish-language preference had increased odds (OR 1.65, 95% CI 1.28 - 2.14) of receiving continuous neuromuscular blockade compared with English-language preference, with substantial variation across sites (Figure 1). This aggregate effect was consistent even after adjustment for COVID status (OR 1.49, 95% CI 1.23 - 1.81). Spanish-language preference was associated with increased IMV duration (27.2 hours, 95% CI 14.2 - 40.2) but not with 30-day mortality (OR 1.12, 95% CI 0.99 - 1.27). Conclusions Patients with a Spanish-language preference are more likely to receive continuous paralytics compared to patients with an English-language preference. The significant heterogeneity across health systems suggests further research is needed to identify mechanisms through which language influences paralytic treatment practices and to mitigate related disparities. This abstract is funded by: Dr. Ortiz is supported by research training grant NIH/T32-HL-007891. Dr. Chesley is supported by NHLBI K01HL171466. Dr. Reilly is supported by NHLBI R01-HL155159. Dr. Weissman is supported by R35GM155262. Dr. Rojas is supported by NIH/NIDA R01DA051464 and the Robert Wood Johnson Foundation and has received consulting fees from Truveta. Dr. Gao is supported by NIH/NHLBI K23HL169815, a Parker B. Francis Opportunity Award, and an American Thoracic Society Unrestricted Grant. Dr. Barker is supported by research training grant NIH/T32-HL-007749. Dr. Hochberg is supported by NIH/NHLBI K23HL169743. Dr. Ingraham is supported by NIH/NHLBI K23HL166783. Dr. Parker is supported by NIH K08HL150291 and R01LM014263 and the Greenwall Foundation. Dr. Lyons is supported by NIH/NCI K08CA270383. Dr. Bhavani is supported by NIH/NIGMS K23GM144867. Dr. Sarma is supported by NIH K23HL163491. The other authors have no conflicts of interest to disclose.
Ortiz et al. (Fri,) studied this question.