A retrospective national cohort study of trends in mechanical ventilation among Veterans living with dementia, 2010-2019

Abstract: BACKGROUND: Despite longstanding concern about an increase in use of invasive mechanical ventilation among persons living with dementia (PLWD), no studies have examined trends in mechanical ventilation use among PLWD in Veterans Affairs (VA) facilities. In this study, we aimed to (1) identify recent trends in use of mechanical ventilation among Veteran PLWD and (2) assess mortality trends of those who received mechanical ventilation. METHODS: In this retrospective national cohort study of all VA medical hospitalizations of Veteran PLWD ≥ 65 years from 2010 to 2019, we used data from the VA Corporate Data Warehouse and defined dementia using a VA-sanctioned list of dementia diagnosis codes. We calculated the percentage of hospitalizations with mechanical ventilation use during the study period and used linear regression to determine a temporal trend. We calculated in-hospital and one-year mortality for hospitalizations of Veteran PLWD involving mechanical ventilation and used linear regression (predictor: time; outcome: mortality) to describe mortality trends. RESULTS: Our cohort included 702,989 hospitalizations at 126 VA medical centers involving 251,545 unique Veteran PLWD. Hospitalized Veteran PLWD were 97.9% male, 89.7% non-Hispanic/Latino, and 72.7% White. Mechanical ventilation use decreased from 1.7% of hospitalizations in 2010 to 1.1% in 2019. Annual in-hospital mortality among those PLWD who received mechanical ventilation decreased from 45.9% in 2010 to 38.0% in 2019 and one-year mortality decreased from 73.4% in 2010 to 70.2% in 2018. CONCLUSIONS: The use of mechanical ventilation in hospitalizations of Veteran PLWD was lower than seen in non-VA facilities and decreased from 2010 to 2019. Among hospitalized Veteran PLWD who received mechanical ventilation, in-hospital and one-year mortality remained high throughout the study period but decreased over time. These descriptive mortality decreases may be attributable to patient selection or improved mechanical ventilation practices. Further research should examine patient- and system-level factors to explain observed trends.