Mortality in individuals with COPD on long-term home non-invasive ventilation

Mathew Cherian,Veronique Adam,Bryan Ross,Jean Bourbeau,Marta Kaminska
DOI: https://doi.org/10.1016/j.rmed.2023.107378
Abstract:Background: Real-world evidence regarding survival of patients with chronic obstructive pulmonary disease (COPD) using chronic non-invasive ventilation (NIV) is scarce. Research question: How do obesity and other factors relate to mortality in patients with COPD on chronic NIV? Study design: and Methods: We retrospectively analyzed data from COPD patients enrolled in a home ventilation program between 2014 and 2018. Survival was compared between obese and non-obese groups using the Kaplan-Meier method. Factors associated with mortality were identified using multivariable Cox proportional regression analyses with Least Absolute Selection and Shrinkage Operator (LASSO) regularization. Univariable analyses were also done stratified by obesity. Results: Median survival was 80.0 (95% CI: 71.0-NA) months among obese (n = 205) and 30.0 (95%CI: 19.0-42.0) months in non-obese (n = 61) patients. NIV adherence was high in both groups. Mortality was associated with male gender [HR 1.44], chronic opioids or benzodiazepines use [HR 1.07], home oxygen use [HR 1.82], fixed pressure mode of ventilation [HR 1.55], NIV inspiratory pressure [HR 1.05], and thoracic cancer [HR 1.27]; obesity [HR: 0.43], age [HR 0.99] and NIV expiratory pressure [HR 0.94] were associated with decreased mortality. In the obese, univariable analyses revealed that chest wall disease, thoracic cancer, home oxygen use, FEV1% predicted, and ventilation parameters were associated with mortality. In the non-obese, male gender and respiratory comorbidities were related to mortality. Interpretation: Obesity is associated with improved survival in COPD patients highly adherent to NIV. Other factors associated with mortality reflect disease severity and ventilator parameters, with differences between obese and non-obese patients.
What problem does this paper attempt to address?