Supplemental oxygen use in older adults with COPD was associated with higher intensive care use (52.8% vs 39.5%, p=0.004) and life support use (42.0% vs 25.5%, p<0.001) at the end of life.
Cross-Sectional (n=878)
Does supplemental oxygen use in older adults with COPD associate with different end-of-life healthcare use or care preferences compared to those without oxygen?
In older adults with COPD, supplemental oxygen use is associated with higher end-of-life intensive care use and longer hospital stays, despite similar strong preferences to limit care.
Absolute Event Rate: 52.8% vs 39.5%
p-value: p=0.004
Abstract Supplemental oxygen use in COPD signals progression to advanced disease and limited prognosis; its relationship to end-of-life (EOL) healthcare use and care preferences is unknown. Do people with COPD using supplemental oxygen have different EOL healthcare use or care preferences compared to those without oxygen? We analyzed 2012-2018 Health and Retirement Survey data for decedents aged ≥50 with COPD and self-reported oxygen use. Proxies completed post-death interviews regarding healthcare use (hospitalization, intensive care, hospice, and palliative case use) in the last two years before death, and advance care planning (ACP including durable power of attorney, advance directives, and care preferences for EOL care) in the final days of life. We examined associations between oxygen use and intensity of healthcare use and ACP at the end of life using logistic regression (adjusted for demographics and comorbidities) and the model-predicted probability estimate of the outcome using the STATA margins command. Among 878 decedents (mean age 81, 53% female, 81% White), 48% used supplemental oxygen for a median of 2 years (range: 0-8) before death. Compared to those without oxygen, decedents who used oxygen had higher mean hospital days (16.4 vs 22.5 days, p = 0.04), intensive care use (52.8% vs 39.5%, p = 0.004) and use of life support (42.0% vs 25.5%, p 0.001), but similar hospital admissions (72.6% vs 69.9%, p = 0.38) and use of palliative care (24.9% vs 25%, p = 0.99) and hospice (47.0% vs 48.5%, p = 0.72), respectively. Compared to those without oxygen, decedents who used oxygen had similar rates (p 0.05) of durable power of attorney completion (no oxygen: 75.8% vs. oxygen: 73.9%), advance directives (55.6% vs. 55.6%), as well as strong preferences to limit care (89.0% vs. 87.4%) and focus on comfort (71.9% vs. 76.2%) rather than to pursue all measures to prolong life (9.5% vs. 8.5%). Our cross-sectional study found that decedents with COPD on oxygen therapy had higher end of life intensive care use and longer hospital stays compared to decedents without oxygen use, despite strong preferences to limit care and focus on comfort at the end of life. These results indicate oxygen use has important prognostic significance and clinicians should use this as an opportunity to address values and preferences. This abstract is funded by: NIH T32AG000212
Suen et al. (2026) conducted a cross-sectional in COPD (n=878). Supplemental oxygen use vs. No supplemental oxygen use was evaluated on Intensive care use in the last two years before death (p=0.004). Supplemental oxygen use in older adults with COPD was associated with higher intensive care use (52.8% vs 39.5%, p=0.004) and life support use (42.0% vs 25.5%, p<0.001) at the end of life.