Projects such as the American Board of Internal Medicine’s “Choosing Wisely” campaign emphasize the importance of identifying and de-implementing ineffective, harmful and/or wasteful practices to improve the quality and safety of delivered care. Among patients with chronic obstructive pulmonary disease (COPD), oxygen has been shown to improve survival among a specific subset of patients with severe resting hypoxemia. The provision of oxygen therapy has been extrapolated to other populations leading to the prescription of oxygen to patients with COPD who do not have an evidence-based indication for its use. For example, oxygen is frequently prescribed upon discharge from hospital after an exacerbation of COPD, then continued indefinitely, resulting in patients who become habituated to having oxygen readily accessible. Published in 2016, the Long-Term Oxygen Treatment Trial (LOTT) demonstrated that the use of supplemental oxygen conferred no benefit to patients with COPD who have moderate resting or ambulatory hypoxemia. In addition to being ineffective in this patient population, supplemental oxygen can pose risks of harm, such as the potential for fire or injury due to falls caused by entanglement in oxygen tubing. It is in this setting that we now have an opportunity, as well as a challenge, to learn how best to improve care quality by de-implementing a therapy that has no demonstrable benefit, is not without risk, but is deeply entrenched in clinical practice. To address this knowledge gap, our proposed study aims to use a mixed-methods approach to inform VA about their current practices for prescribing oxygen therapy to patients with COPD, as well as to identify anticipated barriers and facilitators to its de-implementation among patients in whom it has been shown to be ineffective. To achieve these goals, we will first identify a nationwide cohort of Veterans discharged with supplemental oxygen after hospitalization for a COPD exacerbation and identify patient-, clinician- and system level factors associated with appropriate reassessment of oxygen need within 90 days of discharge. We will then identify a range of high- to low-performing sites in regards to timeliness of oxygen need reassessment. Using a sequential mixed methods approach incorporating quantitative survey collection and in-depth qualitative interviews, we will elicit patient and clinician perspectives on the discontinuation of supplemental oxygen across a range of the high- and low-performing sites as well as at the sites of our planned intervention. Using these initial results, we will develop, refine, and test a patient-, organizational-, and operational partner- informed pilot intervention to increase value-based care by decreasing inappropriate oxygen use. We will randomize 50 patients at two VA sites (VA Puget Sound and Bedford VA) to receive usual care or our de- implementation intervention. In addition to assessing feasibility and acceptability of the intervention, we wi...