PURPOSE: This study aimed to assess the impact of a bundle of care strategy on the duration of awake prone positioning (AW-PP) and other key clinical outcomes in patients with acute respiratory failure (ARF) who require high-flow nasal oxygen (HFNO). METHODS: In this secondary analysis of a prospective, multicenter cohort study, we included patients admitted with COVID-19-related ARF who required HFNO. The protocol encouraged AW-PP for as long as possible. The main exposure was a bundle of care including light sedation, monitoring, and information to patients about the strategy (bundle) compared to no bundle (control). The primary outcome was the duration of AW-PP (hours/day), while secondary outcomes included endotracheal intubation and in-hospital mortality. Directed acyclic graphs (DAGs) were employed to identify variables related to both exposure and outcomes. Four models were used to evaluate exposure-outcome associations: inverse probability of treatment weighting (IPTW), "double-robust" approximation (DR), traditional regression (TR), and mixed-effects model (MEM). RESULTS: Out of 499 patients, 197 were exposed to bundle, and 302 did not. The exposure group had a median (IQR) AW-PP duration of 16 (10-18) hours/day, compared to 10 (7-14) hours/day in the control group. Regression coefficients (95% CI) were 3.39 (1.67-5.11), 3.35 (1.55-5.14), 3.95 (2.63-5.28), and 3.72 (2.5-4.94) for IPTW, DR, TR and MEM, respectively. The odds ratios (95% CI) for intubation were 0.34 (0.15-0.76), 0.23 (0.10-0.50), 0.42 (0.23-0.77), and 0.48 (0.16-0.49), and for in-hospital mortality were 0.38 (0.11-1.27), 0.43 (0.14-1.26), 0.47 (0.22-0.91), and 0.46 (0.12-1.43) in the respective models. CONCLUSION: In the evaluated population of patients with COVID-19-related ARF, implementing a bundle-of-care strategy was associated with a longer AW-PP exposure and a reduced risk of endotracheal intubation. TRIAL REGISTRATION NUMBER: ClinicalTrials.gov. Identifier NCT05178212. Date of registration: January 5th, 2022. STUDY TYPE: Observational.