Background Conventionally, cardiovascular trials evaluate impact on morbid and mortal events. “Win ratio” analyses adding symptom relief and quality of life end points can assess broader intervention effects. Applying win ratio principles, we conducted a post hoc analysis of ADVENT-HF trial findings. Methods ADVENT-HF examined the impact of treating obstructive (OSA) or central (CSA) sleep apnoea with the use of peak flow–triggered adaptive servoventilation (ASVPF) in patients with heart failure with reduced ejection fraction After pre-randomisation stratification into OSA (n = 533) and CSA (n = 198) cohorts, participants were randomly allocated to standard care (n = 375) or standard care plus ASVPF (n = 356). The following trial end points were compared over shared follow-up time of pairs of treated and control participants in a descending hierarchic order: 1) all-cause death; 2) cardiovascular hospitalisations; 3) improvement from baseline at 1 year in 2 or more of the Epworth Sleepiness Scale, Minnesota Living With Heart Failure Quality of Life, or New York Heart Association functional class sccores; 4) change from baseline in the sum of stage N3 and rapid eye movement (REM) sleep stage duration; and 5) change from baseline in arousals. Results The win ratio calculated on sequential addition of outcomes to the hierarchy was 1.18 (95% CI 1.00-1.39). The win difference was 7.6%. The principal contributions to the latter were quality of life (4.2%) and time in stage 3 plus REM sleep (2.1%). Ranking quality of life second increased the win ratio to 1.22 (95% CI 1.04-1.43) and the win difference to 9.1%. Conclusions A win ratio analysis incorporating ADVENT-HF clinical and patient-relevant end points favours overall ASVPF treatment benefit. ASVPF appeared to be more effective in the CSA cohort in terms of adherence, sleep structure, and quality of life. Secondary Analysis of ADVENT-HF: NCT01128816.
Win Ratio Analysis for Peak-Flow Adaptive Servoventilation in Treating Sleep Apnoea in Heart Failure With Reduced Ejection Fraction
Redolfi, S.Membro del Collaboration Group
;
2026-01-01
Abstract
Background Conventionally, cardiovascular trials evaluate impact on morbid and mortal events. “Win ratio” analyses adding symptom relief and quality of life end points can assess broader intervention effects. Applying win ratio principles, we conducted a post hoc analysis of ADVENT-HF trial findings. Methods ADVENT-HF examined the impact of treating obstructive (OSA) or central (CSA) sleep apnoea with the use of peak flow–triggered adaptive servoventilation (ASVPF) in patients with heart failure with reduced ejection fraction After pre-randomisation stratification into OSA (n = 533) and CSA (n = 198) cohorts, participants were randomly allocated to standard care (n = 375) or standard care plus ASVPF (n = 356). The following trial end points were compared over shared follow-up time of pairs of treated and control participants in a descending hierarchic order: 1) all-cause death; 2) cardiovascular hospitalisations; 3) improvement from baseline at 1 year in 2 or more of the Epworth Sleepiness Scale, Minnesota Living With Heart Failure Quality of Life, or New York Heart Association functional class sccores; 4) change from baseline in the sum of stage N3 and rapid eye movement (REM) sleep stage duration; and 5) change from baseline in arousals. Results The win ratio calculated on sequential addition of outcomes to the hierarchy was 1.18 (95% CI 1.00-1.39). The win difference was 7.6%. The principal contributions to the latter were quality of life (4.2%) and time in stage 3 plus REM sleep (2.1%). Ranking quality of life second increased the win ratio to 1.22 (95% CI 1.04-1.43) and the win difference to 9.1%. Conclusions A win ratio analysis incorporating ADVENT-HF clinical and patient-relevant end points favours overall ASVPF treatment benefit. ASVPF appeared to be more effective in the CSA cohort in terms of adherence, sleep structure, and quality of life. Secondary Analysis of ADVENT-HF: NCT01128816.
I metadati presenti in IRIS UNICA sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono protetti da diritto d'autore, salvo diversa indicazione.
