Effect of Positive Airway Pressure Therapy on Mortality in European Obstructive Sleep Apnea Patients: Post Hoc Analysis of a Global Meta-Analysis
Author(s)
Ankit Ghildiyal, M.Pharm, MSc1, Melike Deger Wehr, BSc, MSc2, Fatima Sert Kuniyoshi, PhD3, Florent Lavergne, MSc4, Alison Wimms, PhD5.
1ResMed, Oxfordshire, United Kingdom, 2ResMed, Münich, Germany, 3ResMed, San Diego, CA, USA, 4ResMed, Lyon, France, 5ResMed, Sydney, Australia.
1ResMed, Oxfordshire, United Kingdom, 2ResMed, Münich, Germany, 3ResMed, San Diego, CA, USA, 4ResMed, Lyon, France, 5ResMed, Sydney, Australia.
OBJECTIVES: To evaluate the impact of positive airway pressure (PAP) therapy on all-cause and cardiovascular (CV) mortality specifically among European adults with obstructive sleep apnea (OSA), using subgroup analysis from a recently published global meta-analysis.
METHODS: This post hoc analysis extracted European studies from the forest plots of a published global meta-analysis (Benjafield et al., Lancet Respir Med, 2025). Studies were stratified by design into randomised controlled trials (RCTs) and non-randomised controlled studies (NRCs). Notably, NRCs in the original meta-analysis were adjusted for key confounders. Hazard ratios (HRs) and standard errors were derived from reported confidence intervals assuming log normality. Meta-analyses were conducted using both DerSimonian-Laird (DL) and Hartung-Knapp-Sidik-Jonkman (HKSJ) random-effects models in R (metafor package).
RESULTS: Eighteen studies (n = 254,205) met inclusion criteria: 7 RCTs (n = 2,855) and 11 NRCs (n = 251,350). For all-cause mortality (RCT=7; NRCs=9), the DL model showed a significant overall benefit of PAP (HR = 0.65; 95% CI: 0.54-0.78), with stronger effects in NRCTs (HR = 0.63; 95% CI: 0.52-0.76), while effects in RCTs were not significant (HR = 0.89; 95% CI: 0.49-1.61). HKSJ results were similar (HR = 0.65; 95% CI: 0.55-0.77), with consistent patterns across study types. For CV mortality (RCT=5; NRCs=3), PAP therapy was linked to a significant overall reduction in risk in the DL model (HR = 0.42; 95% CI: 0.25-0.72), with strong effects in NRCTs (HR = 0.38; 95% CI: 0.19-0.73) and non-significant results in RCTs (HR = 0.52; 95% CI: 0.19-1.41). HKSJ estimates confirmed the findings with wider uncertainty (overall HR = 0.42; 95% CI: 0.22-0.81).
CONCLUSIONS: This subgroup meta-analysis suggests that PAP therapy is associated with reduced all-cause and cardiovascular mortality in European OSA populations, particularly in NRCs. Despite inherit the methodological strengths and limitations of the parent meta-analysis, including potential residual confounding and lack of objective PAP adherence data, our findings reinforce the clinical case for broader treatment access
METHODS: This post hoc analysis extracted European studies from the forest plots of a published global meta-analysis (Benjafield et al., Lancet Respir Med, 2025). Studies were stratified by design into randomised controlled trials (RCTs) and non-randomised controlled studies (NRCs). Notably, NRCs in the original meta-analysis were adjusted for key confounders. Hazard ratios (HRs) and standard errors were derived from reported confidence intervals assuming log normality. Meta-analyses were conducted using both DerSimonian-Laird (DL) and Hartung-Knapp-Sidik-Jonkman (HKSJ) random-effects models in R (metafor package).
RESULTS: Eighteen studies (n = 254,205) met inclusion criteria: 7 RCTs (n = 2,855) and 11 NRCs (n = 251,350). For all-cause mortality (RCT=7; NRCs=9), the DL model showed a significant overall benefit of PAP (HR = 0.65; 95% CI: 0.54-0.78), with stronger effects in NRCTs (HR = 0.63; 95% CI: 0.52-0.76), while effects in RCTs were not significant (HR = 0.89; 95% CI: 0.49-1.61). HKSJ results were similar (HR = 0.65; 95% CI: 0.55-0.77), with consistent patterns across study types. For CV mortality (RCT=5; NRCs=3), PAP therapy was linked to a significant overall reduction in risk in the DL model (HR = 0.42; 95% CI: 0.25-0.72), with strong effects in NRCTs (HR = 0.38; 95% CI: 0.19-0.73) and non-significant results in RCTs (HR = 0.52; 95% CI: 0.19-1.41). HKSJ estimates confirmed the findings with wider uncertainty (overall HR = 0.42; 95% CI: 0.22-0.81).
CONCLUSIONS: This subgroup meta-analysis suggests that PAP therapy is associated with reduced all-cause and cardiovascular mortality in European OSA populations, particularly in NRCs. Despite inherit the methodological strengths and limitations of the parent meta-analysis, including potential residual confounding and lack of objective PAP adherence data, our findings reinforce the clinical case for broader treatment access
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
MSR81
Topic
Clinical Outcomes, Medical Technologies, Methodological & Statistical Research
Disease
Respiratory-Related Disorders (Allergy, Asthma, Smoking, Other Respiratory)