Outcomes of positive airway pressure for sleep apnea: Reply

We agree with Drs Plante and Hoyt that using mean adherence to PAP among the relatively few patients remaining on treatment at 6-year follow-up in the trial by Peker et al overestimated mean adherence in that study.1 In response to Plante and Hoyt, we obtained adherence data from the trial investigators (personal written communication, Yüksel Peker, MD, PhD, August 23, 2017), which estimates mean adherence to be 2.8 hours per day rather than 6.6 hours per day and recalculated the meta-regression. The corresponding revised meta-regression for the outcome of major adverse cardiovascular events (MACE) is shown in the Figure and remains nonsignificant for the association between adherence to PAP and risk (P = .17). The associations between adherence to PAP and risk for the MACE plus hospitalization for unstable angina outcome (P = .56) and the cardiovascular death outcome (P = .98) also remained nonsignificant. Revising the adherence subgroup analyses by placing the trial by Peker et al in the subgroup with adherence less than 4 hours per day did not substantively change the conclusions—there was still no definitive evidence of heterogeneity of effects of PAP on the MACE outcome in the trials with 4 or more hours per day adherence (relative risk, 0.43 [95% CI, 0.22 to 0.85]) compared with the trials with less than 4 hours per day adherence (relative risk, 1.01 [95% CI, 0.82 to 1.25]) (P value for homogeneity = .06). The P values for the comparisons between subgroups for other outcomes also remained nonsignificant (all P > .26).