Auto-continuous positive airway pressure (CPAP) has been reported to have no more efficacy than constant CPAP in unselected patients with sleep apnea hypopnea syndrome (SAHS). The aim of this study was to evaluate patients judged to be good candidates for auto-CPAP because of a high within-night variability in pressure requirement.Design:
Single-blind, randomized, cross-over study (2 × 8 weeks) to compare auto-CPAP with constant CPAP.Patients:
Outpatients with moderate-to-severe SAHS attending the chest clinic.Interventions:
Patients were equipped at home in the auto-CPAP mode (model GK418A; Malinckrodt; Nancy, France), using a 4- to 14-cm H2O pressure range. Those individuals having a high within-night variability in pressure requirement, assessed at the end of a 14-day run-in period, were included in the cross-over study. Auto-CPAP was compared with constant CPAP (according to a titration night in the sleep laboratory) in terms of compliance, efficacy on apneas (assessed from the pressure monitor), and sleepiness (assessed on the Epworth sleepiness scale).Measurements and results:
Of 90 consecutive patients with SAHS, 27 patients were selected for a within-night variability in pressure requirement exceeding a given threshold. After completion of the cross-over, 24 patients were evaluable. The median percentage of nights the machine was used was 95.5% (range, 45 to 100%) on constant CPAP, and 96.5% (range, 40 to 100%) on auto-CPAP; the median apnea index recorded by the device was 0.40/h (range, 0 to 2.40/h) on constant CPAP, and 0.45/h (range, 0 to 5.80/h) on auto-CPAP (differences not significant). The mean Epworth sleepiness score was significantly (p < 0.01) lower on auto-CPAP (5.1; SD, 2.8) than on constant CPAP (6.1; SD, 2.8).Conclusions:
In patients selected for a high within-night variability in pressure requirement, auto-CPAP administered via a GK418A device was equivalent to constant CPAP based on a titration night in the sleep laboratory. Subjective ratings for sleepiness were slightly lower on auto-CPAP.