Introduction The MOSAIC multicentre trial investigated the effect of CPAP versus no CPAP on both sleepiness and a vascular risk score, over 6 months, in nearly 400 minimally symptomatic patients with OSA. The results on the vascular risk score, reported previously, suggested no potential vascular benefit of CPAP, although endothelial function improved. The incidence of real events is needed to calculate the size of any future trials in this group of patients.
Methods We have followed the Oxford cohort of 189 patients for two years to assess real vascular event rates and the incidence of accidents and hospital admissions. The events have been verified by inspection of hospital notes and phone calls to patients. The CPAP/No CPAP allocation was based on intention to treat. At the end of the original 6 month trial, 71% of patients, initially randomised to CPAP, stated an intention to continue CPAP.
Results The table shows the 2 year incidence of accidents at home, work and driving (66/189), the incidence of new vascular events (25/189), and all cause mortality (4/189).
Conclusions The event rates are moderate, and there is no statistically significant difference between the groups. A power calculation, based on these figures, suggests that in order not to miss a 30% reduction in relative risk over 2 years (4% absolute reduction), with 90% power, would require 2,540 patients randomised; for accidents, 880; for mortality over 6,000.