To evaluate the efficacy and safety of empiric auto-adjusting CPAP to treat suspected OSA in veterans awaiting polysomnography.
Consecutive patients referred for polysomnography were screened. Exclusion criteria were: known OSA, risk of central apnea, insomnia, narcolepsy, or severe comorbid disease. Inclusion criteria were positive findings in two of the following three categories: severe snoring, daytime sleepiness, and either hypertension or BMI>30. Eligible subjects were seen in clinic and randomized either to treatment with auto-adjusting CPAP or to usual care at the VA. Subjects completed the FOSQ quality of life indicator and the Epworth sleepiness scale at the initial visit and again after 4 weeks.
338 patients were screened. 58 patients met entry criteria and were enrolled. Of these, 39 have completed the initial and follow up questionnaires, 24 in the treatment group and 15 in the control group. The groups did not differ significantly at the start of the study. At follow up, the mean Epworth scores in the treatment and control groups were 11.67 and 15.13 respectively (p=0.01). The mean FOSQ scores in the treatment and control groups were 15.14 and 13.43 respectively (p=0.055). There was a statistically significant improvement in the Epworth scores of patients treated with auto-CPAP. A trend toward improvement in FOSQ was observed that did not reach statistical significance. There were no adverse events documented. One patient withdrew for reasons unrelated to the study.
These data suggest that auto-CPAP significantly improved the Epworth scores of veteran patients likely to have OSA while awaiting polysomnography. This is an encouraging finding and warrants continuation of this trial to its planned enrollment of 300 subjects.
While the prevalence of OSA in the veteran population is likely higher than in the general population, the waiting time for polysomnography can be long. These data suggest empiric treatment while awaiting a polysomnogram may be safe and effective in select patients.
Fitzgerald Drummond, None.