A Modified Definition for Obstructive Sleep Apnea in Home Sleep Apnea Testing after Stroke or Transient Ischemic Attack.

BACKGROUND

Home sleep apnea testing (HSAT) underestimates obstructive sleep apnea (OSA) severity. Overnight oxygen desaturation has been shown to be a strong predictor of incident stroke, and may be helpful in determining which patients with lower apnea-hypopnea indices (AHIs) should be offered treatment.

OBJECTIVES

To examine whether a modified definition for OSA that uses oxygen desaturation (i.e., AHI 5-14.9 per hour and lowest O2 desaturation ≤88%), as compared to an AHI ≥ 15 per hour, would impact: (1) changes in Epworth Sleepiness Scale scores post-continuous positive airway pressure (CPAP) initiation; (2) CPAP compliance rates; and (3) the accuracy of automated versus manual scoring of HSAT.

METHODS

One hundred and six patients with a prior stroke or transient ischemic attack (TIA) underwent HSAT. Epworth Sleepiness Scale and CPAP compliance were measured at baseline and 3-6 months.

RESULTS

Median age was 67.5 years, 57.5% male, and 76.4% presented with stroke. Fifty-nine patients were diagnosed with OSA; of these 54.2% met criteria for the "modified definition" for OSA (AHI 5-14.9 per hour with oxygen desaturation) and 45.8% met criteria for the "classic definition" for OSA (AHI ≥15). The modified (versus classic) definition had: (1) a greater decrease in ESS (P = .05) post-CPAP initiation; (2) comparable CPAP compliance rates; and (3) comparable automatically versus manually scored AHIs (Spearman's rho = .96, Cohen's Kappa ≥ .75 for both definitions, P < .001).

CONCLUSIONS

Using a modified definition of OSA that uses a lower AHI cutoff and includes an oxygen desaturation cutoff in the setting of HSAT for stroke or transient ischemic attack (TIA) patients may improve daytime sleepiness post-CPAP initiation, while not significantly affecting CPAP compliance rates nor the accuracy of automated scoring.