The utility of patient-completed and partner-completed Epworth Sleepiness Scale scores in the evaluation of obstructive sleep apnea.

Abstract

Excessive daytime sleepiness in obstructive sleep apnea (OSA) is often rated differently by patients and their partners. This cross-sectional study compared the utility of patient-completed and partner-completed Epworth Sleepiness Scale (ESS) scores in the evaluation of suspected OSA. Eighty-five patient-partner pairs were enrolled, and 75 patients completed diagnostic sleep studies. The individual and combined utilities of patient-completed and partner-completed ESS scores in identifying OSA and predicting various sleep study-derived indicators of disease severity were determined. Mean partner-completed ESS scores were higher than patient-completed ESS scores (12.3±4.2 vs. 9.4±4.8, p<0.0001); Bland-Altman plot showed significant bias (partner-completed ESS scores 33.5% higher, SD ±55.2%). Partner-completed and combined (but not patient-completed) ESS scores correlated weakly with the apnea-hypopnea index (AHI; partner-completed ESS score r s =0.25, p=0.029; combined ESS score r s =0.29, p=0.013) and oxygen desaturation index (partner-completed ESS score r s =0.26, p=0.025; combined ESS score r s =0.23, p=0.047). None of the ESS scores correlated with body mass index, arousal index, or other parameters of nocturnal oxygen desaturation. In OSA (AHI>15/h) detection, partner-completed ESS scores had greater sensitivity than patient-completed ESS scores (76.9 vs. 46.2%) but poorer specificity (39.1 vs. 65.2%); sensitivity was greatest (82.7%) when either patient-completed or partner-completed ESS score was 10 or higher, and specificity was greatest (80.8%) when both scores were 10 or higher. Neither patient-completed nor partner-completed ESS scores by themselves have great utility in identifying OSA or predicting its severity. However, taking both scores into consideration together improves the sensitivity and specificity of the screening process.