Good and poor sleepers among OSA patients: sleep quality and overnight polysomnography findings

Previous studies aimed to determine if Pittsburgh sleep quality index (PSQI) is a valid screening instrument for obstructive sleep apnea, indicating its disadvantages. However, the rationale of PSQI use in sleep clinics is not the screening, but the assessment of sleep quality itself. Therefore, the aims of this study were to investigate the sleep quality in obstructive sleep apnea patients and to identify the cutoff point for differentiation of "good" and "poor" sleepers among them. We constructed the Croatian version of PSQI and assessed its psychometric properties. The protocol of the study included the assessment of sleep quality in 130 obstructive sleep apnea patients and 75 healthy control subjects. All subjects completed the Croatian version of the PSQI, and the patients underwent overnight polysomnography screening. Obstructive sleep apnea patients had higher values of the global PSQI component score, indicating lower sleep quality, compared to a healthy control group (p < 0.001). The psychometric properties of PSQI scores in the prediction of normal sleep efficiency indicate that the cutoff score of 9.5 differentiates patients in total sleep time (p <  0.001), REM duration (p = 0.014), sleep efficiency (p = 0.001), time spent awake during sleep (p = 0.006), after sleep (p = 0.024), and after sleep onset (p = 0.001). In OSA patients, a PSQI cutoff score of 9.5 differentiated good and poor sleepers significantly in total sleep time, REM duration, time spent awake during sleep, and WASO time. Current findings enhance the interpretability of PSQI results in a population of OSA patients.