The No-apnea score vs. the other five questionnaires in screening for obstructive sleep apnea-hypopnea syndrome in patients with cerebral infarction.

Abstract

Given the growing number of patients suspected of having obstructive sleep apnea-hypopnea syndrome (OSAHS), screening methods have become increasingly important for sleep clinics. We analyzed the clinical value of the No-apnea score which is used to diagnose OSAHS in patients with cerebral infarction, and compared the accuracy of the No-apnea score with the accuracy of the NoSAS score, the STOP-Bang questionnaire (SBQ), the Epworth Sleepiness Scale (ESS), the STOP questionnaire (STOP) and the Berlin questionnaire (BQ). Between January 2014 and December 2018, a total of 221 cerebral infarction patients, suspected of having OSAHS, underwent the polysomnography (PSG) for one night at the sleep medical center of Guangdong Medical University Affiliated Second Hospital. The PSG data were collected and analyzed with the NoSAS score, the SBQ, the ESS, the STOP, the BQ, and patients' demographic information. Based on the apnea-hypopnea index (AHI), the patients were classified into four groups: the normal group (<5 events/h), mild OSAHS group (5-15 events/h), moderate OSAHS group (15-30 events/h) and severe OSAHS group (≥30 events/h). The sensitivity, specificity, positive predictive value, negative predictive value and areas under the curve (AUC) of the Receiver Operating Curve (ROC) were calculated for the five questionnaires to compare their relative efficacies for diagnosing OSAHS. When using the standard of AHI ≥5 for diagnosing OSAHS, the NoSAS score had an AUC of 0.831; the SBQ had an AUC of 0.730; the BQ had an AUC of 0.698; and the STOP had an AUC of 0.735, so these techniques are relatively accurate in diagnosing OSAHS. On the other hand, the No-apnea score and the ESS score are relatively less accurate comparing to the rest: the No-apnea had an AUC of 0.626, and the ESS had an AUC of 0.650. Using the NoSAS score to predict AHI ≥5 events/h, AHI ≥15 events/h and AHI ≥30 events/h, the sensitivity and specificity were 0.867 and 0.731, 0.888 and 0.476, 0.889 and 0.369, respectively; Using the SBQ to predict AHI ≥5 events/h, AHI 15 events/h and AHI ≥30 events/h, the sensitivity and specificity were 0.903 and 0.268, 0.914 and 0.200, 0.903 and 0.268, respectively; Using the STOP to predict AHI ≥5 events/h, AHI ≥15 events/h and AHI ≥30 events/h, the values were 0.830 and 0.500, 0.871 and 0.390, 0.875 and 0.302, respectively; and using the BQ to predict AHI ≥5 events/h, AHI ≥15 events/h and AHI ≥30 events/h, the values were 0.758 and 0.482, 0.810 and 0.429, 0.819 and 0.362, respectively. The study concludes that the NoSAS score and the SBQ had a better predictive value for cerebral infarction patients suspected with OSAHS disease. These questionnaires can also effectively help clinicians quickly address nocturnal hypoxia in patients with cerebral infarction to control subsequent complications in patients with cerebral infarction. More studies are needed to evaluate the efficacy of the NoSAS score in screening for OSAHS in patients with cerebral infarction.