Related Experiment Video
Updated: Dec 6, 2025
Drug-Induced Sleep Endoscopy DISE with Target Controlled Infusion TCI and Bispectral Analysis in Obstructive Sleep Apnea
Published on: December 6, 2016
STOP-Bang and NoSAS questionnaires as a screening tool for OSA: which one is the best choice?
José Coutinho Costa1, Alexandre Rebelo-Marques2,3,4, João Pedro Neiva Machado1
1Pneumology Service, Hospital and University Center of Coimbra, Coimbra, Portugal.
Introduction:
Currently there has been significant growth in the number of patients with suspected obstructive sleep apnea (OSA) referred to sleep clinics. In this sense, screening and stratification methods of the severity of this pathology have become increasingly relevant.
Objective:
To evaluate the performance of the NoSAS and STOP-Bang scores in the screening of OSA in a sleep clinic.
Methods:
Prospectively, for 12 months, all patients referred by primary care physicians to our sleep unit for clinical evaluation and who underwent in-lab polysomnography (PSG), also completed the NoSAS score (Neck circumference, Obesity, Snoring, Age, Sex) and STOP-Bang (Snoring, Tiredness, Observed apnea, Pressure (high blood), BMI, Age, Neck circumference, Gender). A ROC (receiver operating characteristic) analysis was used to find the scores that simultaneously maximize sensitivity and specificity for each diagnosis.
Results:
Of the 294 individuals included, 84% had OSA, of which 28.8% were mild, 34.8% moderate, and 36.4% were severe.
Using The Nosas Score For Predicting Osa, Moderate To Severe Osa, And Severe Osa, The Roc Area Was:
0.770 (95% CI: 0.703-0.837), p<0.001, sensitivity of 57.5%, and specificity of 83.0% for a score of 12; 0.746 (95% CI: 0.691-0.802), p<0.001, sensitivity of 68.2% and specificity of 75.4% for a score of 13; 0.686 (95% CI: 0.622-0.749), p<0.001, sensitivity of 71.1% and specificity of 58.3% for a score of 13, respectively.
Using The Stop-Bang Score For Predicting Osa, Moderate To Severe Osa, And Severe Osa, The Roc Area Was:
0.862 (95% CI: 0.808-0.916), p<0.001, sensitivity of 68.4% and specificity of 85.1% for a score of 5; 0.813 (95% CI: 0.756-0.861), p<0.001, sensitivity of 77.3% and specificity of 66.1% for a score of 5; 0.787 (95% CI: 0.732-0.841), p<0.001, sensitivity of 70.0% and specificity of 79.9% for a score of 6, respectively.
Conclusions:
The ROC area was consistently high for both scores confirming the diagnostic ability of the NoSAS and STOP-Bang questionnaires for all OSA severities. Thus, our results suggest that these questionnaires may be a powerful tool for the screening and stratification of patients in the diagnosis of OSA. Overall, the diagnostic ability of the STOP-Bang was higher than the NoSAS.
