Modification and accuracy of the STOP-Bang questionnaire for screening obstructive sleep apnea in adults

Abstract

Objective:To assess the accuracy of the STOP-Bang questionnaire（SBQ） about screening for obstructive sleep apnea（OSA） in adults, and to establish appropriate cutoffs for body mass index（BMI）, neck circumference（NC）, age, and SBQ score for this population in combination with the body size characteristics of the study subjects, and to evaluate the accuracy of the SBQ for combined screening with waist to height ratio（WHtR）. Methods:The data were collected from October, 2019 to November, 2020 at the Affiliated Hospital of Guizhou Medical University, and 218 patients with suspicious OSA were included in this study, all of them completed the SBQ and underwent overnight PSG. SBQ screening and PSG were tested in a blinded concurrent manner. Using the PSG results as the gold standard, the subjects were divided into a diseased group with moderate to severe OSA and a non-diseased group based on apnea hypopnea index. According to SBQ screening results, the subjects were divided into a positive group with moderate severe OSA and negative group. And calculated in the form of four-fold table, the optimal cutoffs were determined by adjusting age, BMI, NC, and SBQ score cutoffs, and calculating the sensitivity, specificity, positive predictive value（PPV）, negative predictive value（NPV）, Youden index, and area under the receiver operating characteristic curve（ROC）. And the modified version of SBQ was combined with WHtR for joint diagnosis to evaluate the accuracy of combined screening. Results:With all 218 patients completing the experiment, 159（72.94%） patients were in the moderate-severe group. The original SBQ had a sensitivity of 93.71%, specificity of 47.46%, PPV of 82.80%, NPV of 73.70%, Youden index of 0.41, and area under the ROC curve（AUC） of 0.71 for predicting moderate to severe OSA. When adjusting the BMI cutoff to 30 kg/m², NC cutoff to 40 cm, and age cutoff to 50 years, the modified SBQ had a sensitivity of 95.60%, specificity of 47.46%, PPV of 83.10%, NPV of 80.00%, Youden index of 0.43, and AUC of 0.72 for predicting moderate to severe OSA. When the modified SBQ score was 4 or more as the cutoff value, the highest Youden index and AUC were 0.54, 0.77 respectively. When the modified SBQ≥3 points were combined with WHtR>0.55, the highest Youden index and AUC were 0.58, 0.79 respectively. Conclusion:SBQ is effective in screening for moderate to severe OSA and better screening accuracy can be achieved by using a modified version of the SBQ（BMI>30 kg/m², NC>40 cm, age>50 years）. The maximum diagnostic accuracy can be achieved when the modified SBQ total score was 4 or more as the cutoff. Those with the modified SBQ total score=3 and WHtR>0.55 can be classified as high risk for moderate to severe OSA, and the modified version of SBQ can be a screening tool to identify moderate to severe OSA.