Application of the STOP-BANG Questionnaire with a Modified Scoring Model on Detection of Undiagnosed Obstructive Sleep Apnea

<bold>Background:</bold> The STOP-BANG questionnaire (SBQ) is a simple screening tool for Obstructive Sleep Apnea (OSA). It was suggested to use a modified scoring model to further improve the predictive performance in patients with score of 3 – 4, which is intermediate risk for moderate to severe OSA. A specific combination of 2 items from STOP and any 1 item from BANG was proposed. It may, therefore, aid the prioritization of polysomnography (PSG) among those patients with an intermediate risk of OSA with this two-step approach. This study aimed to evaluate the predictive performance of the modified scoring model of the SBQ to detect moderate to severe OSA. <bold>Methods:</bold> This was a retrospective single-centre study. Clinical records of all suspected OSA patients who underwent in-patient PSG after assessments in our sleep clinic were retrieved for analysis. The predictive performance of the modified scoring model of the SBQ was evaluated <bold>Results:</bold> In the subgroup of patients with a SBQ score of 3 - 4, by applying the 4 specific combinations, the specificities of STOP ≥ 2 plus BMI ≥ 35kg/ m2 and STOP ≥ 2 plus neck circumference > 40cm (S2N) were improved at 97.6% and 95.8% respectively for moderate to severe OSA. In terms of probability, only (S2N) could significantly improve the probability to 66.7% for moderate to severe OSA, compared with the probability of 37.6% if only using SBQ score 3 - 4. <bold>Conclusion:</bold> This study showed a significant improvement in the probability for moderate to severe OSA by applying the specific combination of (S2N) only in the patient group with a SBQ score of 3 – 4 with the two-stepped modified scoring model.