Pulmonary Function Influences the Performance of Berlin Questionnaire, Modified Berlin Questionnaire, and STOP-Bang Score for Screening Obstructive Sleep Apnea in Subjects with Chronic Obstructive Pulmonary Disease

PURPOSE: The co-existence of chronic obstructive pulmonary disease (COPD) and obstructive sleep apnea (OSA) has been described as the overlap syndrome. The objective of the study is to investigate the performance of Berlin Questionnaire (BQ), modified Berlin Questionnaire (MBQ), and STOP-Bang score in screening overlap syndrome from COPD subjects and investigate how pulmonary function interferes with questionnaire scoring. SUBJECTS AND METHODS: Among 116 COPD subjects included in this study, 62 were overlap syndrome subjects and 54 were COPD subjects without OSA. Subjects included were asked to fill out the questionnaires to collect demographic characteristics of subjects and questionnaire scores of BQ, MBQ, and STOP-Bang; perform pulmonary function test to confirm their COPD diagnosis; and perform polysomnography. RESULTS: AUC (area under the curve) of BQ, MBQ, and STOP-Bang score in screening OSA among patients with COPD was 0.71 (0.64–0.79), 0.75 (0.67–0.83), and 0.72 (0.64–0.80). In COPD subjects without OSA, FEV1%pred was statistically associated with the misdiagnosis of BQ (P= 0.0091), MBQ (P= 0.0143), and STOP-Bang (P= 0.0453). In patients with overlap syndrome, FVC%pred affected the risk of misdiagnosis of the three questionnaires (BQ: P= 0.0413; MBQ: P= 0.0150; STOP-Bang: P= 0.0241). BMI and neck circumferences (NC) were negatively correlated with FEV1%pred (BMI: P= 0.0454; NC: P= 0.0230) and FVC%pred (BMI: P= 0.0042; NC: P= 0.0367) in overlap subjects. In contrast, BMI was positively correlated with FEV1/FVC (P= 0.0141) and FEV1%pred (P= 0.0391) in COPD subjects without OSA. CONCLUSION: BQ, MBQ, and STOP-Bang score performed well in COPD subjects for screening OSA. The diagnosis of the three questionnaires was more accurate in subjects with lower FEV1%pred or FVC%pred value. Pulmonary function might exert influence on the diagnosis efficacy of the three questionnaires through BMI and neck circumference.