Screening of Obstructive Sleep Apnea during Pregnancy: Differences in Predictive Values of Questionnaires across Trimesters

Study Objectives: Evaluation of Berlin and Stop-Bang questionnaires in detecting obstructive sleep apnea (OSA) across trimesters of pregnancy. Methods: Pregnant women from a high-risk pregnancy clinic were recruited to complete sleep evaluations including Berlin and Stop-Bang questionnaires. Overnight testing with Watch-PAT200 for diagnosis of OSA (cutoff point of apnea-hypopnea index >= 5 events/h) was performed. Results: Seventy-two singleton pregnant women participated in the study. Enrollment consisted of 23, 24, and 25 women during first, second, and third trimesters, respectively. Of 72 pregnancies, 23 patients (31.9%) had OSA. Prevalence of OSA classified by trimesters from first to third was 30.4%, 33.33%, and 32.0%, respectively. Overall predictive values of Berlin and Stop-Bang questionnaires were fair (ROC area under curve, AUC 0.72 for Berlin, p = 0.003; 0.75 for Stop-Bang, p = 0.001). When categorized according to trimesters, predictive values substantially improved in second (AUC: 0.84 for Berlin; 0.78 for Stop-Bang) and third trimesters (AUC: 0.81 for Berlin; 0.75 for Stop-Bang), whereas performances of both questionnaires during first trimester were poorer (AUC: 0.49 for Berlin; 0.71 for Stop-Bang). Multivariate analyses show that pre-pregnancy body mass index (BMI) in first trimester, snore often in second trimester, and weight gain and pregnancy BMI in third trimester were significantly associated with OSA. Conclusions: In high-risk pregnancy, Berlin and Stop-Bang questionnaires were of limited usefulness in the first trimester. However their predictive values are acceptable as pregnancy progresses, particularly in second trimester. OSA in pregnancy seems to be a dynamic process with different predictors association during each trimester.