Diagnostic Accuracy of the Screening Questionnaires for Obstructive Sleep Apnoea in Pregnancy: A Meta-Analysis and Updated Systematic Review.

Abstract

This systematic review and meta-analysis evaluate the diagnostic performance of obstructive sleep apnoea (OSA) screening tools in pregnant populations and the efficacy of common sleep questionnaires. A comprehensive search was conducted from the beginning to March 2024 using MEDLINE, Scopus, Cochrane CENTRAL, and Google Scholar, adhering to PRISMA guidelines. Studies were included if they adopted OSA screening questionnaires in pregnant women and compared results with overnight polysomnography. Keywords included terms related to pregnancy and OSA (e.g., Berlin questionnaire, ESS, PSQI, PSG, Watch-PAT). Sensitivity, specificity, diagnostic odds ratios (DOR), and likelihood ratios (LR) were calculated for each study. Tests for equality of sensitivity and specificity were conducted to evaluate variability. Eight studies involving 10,043 pregnant women were included. Reported OSA prevalence ranging from 12% to 72.3%. The Berlin questionnaire demonstrated significant heterogeneity in both sensitivity (χ² = 23.54, df = 6, p = 0.0006) and specificity (χ² = 33.74, df = 6, p = 7.56e-06), with a strong positive correlation between sensitivity and false positive rate (max. correlation coefficient = 0.994; 95% CI 0.092-0.994). The ESS showed similar variability (sensitivity: χ² = 10.55, df = 4, p = 0.0321; specificity: χ² = 74.18, df = 4, p = 2.97e-15), also revealing a positive correlation between sensitivity and false positives. Traditional screening tools such as the Berlin questionnaire and ESS exhibit poor diagnostic accuracy for OSA during pregnancy. These findings highlight the need for pregnancy-specific screening instruments and further research into OSA risk factors unique to this population.