Violet Vines Marshman Centre for Rural Health Research, La Trobe Rural Health School, La Trobe University, Edward Street, Bendigo, Victoria, Australia.
Institute for Breathing and Sleep, Austin Health, Heidelberg, Victoria, Australia.
BMC Pregnancy Childbirth. 2024 Aug 30;24(1):565. doi: 10.1186/s12884-024-06753-z.
Sleep Disorder Breathing (SDB) in pregnant patients ranges from 3 to 27% and varies depending on gestational age and method used to diagnose. SDB increases the risk of advanced pregnancy complications such as gestational diabetes mellitus, pregnancy-induced hypertension, and preeclampsia. Screening and diagnosis of SDB during pregnancy remains a challenge, with existing screening tools underperforming during pregnancy. This study aimed to validate a previously developed model for predicting SDB during late pregnancy and compare the predictive value of bedpartner responses.
Ninety-six women in the third trimester of pregnancy underwent polysomnography and completed the Berlin Questionnaire (BQ), with 81 bedpartners completing the BQ about their pregnant partner. A subset of BQ items (snoring volume and tiredness upon awakening) along with BMI > 32 kg/m was utilised to calculate the Wilson Optimized Model (WOM), which demonstrated strong predictive properties in development.
SDB (RDI/hr ≥ 5) was detected in 43.8% of women. BQ identified 72% of pregnant mothers as high risk for SDB (Sensitivity = 83%, Specificity = 37%), compared to 29% of mothers identified by the WOM (Sensitivity = 45%, Specificity = 83%). At RDI of ≥ 15, the WOM correctly classified more women according to SDB risk than the BQ (76.0% vs. 41.7% cases correct, X(1) = 23.42, p < .001), with no difference at RDI ≥ 5. Bedpartners were more likely to report high risk for SDB on the WOM than pregnant women themselves (38.3% vs. 28.4%), however predictive ability was not improved by bedpartner input (RDI ≥ 5 bedpartner AUC = 0.69 v mother AUC = 0.73).
BQ largely overestimates the prevalence of SDB in pregnancy compared to the WOM which underestimates. Utilising bedpartner responses didn't improve screening for SDB in late pregnancy. More work is needed to develop a pregnancy-specific tool for quick and accurate screening for SDB.
孕妇睡眠呼吸障碍（SDB）的发生率为 3%至 27%，具体取决于妊娠年龄和用于诊断的方法。SDB 增加了妊娠并发症的风险，如妊娠糖尿病、妊娠高血压和子痫前期。怀孕期间 SDB 的筛查和诊断仍然是一个挑战，现有的筛查工具在怀孕期间表现不佳。本研究旨在验证之前开发的一种预测晚期妊娠 SDB 的模型，并比较床伴反应的预测价值。
96 名处于妊娠晚期的妇女接受了多导睡眠图检查，并完成了柏林问卷（BQ），其中 81 名床伴完成了关于其孕妇伴侣的 BQ。BQ 的一部分项目（打鼾音量和醒来时的疲倦感）以及 BMI>32kg/m2 被用来计算威尔逊优化模型（WOM），该模型在开发过程中表现出了很强的预测性能。
43.8%的女性被诊断为 SDB（RDI/hr≥5）。BQ 识别出 72%的孕妇有 SDB 的高风险（敏感性=83%，特异性=37%），而 WOM 识别出的母亲只有 29%（敏感性=45%，特异性=83%）。在 RDI≥15 的情况下，WOM 根据 SDB 风险正确分类的女性多于 BQ（76.0% vs. 41.7%的病例正确，X(1)=23.42，p<0.001），而在 RDI≥5 时则没有差异。与孕妇自身相比，床伴更有可能报告在 WOM 上有 SDB 的高风险（38.3% vs. 28.4%），但床伴的输入并不能提高预测能力（RDI≥5 时床伴 AUC=0.69 与母亲 AUC=0.73）。
BQ 大大高估了与 WOM 相比孕妇 SDB 的患病率，而 WOM 则低估了。利用床伴的反应并不能提高晚期妊娠 SDB 的筛查效果。需要进一步研究开发一种针对妊娠的快速准确筛查 SDB 的特定工具。
