Validation of obstructive sleep apnea questionnaires in high-risk pregnancy

Abstract

Study objectives: Obstructive sleep apnea (OSA) is an under-recognized risk factor for pregnancy complications. Sleep questionnaires vary in ability to predict pregnant patients at risk for OSA. We aimed to assess the fidelity of the DoISnorePreg questionnaire, compare its performance with the Berlin Questionnaire, and assess the impact of the Epworth Sleepiness Scale (ESS) in high-risk obstetric patients.

Methods: Patients without history of preeclampsia or gestational diabetes were recruited from a high-risk obstetric clinic, completed sleep questionnaires, and underwent home sleep apnea test at 28 weeks gestation. Patients were followed through pregnancy to assess maternal and neonatal outcomes. The predictive value of the questionnaires for OSA was assessed using C-statistics.

Results: 39% of participants had an apnea-hypopnea index ≥ 5, among whom, 29% had moderate or severe OSA. Patients with OSA had higher BMIs, older ages, higher sleepiness scores, more profound oxygen desaturations, longer post-partum lengths of stay, and newborns with increased NICU admissions. DoISnorePreg outperformed the Berlin Questionnaire with an area under the receiver operating curve of 0.74 versus 0.58. ESS did not enhance the performance of either questionnaire. A threshold of ≥ 4 positive responses on DoISnorePreg best predicted patients at risk for OSA with 89% sensitivity and 43% specificity (p < 0.01).

Conclusions: Prevalence of OSA among high-risk obstetric patients is high. Clinical implications of oxygen desaturations, longer hospital stays, and NICU admissions highlight the need for screening. DoISnorePreg is a valid predictive tool that helps fill a gap in screening this high-risk population.

Keywords: DoISnorePreg; high-risk pregnancy; obstructive sleep apnea; sleep questionnaires.