Abstract

OBJECTIVE: We sought to determine the role of respiratory assessment by cardiorespiratory symptoms and/or oxygen saturation by pulse oximetry (SpO2) in predicting adverse maternal outcomes in women admitted to hospital with preeclampsia. METHODS: These data derive from an international, prospective multicentre cohort study, PIERS (Pre-eclampsia Integrated Estimate of RiSk), which assesses predictors of adverse outcomes in women admitted to tertiary perinatal units with preeclampsia. Univariate and multivariate analyses of cardiorespiratory symptoms and pulse oximetry were performed to assess their ability to predict a combined adverse maternal outcome developed through international Delphi consensus. RESULTS: SpO2 successfully predicted adverse maternal outcomes; the area under the receiver-operator characteristic curve (AUC ROC) was 0.71 (95% CI 0.65 to 0.77). Combining the symptoms of chest pain and/or dyspnea with pulse oximetry improved this predictive ability (AUC ROC 0.73; 95% CI 0.67 to 0.78). When SpO2 was stratified into risk groups using inflection points on the ROC curve, the highest risk group (SpO2 90% to 93%) had an odds ratio of 18.1 (95% CI 8.2 to 40.2) for all outcomes within 48 hours when compared with the baseline group (SpO2 98% to 100%). CONCLUSION: Assessing SpO2 aids in the assessment of maternal risk in women admitted to hospital with preeclampsia. An SpO2 value of ≤ 93% confers particular risk. The symptom complex of chest pain and/or dyspnea adds to the association.