RESUMO
OBJECTIVE: Several authorities have recommended the use of an obstetric early warning system (OEWS) to prevent severe morbidity and mortality. Data on the accuracy of OEWS in different clinical settings and maternal populations are still scarce. Our aim was to validate OEWS to detect maternal morbidity among high-risk women in the postnatal ward. METHODS: An OEWS was assigned to women with a body mass index >35 kg/m2, postpartum hemorrhage (PPH) >1500 mL, preeclampsia, concern over the maternal condition, chorioamnionitis, or type 1 diabetes. Morbidity was defined as worsening preeclampsia, action on hemorrhage, thromboembolia, diabetic ketoacidosis, puerperal infections, transfer to the intensive care unit, cardiopulmonary dysfunction, or death during the stay in the postnatal ward. The use of OEWS was implemented on November 1, 2016, and the study period ended on April 30, 2018. RESULTS: The study group included 827 women. The incidence of maternal morbidity was 29%. Women with PPH (odds ratio [OR], 6.4 [95% confidence interval, 3.5-11.6]) and preeclampsia (OR, 5.7 [3.5-9.6]) had the highest risk for morbidity. The sensitivity of OEWS for any morbidity was 42% (35%-48%), the specificity was 83% (80%-86%), the positive predictive value was 50% (44%-56%), and the negative predictive value was 78% (76%-80%). Systolic (OR, 6.8 [4.0-11.5]) and diastolic (OR, 3.3 [1.8-6.0]) blood pressure as well as pulse (OR, 2.1 [1.1-4.2]) predicted morbidity the most. CONCLUSIONS: In high-risk women, OEWS revealed one-half of the morbidity. Women with PPH and preeclampsia benefited most from it. Abnormal blood pressure and pulse had the strongest associations with morbidity.