Gomes Iury, Marçal Franz, Reis João Victor de Carvalho, Silva Ana Paula Dos Santos, Sampaio Lucas, Moreira Leonardo Alves, Costa Guilherme Lelis, Correa Mário Dias, Reis Zilma Silveira Nogueira, Mayrink Jussara
Universidade Federal de Minas Gerais Belo HorizonteMG Brazil Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil.
Rev Bras Ginecol Obstet. 2025 Jul 15;47. doi: 10.61622/rbgo/2025rbgo39. eCollection 2025.
Low- and middle-income countries face significant challenges in managing women diagnosed with pre-eclampsia, from making the clinical decision about whether to deliver to transferring these women to healthy facilities where they can receive appropriate care. The aim of this study was to evaluate the performance and accuracy of the fullPIERS model in a referral Brazilian maternity hospital - to assess maternal and fetal morbidity and impatient mortality at birth admission.
A cross-sectional study analyzed pregnant women with preeclampsia diagnosis, between 2014 and 2023. The full PIERS model was applied to a database retrospectively collected and its accuracy to predict maternal and perinatal outcomes during the hospital stay was determined through a receiver operating curve.
Analyzing 207 pregnant women with fullPIERS had an Area Under the Curve (AUC) for adverse maternal outcome discrimination of 0.672 (0.576-0.767 95% CI, p<0.001) and AUC 0.582, (0.504-0.6661 95% CI, p = 0.041) for maternal and perinatal outcomes. Nevertheless, the model had no discrimination utility to assess perinatal outcomes (AUC 0.561, 0.480-0.642 95% CI, p = 0.642).
The fullPIERS model had limited performance in identifying women at increased risk of adverse outcomes birth admission and absent utility to assess perinatal outcomes. Future studies, combining different tools and validated in low- and middle-income countries should be carried out to improve maternal health.
低收入和中等收入国家在管理被诊断为子痫前期的妇女方面面临重大挑战，从做出是否分娩的临床决策到将这些妇女转移到能够提供适当护理的健康设施。本研究的目的是评估fullPIERS模型在巴西一家转诊妇产医院的性能和准确性——评估入院分娩时的孕产妇和胎儿发病率以及住院死亡率。
一项横断面研究分析了2014年至2023年间被诊断为子痫前期的孕妇。将fullPIERS模型应用于回顾性收集的数据库，并通过受试者工作特征曲线确定其预测住院期间孕产妇和围产期结局的准确性。
分析207名孕妇，fullPIERS模型对不良孕产妇结局的鉴别曲线下面积（AUC）为0.672（0.576 - 0.767，95%置信区间，p<0.001），对孕产妇和围产期结局的AUC为0.582（0.504 - 0.6,661，95%置信区间，p = 0.041）。然而，该模型在评估围产期结局方面没有鉴别效用（AUC 0.561，0.480 - 0.642，95%置信区间，p = 0.642）。
fullPIERS模型在识别入院分娩时不良结局风险增加的妇女方面性能有限，且在评估围产期结局方面没有效用。未来应开展结合不同工具并在低收入和中等收入国家进行验证的研究，以改善孕产妇健康状况。
