Initial establishment and validation of a predictive model for preeclampsia in twin pregnancies based on maternal characteristics and echocardiographic parameters

• Published in: Acta obstetricia et gynecologica Scandinavica Vol. 102; no. 11; pp. 1566 - 1574
• Main Authors: Xiang, Qianqian, Wei, Yuan, Feng, Xinheng, Chen, Shaomin, Zhao, Yangyu
• Format: Journal Article
• Language: English
• Published: United States John Wiley & Sons, Inc 01.11.2023; John Wiley and Sons Inc; Wiley
• Subjects: cardiac function; Health risks; Original; prediction model; Preeclampsia; Pregnancy; transthoracic echocardiography; twin pregnancy; Twins; Ultrasonic imaging; transthoracic echocardiography; cardiac function; prediction model; preeclampsia; twin pregnancy
• ISSN: 0001-6349; 1600-0412; 1600-0412
• DOI: 10.1111/aogs.14651

Introduction The primary objective of this study was to evaluate the association between maternal echocardiographic changes and the occurrence of preeclampsia (PE) in twin pregnancies. Additionally, we established and validated a prediction model for PE in twin pregnancies. Material and methods The first part of this study was retrospective and included data from 854 twin pregnancies that received antenatal care at Peking University Third Hospital from April 2017 to April 2021 (training cohort). Overall, 159 women who underwent transthoracic echocardiography were included in the analysis. To build a predictive model, cardiac findings were compared between normotensive women and those with PE. The model was then validated in a prospective longitudinal cohort (test cohort) that included 109 women with twin pregnancies who underwent two consecutive transthoracic echocardiography examinations during the second and third trimesters. Results Fifty‐four normotensive women and 105 women with PE were analyzed in the retrospective cohort that was used to build the model in which later preeclampsia was associated with higher left ventricular mass index (>61 g/m2), interventricular septal thickness (>7.87 mm), left atrial anteroposterior diameter (>33.5 mm), mitral inflow late diastolic velocity (A) (>0.685 m/s), ratio of early diastolic velocity (E) and peak early diastolic myocardial velocity of the lateral mitral annulus (>6.5), and lower peak early diastolic myocardial velocity (<13.1 cm/s). The optimized PE prediction model based on the interventricular septal thickness, left atrial anteroposterior diameter, A, peak early diastolic myocardial velocity and pre‐pregnancy bodyweight index was then established (area under the curve [AUC] = 0.840, 95% CI 0.778–0.903, P < 0.001). The model was tested in the prospective cohort including 87 normotensive women and 22 women with PE; the validation test showed that the prediction model in the second (AUC = 0.801) and third (AUC = 0.811) trimesters had high discriminative ability and calibration. Conclusions Maternal echocardiographic changes in twin pregnancies are associated with the development of preeclampsia. The model constructed, based on the echocardiographic parameters and body mass index, provides novel ideas for the prediction of PE. A predictive model for PE in twin pregnancy was constructed using maternal cardiac function parameters and characteristics, including IVST, LAAP, A, Em and BMI.