Prediction of clinical outcomes in women with placenta accreta spectrum using machine learning models: an international multicenter study

Placenta accreta spectrum is a major obstetric disorder that is associated with significant morbidity and mortality. The objective of this study is to establish a prediction model of clinical outcomes in these women PAS-ID is an international multicenter study that comprises 11 centers from 9 countries. Women who were diagnosed with PAS and were managed in the recruiting centers between 1 January 2010 and 31 December 2019 were included. Data were reanalyzed using machine learning (ML) models, and 2 models were created to predict outcomes using antepartum and perioperative features. ML model was conducted using python^® programing language. The primary outcome was massive PAS-associated perioperative blood loss (intraoperative blood loss ≥2500 ml, triggering massive transfusion protocol, or complicated by disseminated intravascular coagulopathy). Other outcomes include prolonged hospitalization &gt;7 days and admission to the intensive care unit (ICU). 727 women with PAS were included. The area under curve (AUC) for ML antepartum prediction model was 0.84, 0.81, and 0.82 for massive blood loss, prolonged hospitalization, and admission to ICU, respectively. Significant contributors to this model were parity, placental site, method of diagnosis, and antepartum hemoglobin. Combining baseline and perioperative variables, the ML model performed at 0.86, 0.90, and 0.86 for study outcomes, respectively. Ethnicity, pelvic invasion, and uterine incision were the most predictive factors in this model. ML models can be used to calculate the individualized risk of morbidity in women with PAS. Model-based risk assessment facilitates <i>a priori</i> delineation of management.

胎盘植入谱系（Placenta accreta spectrum, PAS）是一类可导致显著发病率与死亡率的严重产科疾病。本研究旨在构建此类患者的临床结局预测模型。PAS-ID是一项国际多中心研究，纳入了来自9个国家的11个研究中心。研究对象为2010年1月1日至2019年12月31日期间，在各招募中心确诊并接受治疗的PAS患者。本研究采用机器学习（Machine Learning, ML）模型对数据进行重新分析，构建了两个分别基于产前特征与围手术期特征的结局预测模型。机器学习模型通过Python^®编程语言实现。本研究的主要结局指标为与PAS相关的大量围手术期失血，定义为术中失血量≥2500ml、触发大量输血方案，或并发弥散性血管内凝血。次要结局指标包括住院时间延长＞7天以及入住重症监护病房（Intensive Care Unit, ICU）。最终共纳入727例PAS患者。基于产前特征的机器学习预测模型，在大量失血、住院时间延长及入住ICU三个结局中的曲线下面积（Area Under Curve, AUC）分别为0.84、0.81和0.82。该模型的重要预测因素包括产次、胎盘着床部位、诊断方法及产前血红蛋白水平。结合基线变量与围手术期特征构建的机器学习模型，在上述三项研究结局中的AUC分别为0.86、0.90和0.86。其中种族、盆腔侵犯情况及子宫切口类型是该模型中预测效能最强的因素。机器学习模型可用于评估PAS患者的个体化发病风险。基于模型的风险评估有助于<i>a priori</i>规划诊疗管理策略。