Supplementary Material for: Role of cerebroplacental ratio in predicting the outcome of pregnancies complicated by diabetes

Introduction: Our objective was to evaluate the strength of association and diagnostic performance of cerebroplacental ratio (CPR) in predicting the outcome of pregnancies with pre- and gestational diabetes mellitus. Methods: Pubmed, Embase, Cochrane and Google Scholar databases were searched. Inclusion criteria were pregnancies with gestational or pre-gestational diabetes undergoing assessment of CPR. The primary outcome was a composite score of perinatal mortality and morbidity as defined by the original publication. Secondary outcomes included preterm birth (PTB), gestational age (GA) at birth, mode of delivery, birthweight, perinatal death (PND), Apgar score < 7 at 5 minutes, abnormal acid-base status, neonatal hypoglycemia, admission to neonatal intensive care unit (NICU) Head-to-head meta-analyses were used to directly compare the risk of each of the explored outcomes. For those outcomes found to be significant, computation of diagnostic performance of CPR was assessed using bivariate model. Results: Six studies (2743 pregnancies) were included. The association between low CPR and adverse composite perinatal outcome was not statistically significant(p= 0.096. In pregnancies complicated by GDM, fetuses with a low CPR had a significantly higher risk of birthweight <10th percentile (OR: 5.83, 95% CI 1.98-17.12) and this association remains significant when using a CPR <10th centile (p<0.001). Fetuses with low CPR had also a significantly higher risk of perinatal death (PND) (OR: 6.15 p<0.001) and admission to neonatal intensive care unit (NICU) (OR 3.32 , p<0.001), but not of respiratory distress syndrome (RDS) (p= 0.752), Apgar score <7 at 5 minutes (p=0.920), abnormal acid base status (p= 0.522) or neonatal hypoglycemia (p= 0.005). However, CPR showed a low diagnostic accuracy for detecting perinatal outcomes. Conclusion: CPR is associated but not predictive of adverse perinatal outcome in pregnancies complicated by gestational diabetes. The findings from this systematic review do not support the use of CPR as a universal screening for pregnancy complication in women with diabetes.

引言：本研究旨在评估脑胎盘比率（cerebroplacental ratio, CPR）在预测孕前糖尿病及妊娠糖尿病孕妇妊娠结局的关联强度与诊断效能。方法：检索PubMed、Embase、Cochrane图书馆及Google Scholar数据库。纳入标准为接受CPR评估的孕前糖尿病或妊娠糖尿病妊娠病例。主要结局为原始文献定义的围产期死亡率与患病率复合终点；次要结局包括早产（preterm birth, PTB）、分娩时孕周（gestational age, GA）、分娩方式、出生体重、围产期死亡（perinatal death, PND）、5分钟Apgar评分<7分、酸碱平衡异常、新生儿低血糖及新生儿重症监护病房（neonatal intensive care unit, NICU）收治。采用头对头Meta分析直接比较各探索结局的风险；对于存在统计学意义的结局，采用双变量模型评估CPR的诊断效能。结果：最终纳入6项研究，共2743例妊娠病例。低CPR与不良围产期复合结局的关联无统计学意义（p=0.096）。在合并妊娠糖尿病（GDM）的妊娠中，低CPR胎儿的出生体重低于第10百分位数的风险显著升高（比值比OR=5.83，95%置信区间CI：1.98~17.12）；当采用CPR<第10百分位数作为截断值时，该关联仍具有统计学意义（p<0.001）。低CPR胎儿的围产期死亡（PND）风险（OR=6.15，p<0.001）及新生儿重症监护病房（NICU）收治风险（OR=3.32，p<0.001）均显著升高，但与呼吸窘迫综合征（RDS）发生风险（p=0.752）、5分钟Apgar评分<7分（p=0.920）、酸碱平衡异常（p=0.522）及新生儿低血糖（p=0.005）无显著关联。不过，CPR在预测围产期结局方面的诊断准确性较低。结论：在合并妊娠糖尿病的妊娠中，CPR与不良围产期结局存在关联，但无法对其进行有效预测。本系统评价的结果不支持将CPR作为糖尿病女性妊娠并发症的通用筛查手段。