Department of Obstetrics and Gynecology, Villalba General Hospital, Camino de Moralzarzal M-608 Km, Calle Alpedrete 41, 28400 Collado Villalba, Madrid, Spain.
Department of Applied mathematics, Escuela Universitaria Politécnica de La Almunia, Universidad de Zaragoza, Calle Mayor 5, 50100, La Almunia de Doña Godina, Zaragoza, Spain.
Eur J Obstet Gynecol Reprod Biol. 2020 Oct;253:238-248. doi: 10.1016/j.ejogrb.2020.08.044. Epub 2020 Aug 27.
To develop fetal growth standards for twin gestations by placental chorionicity in a Spanish population and compare them with European and American standards to estimate the suitability of their use in clinical practice.
This was a retrospective cohort study of 518 twin pregnancies, 435 dichorionic-diamniotic and 83 monochorionic-diamniotic, performed between January 2012 and December 2017. A total of 4,783 and 1,455 estimated fetal weights were considered from the 17 to the 37 week of gestation, using multilevel models, to build dichorionic-diamniotic and monochorionic-diamniotic standards, respectively. The percentages of small and large for gestational age were calculated as a model adjustment measure and adjustment to the studied data and the values provided by our model were compared against those of six European and American twin standards and three singleton standards. Correlation analyses between percentile predictions were performed using Cohen kappa coefficient. The predictive ability to detect small for gestational age was also provided by the sensitivity and positive predictive value.
We found slight differences between standards by chorionicity, being dichorionic-diamniotic percentiles slightly higher than monochorionic-diamniotic ones from the 17 to 37 weeks' gestation. For dichorionic-diamniotic cases, both our standard (9.8-8.2) and that of Grantz (8.2-10.5) showed good adjustments for the 10 and 90 percentiles while the other compared standards underestimated or overestimated them. For monochorionic-diamniotic cases, both our standard (10.2-8.5) and that of Shivkumar (11.4-6.8) had the most suitable adjustment. The correlation analysis between small and large for gestational age cases provided by standards, showed clear differences among them. Kappa's coefficient showed a substantial agreement between both Ananth (0.7) and Stirrup (0.69) dichorionic-diamniotic cases and our standard. There was also a substantial agreement between the Shivkumar (0.77) standard and our results for monochorionic-diamniotic cases. The correlation was moderate for all other comparisons.
Our model showed a good adjustment to the studied population. There are clear differences among small and large for gestational age cases provided by twin standards in our studied population. The twin growth standards depend on the population characteristics and model structure. We found the use of singleton standards for twin pregnancies inadequate.
通过胎盘绒毛膜性在西班牙人群中制定双胎妊娠的胎儿生长标准，并与欧洲和美国的标准进行比较，以评估其在临床实践中的适用性。
这是一项回顾性队列研究，纳入了 2012 年 1 月至 2017 年 12 月期间的 518 例双胎妊娠，其中 435 例为双绒毛膜双羊膜囊双胎，83 例为单绒毛膜双羊膜囊双胎。使用多层模型分别对 17 周至 37 周的 4783 例和 1455 例估计胎儿体重进行分析，以建立双绒毛膜双羊膜囊和单绒毛膜双羊膜囊的标准。小胎龄儿和大胎龄儿的百分比作为模型调整指标和对研究数据的调整，并将模型提供的值与六个欧洲和美国的双胎标准以及三个单胎标准进行比较。使用 Cohen kappa 系数进行百分位预测的相关性分析。通过灵敏度和阳性预测值也提供了对小胎龄儿的检测能力。
我们发现不同绒毛膜性的标准之间存在细微差异，从 17 周到 37 周，双绒毛膜双羊膜囊的百分位值略高于单绒毛膜双羊膜囊。对于双绒毛膜双羊膜囊病例，我们的标准（9.8-8.2）和 Grantz 的标准（8.2-10.5）在第 10 和 90 百分位值上都有很好的调整，而其他比较标准则低估或高估了这些百分位值。对于单绒毛膜双羊膜囊病例，我们的标准（10.2-8.5）和 Shivkumar 的标准（11.4-6.8）的调整最为合适。标准提供的小胎龄儿和大胎龄儿病例的相关性分析显示，它们之间存在明显差异。Kappa 系数显示，Ananth（0.7）和 Stirrup（0.69）的双绒毛膜双羊膜囊病例与我们的标准之间存在显著一致性。Shivkumar（0.77）标准与我们的单绒毛膜双羊膜囊病例结果之间也存在显著一致性。对于所有其他比较，相关性都是适度的。
我们的模型对研究人群有很好的调整。在我们研究的人群中，不同标准提供的小胎龄儿和大胎龄儿病例之间存在明显差异。双胎生长标准取决于人群特征和模型结构。我们发现将单胎标准用于双胎妊娠是不充分的。
