Department of Gynecology and Obstetrics, Atatürk University School of Medicine, Erzurum, Turkey.
Department of Biostatistics, Biruni University School of Medicine, Istanbul, Turkey.
Int J Gynaecol Obstet. 2021 Feb;152(2):256-261. doi: 10.1002/ijgo.13386. Epub 2020 Oct 28.
To analyze all the variables in women who received fibrinogen for postpartum hemorrhage (PPH) using hierarchical cluster analysis, to provide greater insight into the risk variables involved in these women.
This retrospective study of women with at least 500 mL of bleeding at birth or during the postpartum period and treated with fibrinogen was conducted at the Department of Obstetrics and Gynecology, Atatürk University School of Medicine from January 2013 to January 2018. Data on the women were obtained from medical records and various risk variables were recorded and analyzed using hierarchical cluster analysis.
A total of 114 women with PPH were included in the study. Based on a dendrogram, three main clusters of similar quality variables were created: 1) gravida, parity, age, cervical/vaginal hematoma, laparotomy, hypogastric artery ligation, uterine artery embolization, uterine artery ligation, uterine atony, distance from outer center, lowest hemoglobin, preoperative platelets, endometritis, preoperative white blood cells; 2) lowest fibrinogen, highest fibrinogen, type of birth, placenta invasion anomaly, Bakri balloon tamponade, postpartum hysterectomy, preoperative activated partial thromboplastin time (APTT), preoperative international normalized ratio (INR), placental abruption, in-utero ex fetus; 3) postoperative APTT, postoperative INR, maternal mortality, erythrocyte transfusion, plasma transfusion, hospital stay time, disseminated intravascular coagulation/HELLP syndrome, highest hemoglobin, blood group, postoperative platelets, platelet transfusion, pre-eclampsia/eclampsia, fibrinogen extract.
According to the cluster analysis, we should keep fibrinogen extract in the foreground especially in the treatment of hemorrhage in patients with variable conditions. As a result, we can determine whether fibrinogen extract, which has a high economic cost, should be kept at each center. We can also direct which patient will be referred in accordance with the referral steps.
通过层次聚类分析，对接受纤维蛋白原治疗产后出血（PPH）的女性的所有变量进行分析，以期深入了解这些女性所涉及的风险变量。
本研究为回顾性研究，对象为 2013 年 1 月至 2018 年 1 月期间在阿塔图尔克大学医学院妇产科分娩或产后至少出血 500ml 并接受纤维蛋白原治疗的女性。从病历中获取女性数据，并记录和分析各种风险变量，使用层次聚类分析。
本研究共纳入 114 例 PPH 患者。基于聚类树状图，创建了 3 个具有相似质量变量的主要聚类：1）产妇的孕次、产次、年龄、宫颈/阴道血肿、剖腹术、下腹动脉结扎术、子宫动脉栓塞术、子宫动脉结扎术、子宫收缩乏力、距外中心距离、最低血红蛋白、术前血小板、子宫内膜炎、术前白细胞；2）最低纤维蛋白原、最高纤维蛋白原、分娩方式、胎盘植入异常、Bakri 球囊填塞、产后子宫切除术、术前活化部分凝血活酶时间（APTT）、术前国际标准化比值（INR）、胎盘早剥、胎儿宫内死亡；3）术后 APTT、术后 INR、产妇死亡率、红细胞输注、血浆输注、住院时间、弥散性血管内凝血/HELLP 综合征、最高血红蛋白、血型、术后血小板、血小板输注、子痫前期/子痫、纤维蛋白原提取物。
根据聚类分析，我们应该特别关注纤维蛋白原提取物在不同情况下治疗出血的作用。因此，我们可以确定每个中心是否应该保留纤维蛋白原提取物，因为它的成本很高。我们还可以根据转诊步骤确定应将哪个患者转诊。
