Xin yixue (Aug 2022)

Establishment of a risk scoring model to predict intraoperative massive hemorrhage in infants with intracranial tumors

  • Liu Jianhua, Zhang Na, Xu Yingyi, Wei Wei, Chen Cheng, Tan Yonghong

DOI
https://doi.org/10.3969/j.issn.0253-9802.2022.08.008
Journal volume & issue
Vol. 53, no. 8
pp. 575 – 581

Abstract

Read online

Objective To analyze the factors influencing intraoperative massive hemorrhage in infants with intracranial tumors and establish a bleeding risk scoring （BRS） model. Methods Clinical data of 278 infants who underwent elective intracranial tumor resection were retrospectively analyzed. According to the ratio of blood loss （BL） to estimated blood volume （EBV）, all infants were divided into the slight or moderate hemorrhage group （BL/EBV≤0.5） and massive hemorrhage group （BL/EBV > 0.5）. Relevant clinical data and hemorrhage were statistically analyzed. Univariate and multivariate logistic regression analyses were used to identify the independent risk factors for intraoperative massive hemorrhage and establish a BRS model. Forty-five infants undergoing intracranial tumor resection were recruited in this study. The receiver operating characteristic （ROC） curve was delineated to evaluate the performance of this model. Results One hundred and eighty-three infants were assigned into in the slight or moderate hemorrhage group and 95 in the massive hemorrhage group. Logistic regression analysis demonstrated that body weight ≥12 kg, tumor diameter ≥4.5 cm,and operation time ≥ 5 h were the independent risk factors for intraoperative massive hemorrhage （all P < 0.05）. The score of tumor diameter ≥4.5 cm was set as 5, and 2 for operation time≥5 h. The total score of infant was calculated as BRS. The area under the ROC curve （AUC） of BRS model in the retrospective population was 0.788. If the critical value for massive hemorrhage was determined as 5, the sensitivity was 66.8%, the specificity was 81.3%, and the accuracy was 71.0%, while in the prospective population, the ROC AUC was 0.909, the sensitivity, specificity and accuracy were 78.6%, 88.9% and 87.0%, respectively. Conclusions Preoperative tumor diameter≥4.5 cm and operation time≥5 h are the independent risk factors for intraoperative massive hemorrhage in infants undergoing intracranial tumor resection. The BRS model established based on these factors has high predictive ability for intraoperative massive hemorrhage in infants with intracranial tumors.

Keywords