Page 96 - 《中国药房》2026年4期
P. 96
Boosting Machine, Random Forest, and Categorical Boosting models were constructed. Model performance was evaluated using
metrics including the area under the curve (AUC) of receiver operating characteristic curve and recall rate. The Shapley Additive
exPlanations (SHAP) method was applied to analyze and interpret the contribution of each variable. A nomogram was constructed
based on the optimal model. RESULTS A total of 38 trigger items for active monitoring of bevacizumab-related ADR were
determined, comprising 17 laboratory indicators, 13 clinical manifestations, and 8 intervention measures. In total, 483 patients
with positive trigger items were included, and 318 patients with bevacizumab-induced ADR were identified, including 83 SARs.
The positive predictive values for the trigger items and cases were 43.57% (708/1 625) and 63.84% (318/483), respectively.
Bevacizumab-induced ADR involved 7 systems/organs, with the hematological system being the most frequently involved
(64.15%). The Boruta algorithm selected 7 variables: serum potassium, hematocrit, albumin-to-globulin ratio, prealbumin,
hypertension history, age and red blood cell count. Multivariable Logistic regression showed that elevated serum potassium levels
were associated with a decreased risk of bevacizumab-induced SAR (OR=0.234, P=0.002), while a history of hypertension
(OR=2.642, P=0.006) and increased age (OR=1.040, P=0.025) were associated with an increased risk. The Logistic
Regression model demonstrated superior performance with higher AUC, F1 score and recall rate (0.761, 0.447, 0.607), compared
to other models. SHAP evaluation results indicated that variables such as serum potassium, hematocrit, and age ranked highest in
importance. CONCLUSIONS Totally 38 trigger entries have been successfully identified for active screening of bevacizumab-
related ADR. Elevated serum potassium levels are a protective factor against bevacizumab-induced SAR, whereas the hypertension
history and increased age are risk factors. The Logistic Regression model is the optimal predictive model.
KEYWORDS bevacizumab; adverse drug reaction; global trigger tool; machine learning; predictive model
癌症是导致疾病发生和死亡的重要原因。在全球 1 资料与方法
范围内,我国新发癌症及死亡病例数的占比均较为突 1.1 触发器条目确定
出,给卫生系统和患者家庭造成了沉重的经济负担 [1―2] 。 依据GTT白皮书、药品说明书及相关文献 [8―11] ,结合
近年来,分子靶向治疗发展迅速,贝伐珠单抗(bevaci‐ 桂林市人民医院(以下简称“我院”)实际情况,本研究初
zumab)作为代表性抗血管生成单克隆抗体,可特异性结 步拟定45项触器发条目,涵盖检验指标、临床表现、干预
合血管内皮生长因子A,阻断其与血管内皮生长因子受 措施三大类。通过单轮德尔菲法就上述初拟条目向8名
体2结合,进而抑制肿瘤细胞新生血管形成,有效延缓癌 专家进行函询,并采用专家打分法计算各条目的重要性
[3]
症进展 。然而,贝伐珠单抗可引发蛋白尿、出血、消化 均值(Mj )、专家权威系数(Cr )、变异系数(Vj ),以评估专
[12]
道穿孔等严重不良反应(severe adverse reaction,SAR), 家意见的一致性与可靠性 。最终以 Vj>0.35 且 Mj<
[12]
其机制主要涉及血管舒张障碍、肾脏滤过屏障受损以及 3.5为标准 ,确定贝伐珠单抗相关ADR触发器条目。
[4]
血管修复功能减弱 ;同时,贝伐珠单抗相关SAR还可能 1.2 病例筛选
本研究采用回顾性研究方法,在中国医院药物警戒
导致患者治疗中断或住院时间延长,亟须临床加强
系统(China Hospital Pharmacovigilance System,CHPS)
识别。
中检索 2020 年 1 月 1 日至 2024 年 9 月 10 日于我院住院
目前,药物不良反应(adverse drug reaction,ADR)监
并使用贝伐珠单抗患者的相关记录。具体操作如下:依
测主要依赖自发上报系统,但该系统存在上报率低、信
[5]
息不完整、易遗漏等不足 。全面触发工具(global trig‐ 据“1.1”项下确定的触发器条目,药师在 CHPS 的“药品
评价系统”模块中完成规则配置及关键字段术语匹配;
[6]
ger tool,GTT)作为可主动识别ADR的工具之一 ,已广
随后,CHPS 基于医嘱、病程及检验信息自动筛查、关联
泛应用于抗真菌药、抗肿瘤药物的相关研究 [5,7] ,但其用
并导出相关病历资料。本研究患者的纳入标准包括:
于如贝伐珠单抗等具体药物安全性监测的报道相对缺
(1)接受贝伐珠单抗治疗的所有住院患者;(2)年龄≥18
乏。随着电子病历与临床大数据的积累,机器学习技术
岁;(3)触发器条目阳性;(4)住院时间≥24 h。本研究患
在 ADR 监测、安全风险评估方面展现出较高的建模灵
者的排除标准包括:(1)主要住院信息缺失者;(2)经核
活性和良好的预测性能,可在传统统计学分析基础上进 实未使用贝伐珠单抗者。本研究方案已获我院伦理委
一步强化临床对高危患者的识别能力。基于此,本研究 员会批准(批准号为2021-033KY)。
首先结合GTT理念确定贝伐珠单抗相关ADR触发器条 1.3 ADR判定与分级
目,据此筛选触发器条目阳性的患者病历;随后,通过人 针对病历资料,由 2 名初级职称药师判断是否发生
工审核判定ADR并描述其临床特征;最后,挖掘贝伐珠 ADR;若存在分歧,则由高级职称药师复核裁决。具体
单抗致 SAR 的危险/保护因素,构建并优选机器学习模 操作如下:依据WHO-乌普萨拉监测中心制定的ADR判
型,将最优模型以列线图形式呈现,以期提高临床对该 定标准,对可疑ADR(触发器条目阳性)与贝伐珠单抗的
药相关 ADR 的早期识别及风险评估能力,为临床优化 因果关系(包括“肯定”“很可能”“可能”“不太可能”“待
[13]
治疗方案提供参考。 定”“无法评价”)进行评估 ;随后,结合不良事件通用
· 498 · China Pharmacy 2026 Vol. 37 No. 4 中国药房 2026年第37卷第4期

