目的：探讨开角型青光眼患者并发认知功能障碍的影响因素，并构建决策树模型。

方法：回顾性研究。收集2022年2月至2024年11月在本院治疗的开角型青光眼患者的临床资料，根据患者入院时蒙特利尔认知评估量表(MoCA)评估结果分为认知功能障碍组和认知功能正常组。比较两组患者临床资料。采用Logistic回归模型分析开角型青光眼患者并发认知功能障碍的影响因素，同时基于两组临床资料利用决策树CHAID算法分析构建决策树模型，并采用受试者工作特征(ROC)曲线比较两种模型的预测效能。

结果：本研究共纳入开角型青光眼患者179例，根据患者入院时MoCA评估结果分为认知功能障碍组107例(男59例、女48例，年龄≥60岁者66例)和认知功能正常组72例(男34例、女38例，年龄≥60岁者28例)。认知功能障碍发生率为59.8%(107/179)。认知功能障碍组年龄≥60岁、受教育程度初中及以下、高血压、睡眠障碍和疾病中/晚期占比均高于认知功能正常组(均P<0.05)。Logistic回归分析显示，年龄≥60岁、受教育程度初中及以下、高血压、睡眠障碍、疾病中/晚期均为开角型青光眼患者并发认知功能障碍的危险因素(均P<0.05)。采用决策树CHAID算法构建的决策树模型共4层，11个节点，输出疾病严重程度、年龄、睡眠障碍、受教育程度、高血压共5个风险变量，其中疾病严重程度是首层风险变量。ROC曲线显示，决策树模型和Logistic回归模型预测开角型青光眼患者并发认知功能障碍的曲线下面积分别为0.849、0.842，二者预测价值比较无差异(P>0.05)。

结论：疾病严重程度、年龄、睡眠障碍、受教育程度、高血压是开角型青光眼患者并发认知功能障碍的影响因素，且由决策树CHAID算法分析构建的决策树模型对认知功能障碍具有良好的预测价值。

METHODS:Retrospective study. The clinical data of patients with open angle glaucoma treated in the hospital from February 2022 to November 2024 were collected, and patients were divided into a cognitive dysfunction group and a cognitive function normal group according to the evaluation results of the Montreal Cognitive Assessment Scale(MoCA)at the time of admission, and the clinical data of the two groups were compared. The influencing factors of cognitive dysfunction in patients with open angle glaucoma were analyzed through Logistic regression model, and the decision tree model was analyzed and constructed based on the clinical data of the two groups through the decision tree CHAID algorithm, and the predictive performance of two models was compared using receiver operating characteristic(ROC)curves.

RESULTS:The total of 179 patients with open angle glaucoma were included in this study, and were divided into a cognitive dysfunction group of 107 cases(59 males and 48 females, with 66 cases aged ≥60 y)and a cognitive function normal group of 72 cases(34 males and 38 females, with 28 cases aged ≥60 y)according to the MoCA assessment results at the time of patient admission.The incidence of cognitive dysfunction was 59.8%(107/179). The proportion of age ≥60y, education level of junior high school or below, hypertension, sleep disorders, and middle/late stage diseases in the cognitive dysfunction group were higher than those in the cognitive function normal group(all P<0.05). Logistic regression analysis showed that age ≥60 y, education level of junior high school or below, hypertension, sleep disorders, and middle/late stage disease were all risk factors for cognitive dysfunction in patients with open angle glaucoma(all P<0.05). The risk prediction decision tree model constructed using the decision tree CHAID algorithm included 4 layers and 11 nodes, and it outputed 5 risk variables: disease severity, age, sleep disorders, education level, and hypertension. Among them,the disease severity was the first layer risk variable. The ROC curve showed that the area under curve for predicting cognitive dysfunction in patients with open angle glaucoma using the risk prediction decision tree model and logistic regression model was 0.849 and 0.842 respectively, and there was no statistically significant difference in the predictive value between the two models(P>0.05).

CONCLUSION: The disease severity, age, sleep disorders, education level, and hypertension are influencing factors for cognitive dysfunction in patients with open angle glaucoma, and the risk prediction decision tree model analyzed and constructed by the decision tree CHAID algorithm has good predictive value for cognitive dysfunction.