| 曾勍,朱锐.福州社区糖尿病视网膜病变发生预测模型的构建.[J].中南医学科学杂志.,2021,(6):735-737. |
| 福州社区糖尿病视网膜病变发生预测模型的构建 |
| Prediction model construction of diabetic retinopathy in Fuzhou community |
| 投稿时间：2021-02-11 修订日期：2021-07-27 |
| DOI：10.15972/j.cnki.43-1509/r.2021.06.030 |
| 中文关键词: 福州社区 糖尿病视网膜病变 预测模型 预测效能 |
| 英文关键词:Fuzhou community diabetic retinopathy prediction model prediction efficiency |
| 基金项目:福州市科技计划项目(2018-S-100-3) 作者简介:曾勍,硕士,主治医师,研究方向为视网膜病变与眼底疾病的临床诊断与治疗,E-mail为zr13860609862@sina.com。通信作者朱锐,硕士,主任医师,研究方向为视网膜病变与眼底疾病的临床诊断与治疗,E-mail为52625694@qq.com。 |
|
| 摘要点击次数: 423 |
| 全文下载次数: 379 |
| 中文摘要: |
| 目的探讨福州社区糖尿病视网膜病变(DR)发生预测模型的构建。方法选取糖尿病患者2 000例,采用TRC-50DX眼底彩色机拍摄,根据患者是否发生DR分为DR组和非DR组。采用单因素及多因素Logistic分析两组病例资料。利用R软件建立预测福州社区DR发生的列线图模型；绘制ROC曲线,分析该模型在福州社区DR发生中的预测效能。结果福州社区DR发生率与年龄、病程、总胆红素、纤维蛋白原(FIB)、合并高血压有关(P＜0.05)。列线图模型成功建立并预测各因素共存时发生DR的概率为92.0%,DR发生的C-index为0.816,具有良好的区分度；ROC曲线下面积为0.832,95%CI为0.757～0.899。结论福州社区DR发生率较高,构建的预测模型具有良好的区分度,能获得较高的预测效能。 |
| 英文摘要: |
| To explore the prediction model construction of diabetic retinopathy(DR)in Fuzhou community. Methods2 000 diabetic patients were selected and photographed by TRC-50DX fundus color machine. According to the occurrence of DR, the patients were divided into DR group and non-DR group. Univariate and multivariate Logistic analysis were used to analyze the case data of the two groups. R software was used to establish a line graph model to predict the occurrence of DR in Fuzhou community. ROC curve was drawn to analyze the predictive efficiency of the model in the occurrence of DR in Fuzhou community. ResultsThe incidence of DR in Fuzhou community was statistically significant with age, course of disease, total bilirubin, fibrinogen (FIB) and hypertension (P＜0.05). The linear graph model was established successfully, and the probability of DR occurrence was 92.0% when all factors co-existed. The C-index of DR was 0.816, showing good differentiation. The area under the curve was 0.832, and the 95%CI was 0.757－0.899. ConclusionThe DR incidence rate in Fuzhou community is high, and the constructed prediction model has good discrimination and can obtain high prediction efficiency. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|