Abstract:

Objective To construct a nomogram prediction model for aspiration in ICU patients receiving enteral nutrition support，providing a reference for the early clinical identification of high-risk patients and the formulation of individualized prevention strategies. Methods By using the convenience sampling method，543 patients who received enteral nutrition support and were admitted to the ICU of a tertiary general hospital in Beijing from May 2021 to May 2025 were retrospectively selected. They were divided into a modeling group（n=380） and a validation group（n=163） at a ratio of 7 ∶ 3. Based on LASSO regression and Logistic regression to analyze the influencing factors of aspiration in ICU patients with enteral nutrition support，a nomogram was constructed and the receiver operating characteristic curve was drawn. The Hosmer-Lemeshow goodness-of-fit test was performed to evaluate the model. Results Aspiration occurred in 102 cases（26.84%） in the modeling group and 44 cases（26.99%） in the validation group.Age，the length of nasogastric tube insertion，whether there is tachycardia，white blood cell count，swallowing function，and nutritional risk are the influencing factors of aspiration in ICU patients with enteral nutrition support（all P<0.05）.The area under the operating characteristic curve of the subjects in the modeling group was 0.902，and that in the validation group was 0.921. The Hosmer-Lemeshow test results showed that χ²=4.445 and 8.502，both P>0.05. Conclusion Advanced age，excessively long nasogastric tube insertion length，combined tachycardia，high white blood cell count，poor swallowing function，and high nutritional risk are the influencing factors for aspiration in ICU patients receiving enteral nutrition support. The nomogram prediction model constructed based on the above factors has certain efficacy.

Key words: Intensive Care Unit, Enteral Nutrition, Aspiration, Risk Assessment, Prediction Models, Nomogram, Nursing Care

摘要：

目的 构建ICU肠内营养支持患者误吸发生的列线图预测模型，为临床早期识别高危患者、制订个体化预防策略提供参考。方法 采用便利抽样法，回顾性选取2021年5月至2025年5月北京市某三级甲等综合医院ICU收治的543例实施肠内营养支持的患者，依照7 ∶ 3将其分为建模组（n=380），验证组（n=163）。基于LASSO回归与Logistic回归分析ICU肠内营养支持患者发生误吸的影响因素，构建列线图并绘制受试者操作特征曲线，行Hosmer-Lemeshow拟合优度检验来评估模型的效能。结果 建模组有102例（26.84%）、验证组有44例（26.99%）发生误吸。年龄、鼻饲管置入长度、是否合并心动过速、白细胞计数、吞咽功能评分、营养风险评分是ICU肠内营养支持患者发生误吸的影响因素（均P<0.05）；建模组受试者操作特征曲线下面积为0.902，验证组为0.921；Hosmer-Lemeshow检验结果显示，χ²=4.445、8.502，均P>0.05。结论 年龄大、鼻饲管置入长度过长、合并心动过速、白细胞计数高、吞咽功能差、营养风险高是ICU肠内营养支持患者发生误吸的影响因素，基于上述因素构建的列线图预测模型具有一定效能。

关键词: 重症监护室, 肠内营养, 误吸, 风险评估, 预测模型, 列线图, 护理