| 巨恬,刘自双,薛慧,等.急性脑卒中病人肺部感染危险因素分析及列线图模型构建和验证[J].安徽医药,2025,29(10):2013-2017. |
| 急性脑卒中病人肺部感染危险因素分析及列线图模型构建和验证 |
| Analysis of risk factors for pulmonary infection in patients with acute stroke and construction and validation of nomogram model |
| |
| DOI：10.3969/j.issn.1009-6469.2025.10.020 |
| 中文关键词: 卒中 呼吸道感染 列线图 预测模型 危险因素 |
| 英文关键词: Stroke Respiratory tract infections Nomogram Prediction model Risk factors |
| 基金项目: |
|
| 摘要点击次数: 40 |
| 全文下载次数: 46 |
| 中文摘要: |
| 目的探索急性脑卒中(AS)病人发生肺部感染的危险因素，并构建列线图风险预测模型。方法回顾性分析 2021年 1月至 2024年 5月首都医科大学附属北京康复医院收治的 502例 AS病人的病例资料。将数据集按照 7∶3随机分为训练集(352例)和验证集(150例)。对训练集病例资料进行单因素和多因素 logistic回归分析，筛选出 AS病人发生肺部感染的独立影响因素，采用 R软件制作列线图，用验证集中样本进行验证，模型区分度评估采用一致性指数(C‐index)，校准度评估采用校准图。结果 502例 AS病人中肺部感染 95例(18.9%)较于对照组，感染组病人的年龄 ≥70岁、吸烟史、合并糖尿病、卧床时间 ≥7 d、吞咽障碍、意识障碍、侵入性操作和 ALB<30 g/L的病，人比例显著增高［70.31%比 46.53%，59.38%比 44.79%，37.50%比 23.96%，62.50%比 48.26%，57.81%比 27.78%，45.31%比 33.33%，60.94%比 34.38%，70.31%比 51.04%］均差异有统计学意义(P<0.05)多因素分析显示年龄［OR=2.68，95%CI:(1.60，4.51)］、合并糖尿病［OR=2.12，95%CI:(1.24，3.63、吞咽障碍［OR=3.19，95%CI3，5.26)］意识障碍［OR=1.64，95%CI:(0.98，2.78)］、侵入性操作［OR=3.22，95%CI:(1.93，5.35)］和血清白蛋白［OR=0.50，95%CI:(0.30)，:(1.9，，、0.83)］是 AS病人中肺部感人发生的最佳预测因子。预测模型的 C-index为 0.79(0.70，0.84)，特异度和灵敏度分别为 82.4%和 60.6%。外部验证的 C-index为 0.74，校准曲线显示模型校准度较好。结论该研究构建的急性脑卒中病人肺部感染列线图风险预测模型区分度与校准度较好，可为临床制订肺部感染防控方案和控制感染率提供参考价值。 |
| 英文摘要: |
| Objective To explore the risk factors for pulmonary infection in patients with acute stroke (AS) and to construct a nomo-gram risk prediction model.Methods A retrospective analysis was conducted on the medical records of 502 AS patients admitted toBeijing Rehabilitation Hospital, Capital Medical University, from January 2021 to May 2024. The dataset was randomly divided into atraining set (352 cases) and a validation set (150 cases) in a 7∶3 ratio. Univariate and multivariate logistic regression analyses were per-formed on the training set to identify independent factors influencing pulmonary infection in AS patients. A nomogram was developedusing R software, and its performance was validated using samples from the validation set. The model's discrimination was assessed us-ing the concordance index (C-index), and calibration was evaluated using calibration plots.Results Among 502 patients with bronchi-ectasis, 95 cases (18.9%) were diagnosed with pulmonary infection. Compared to the control group, patients in the infected group exhib-ited a significantly higher prevalence of several risk factors, including age ≥ 70 years, smoking history, concomitant diabetes, bedrestduration ≥ 7 days, dysphagia, consciousness disturbance, invasive procedures, and serum albumin (ALB) levels < 30 g/L. Specifically,the proportions were as follows: age ≥ 70 years (70.31% vs. 46.53%), smoking history (59.38% vs. 44.79%), diabetes comorbidity (37.50% vs. 23.96%), bedrest ≥7 days (62.50% vs. 48.26%), dysphagia (57.81% vs. 27.78%), consciousness disturbance (45.31% vs. 33.33%), invasive procedures (60.94% vs. 34.38%), and ALB < 30 g/L (70.31% vs. 51.04%). All differences were statistically signifi-cant (P < 0.05). Multivariate analysis indicated that age [OR=2.69, 95%CI: (1.60, 4.51)], presence of diabetes mellitus [OR=2.12, 95%CI: (1.24, 3.63)], dysphagia [OR=3.19, 95%CI: (1.93, 5.26)], consciousness disturbance [OR=1.64, 95%CI: (0.98, 2.78)], invasive procedures [OR=3.22, 95%CI: (1.93, 5.35)], and serum albumin levels [OR=0.50, 95%CI: (0.30, 0.83)] were the most significant predic-tive factors for pulmonary infection in patients with AS. The C-index of the prediction model was [0.79 (0.70, 0.84)], with a specificity of82.4% and a sensitivity of 60.6%. The external validation yielded a C-index of 0.74, and the calibration curve demonstrated that the model had good calibration.Conclusion The nomogram risk prediction model for pulmonary infection in patients with acute strokepresent strong discrimination and calibration, offering valuable insights for clinical strategies in infection prevention and control. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
