Department of Respiratory and Critical Care Medicine, West China Hospital, Sichuan University, No. 37 Guoxue Alley, Chengdu, 610041, Sichuan, China.
Laboratory of Pathology, Key Laboratory of Transplant Engineering and Immunology, NHC, West China Hospital, Sichuan University, No. 37 Guoxue Alley, Chengdu, 610041, Sichuan, China.
BMC Pulm Med. 2022 Aug 13;22(1):312. doi: 10.1186/s12890-022-02113-9.
Researchers have linked cardiovascular disease (CVD) with advancing age; however, how it drives disease progression in elderly severe community acquired pneumonia (SCAP) patients is still unclear. This study aims to identify leading risk predictors of in-hospital mortality in elderly SCAP patients with CVD, and construct a comprehensive nomogram for providing personalized prediction.
The study retrospectively enrolled 2365 elderly patients identified SCAP. Among them, 413 patients were found to have CVD. The LASSO regression and multivariate logistic regression analysis were utilized to select potential predictors of in-hospital mortality in elderly SCAP patients with CVD. By incorporating these features, a nomogram was then developed and subjected to internal validations. Discrimination, calibration, and clinical use of the nomogram were assessed via C-index, calibration curve analysis, and decision plot.
Compared with patients without CVD, elderly SCAP patients with CVD had a significant poor outcome. Further analysis of the CVD population identified 7 independent risk factors for in-hospital mortality in elderly SCAP patients, including age, the use of vasopressor, numbers of primary symptoms, body temperature, monocyte, CRP and NLR. The nomogram model incorporated these 7 predictors showed sufficient predictive accuracy, with the C-index of 0.800 (95% CI 0.758-0.842). High C-index value of 0.781 was obtained in the internal validation via bootstrapping validation. Moreover, the calibration curve indicative a good consistency of risk prediction, and the decision curve manifested that the nomogram had good overall net benefits.
An integrated nomogram was developed to facilitate the personalized prediction of in-hospital mortality in elderly SCAP patients with CVD.
研究人员已经将心血管疾病（CVD）与年龄增长联系起来；然而，它如何驱动老年重症社区获得性肺炎（SCAP）患者的疾病进展尚不清楚。本研究旨在确定患有 CVD 的老年 SCAP 患者住院死亡率的主要风险预测因素，并构建一个综合列线图以提供个性化预测。
本研究回顾性纳入了 2365 名确诊为 SCAP 的老年患者。其中，413 名患者患有 CVD。采用 LASSO 回归和多变量逻辑回归分析来选择患有 CVD 的老年 SCAP 患者住院死亡率的潜在预测因素。通过纳入这些特征，然后开发了一个列线图，并进行了内部验证。通过 C 指数、校准曲线分析和决策图评估列线图的区分度、校准度和临床应用。
与没有 CVD 的患者相比，患有 CVD 的老年 SCAP 患者的预后明显较差。对 CVD 人群的进一步分析确定了 7 个与老年 SCAP 患者住院死亡率相关的独立风险因素，包括年龄、使用血管加压药、主要症状数量、体温、单核细胞、CRP 和 NLR。纳入这 7 个预测因素的列线图模型显示出足够的预测准确性，C 指数为 0.800（95%CI 0.758-0.842）。通过 bootstrap 验证在内部分值验证中获得了较高的 C 指数值 0.781。此外，校准曲线表明风险预测具有良好的一致性，决策曲线表明列线图具有良好的整体净收益。
开发了一个综合列线图，以方便对患有 CVD 的老年 SCAP 患者的住院死亡率进行个性化预测。
