Development of a risk prediction model for cancer-related cognitive impairment in lung cancer patients from the perspective of precision health in nursing science
10.3760/cma.j.cn211501-20240914-02513
- VernacularTitle:护理科学精准健康模型视角下的肺癌患者癌症相关认知障碍风险预测模型的构建
- Author:
Xiaoyu XU
1
;
Lei YE
;
Fangmei CHEN
;
Pan GAO
;
Guanghui XIA
Author Information
1. 南京医科大学附属脑科医院重症医学科，南京 210029
- Publication Type:Journal Article
- Keywords:
Lung neoplasms;
Cancer-related cognitive impairment;
Precision health;
Predictive modelling;
Biomarkers
- From:
Chinese Journal of Practical Nursing
2025;41(14):1063-1071
- CountryChina
- Language:Chinese
-
Abstract:
Objective:Predictive modelling of risk of cancer-related cognitive impairment (CRCI) in lung cancer patients from the perspective of precision health in nursing science.Methods:Prospectively collected lung cancers treated in the Department of Respiratory Medicine of Affiliated Nanjing Brain Hospital, Nanjing Medical University from October 2023 to April 2024 as study subjects by a convenience samphing method. Lasso regression was used to screen the characteristic variables and construct the prediction model, and the predictive ability was evaluated by the AUC of the subjects′operating characteristics; Bootstrap resampling (1 000 times) internal validation of the model; the Hosmer-Lemeshow goodness-of-fit test was performed and the calibration curve was plotted to evaluate the calibration of the model; the clinical validity of the model was evaluated by decision cure analysis (DCA).Results:A total of 142 patients with lung cancer were included, 94 males and 38 females. The incidence of CRCI in lung cancer patients was 69.7%(99/142). Lasso regression showed that age(≥65), education, tumor stage, serum albumin, and PLR were independent risk factors for CRCI (coefficients of 0.372 048 72, - 0.361 265 78, 0.068 728 00, - 0.039 940 32, 0.001 639 92 respectively). The model AUC was 0.874 (95% CI: 0.815-0.933), with a sensitivity of 0.768, and a specificity of 0.860; the H-L goodness-of-fit test showed good agreement ( χ2 = 4.51, P>0.05), and Bootstrap re-sampling internal validation showed an AUC of 0.826. Calibration curves showed good agreement and accuracy between the model predicted probabilities and the actual observed probabilities. DCA showed that the model had clinical benefit when the threshold probability was approximately>25%. Conclusions:The CRCI column-line diagram risk model constructed in this study has good predictive efficacy and can effectively predict the occurrence of CRCI in lung cancer patients.
