Development and validation of a prediction model for 3-year mortality risk in patients with idiopathic pulmonary fibrosis
10.3760/cma.j.cn114798-20250703-00832
- VernacularTitle:特发性肺纤维化患者3年死亡风险预测模型的构建及效能评价
- Author:
Yan WANG
1
;
Xueting YUAN
1
;
Jin JIN
1
;
Xiaomao XU
1
Author Information
1. 北京医院呼吸与危重症医学科 国家老年医学中心 中国医学科学院老年医学研究院，北京 100730
- Publication Type:Journal Article
- Keywords:
Pulmonary fibrosis;
Comorbidity;
Mortality;
Predictive model
- From:
Chinese Journal of General Practitioners
2025;24(10):1232-1239
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To develop a 3-year mortality risk prediction model for patients with idiopathic pulmonary fibrosis (IPF) and evaluate its performance efficiency.Methods:This retrospective study enrolled consecutive patients with idiopathic pulmonary fibrosis (IPF) at Beijing Hospital between January 2013 and December 2021. Patient was followed for ≥3 years. Cox regression analyses were used to identify risk factors of 3-year mortality and a risk prediction model for mortality risk in patients with IPF was developed and evaluated.Results:A total of 204 patients were enrolled, among whom 60 cases died and 144 cases survived during the follow-up. Patients were randomly divided into a training set ( n=142) and a validation set ( n=62) in a 7∶3 ratio. Multivariate Cox regulation analysis revealed that Charlson Comorbidity Index (CCI) score ( HR=1.589,95% CI: 1.310-1.928, P<0.001, C-index=0.716), High Resolution CT (HRCT) reticular pattern ( HR=6.901, 95% CI: 2.763-17.239, P<0.001, C-index=0.752), HRCT honeycombing sign ( HR=3.126, 95% CI: 1.871-5.223, P=0.001, C-index=0.717), and HRCT traction bronchiectasis ( HR=3.875, 95% CI:2.190-6.858, P=0.001, C-index=0.711) were risk factors for 3-year mortality. A 3-year mortality risk model for IPF patients was developed: risk score=0.654×CCI+2.174×reticular pattern (Yes=1)+2.355×honeycombing (Yes=1)+0.511×traction bronchiectasis (Yes=1). The HR of the model for training set, validation set and overall cohort were 2.718 (95% CI: 1.930-3.828, P<0.001, C-index=0.880), 2.537 (95% CI: 1.255-5.131, P=0.010, C-index=0.853) and 2.590 (95% CI: 1.910-3.512, P<0.001, C-index=0.865), respectively. The performance was evaluated by ROC curve, the areas under the curve of the model for predicting 3-year mortality were 0.903 in training set and 0.826 in validation set, respectively. Conclusion:The prediction model composing of CCI score and HRCT signs developed in this study demonstrates a good performance for 3-year mortality risk in IPF patients.
