Construction and validation of a nomogram prediction model for the risk of gastrointestinal bleeding in hospitalized patients with coronary heart disease
- VernacularTitle:住院冠心病患者消化道出血发生风险列线图预测模型的构建与验证
- Author:
Yutao DING
1
;
Yuhang WEI
;
Rujun LI
;
Xin PAN
;
Yang GAO
Author Information
- Keywords: coronary heart disease; gastrointestinal bleeding; nomogram prediction model; inflammatory markers; nutritional status; antithrombotic therapy; calibration curve; decision curve analysis
- From: Journal of Clinical Medicine in Practice 2025;29(19):12-18
- CountryChina
- Language:Chinese
- Abstract: Objective To screen the independent influencing factors for gastrointestinal bleeding(GIB)in hospitalized patients with coronary heart disease(CHD)and to construct and validate a no-mogram prediction model.Methods A total of 440 CHD patients who developed GIB during hospi-talization were selected as GIB group,and another 320 CHD patients hospitalized in the department of cardiovascular medicine were randomly selected as non-GIB group.The clinical data of the two groups were analyzed and compared.Multivariate logistic regression analysis was used to screen the indepen-dentinfluencing factors for GIB.Based on these factors,a nomogram prediction model for the risk of GIB in hospitalized CHD patients was constructed.The entire dataset was randomly divided into train-ing set(n=532)and validation set(n=228)in a 7∶3 ratio.The performance of the nomogram model was evaluated using the receiver operating characteristic(ROC)curve,calibration curve,and decision curve analysis(DCA).Results Multivariate logistic regression analysis showed that body mass index(BMI),history of digestive system diseases,CHD classification,albumin,white blood cell count,monocyte-to-lymphocyte ratio(MLR),and low-density lipoprotein were all independent influencing factors for GIB in CHD patients(P＜0.05).ROC curve analysis indicated that the nomo-gram model(excluding low-density lipoprotein)constructed based on independent influencing factors exhibited good discrimination in both the training set(area under the curve:0.839,95％CI,0.805 to 0.873)and the validation set(area under the curve:0.810,95％CI,0.751 to 0.868).Calibration curve analysis demonstrated good consistency between the predicted probabilities and the observed incidence of GIB in hospitalized CHD patients in both the training and validation sets.DCA results revealed that the nomogram model had a good clinical net benefit.Conclusion The nomogram model constructed based on independent influencing factors has good predictive performance for the risk of GIB in hospitalized CHD patients and can provide a basis for clinicians to promptly identify GIB and adjust medication regimens.
