Development and validation of a BMI stratified mortality prediction model for patients with COPD complicated by HF using the MIMIC-IV database.

UNLABELLED

Given the high mortality rate of chronic obstructive pulmonary disease (COPD) complicated by heart failure (HF), early identification of high-risk patients and timely intervention are crucial. There is currently no in-hospital mortality risk prediction model for COPD complicated by HF patients with different Body Mass Index (BMI). This study aims to explore the risk factors of COPD complicated by HF and construct an in-hospital mortality risk prediction model.

METHOD

Select a population that meets the diagnostic criteria for COPD complicated by HF from the Medical Information Mart for Intensive Care IV (MIMIC-IV) and analyze the baseline characteristics of the patients. Univariate Cox regression analysis and multivariate Cox regression analysis were used to determine the risk factors for mortality in patients with different BMIs and to construct a prediction model. Evaluate the model's consistency, discriminability, and clinical application value using the calibration curve, area under the curve (AUC), and decision curve analysis (DCA), respectively.

RESULT

A total of 907 patients with COPD complicated by HF were included, and risk factors such as age, heart rate (HR), systolic blood pressure (SBP), diastolic blood pressure (DBP), white blood cell count (WBC), heart rate(HR), respiratory rate (RR), blood urea nitrogen (BUN), prothrombin time (PT), activated partial thromboplastin time (aPTT), diabetes, peripheral vascular disease, sequential organ failure assessment (SOFA), and Glasgow Coma Scale(GCS) were included in the prediction model. AUC, calibration, and decision curves indicate that most models have good discrimination, calibration, and clinical application value.

CONCLUSION

The in-hospital mortality risk prediction model for COPD complicated by HF based on MIMIC-IV has good recognition ability and significant clinical reference value for patient prognosis risk assessment and intervention treatment.