Study on the construction of nomogram prediction model for prognostic assessment of heart failure patients based on serological markers and echocardiography
F Yu,Z-X Wang,J Yu,F-J Hu,R-G Zhang,Y Yuan,W Yang
DOI: https://doi.org/10.26355/eurrev_202404_35913
Abstract:Objective: We aimed to construct a nomogram prediction model for prognostic assessment of patients with heart failure (HF) based on serological markers and echocardiography. Patients and methods: A total of 200 HF patients admitted to the Second Affiliated Hospital of Nanchang University from January 2018 to January 2020 were selected as the research objects. According to the New York Heart Association (NYHA) cardiac function classification, they were divided into 3 groups, including 65 cases of grade II, 97 cases of grade III, and 38 cases of grade IV. Three groups of echocardiographic parameters were compared [including left ventricular ejection fraction (LVEF), left ventricular end-diastolic diameter (LVEDD), left ventricular end-systolic diameter (LVESD), left ventricular end-systolic volume (LVESV)], differences in serum markers brain natriuretic peptide (BNP), soluble growth-stimulating expression gene 2 (sST2) and the Modified Early Warning Score (MEWS). The patients were divided into two groups according to their clinical outcomes during the follow-up period, including 52 cases in the death group and 148 cases in the survival group. The clinical data of the two groups were compared, and multi-factor logistic regression analysis was performed to screen out the independent risk factors affecting the patient's death. A nomogram model of the patient's mortality risk was constructed based on the independent risk factors. Receiver operating characteristic (ROC) curves and calibration curves were used to evaluate the discrimination and accuracy of the nomogram model. Results: As the cardiac function class of elderly chronic heart failure (CHF) patients increases, LVEDD, LVESD, sST2, and MEWS increase and LVEF decreases (p<0.05). Multifactor analysis results showed that LVEF, LVEDD, sST2, and MEWS were independent factors affecting the clinical outcome of patients. The AUCs predicted using LVEF, LVEDD, sST2, and MEWS alone were 0.738, 0.775, 0.717, 0.831, and 0.768, respectively. There is a certain degree of discrimination, and the model has extremely high accuracy. Conclusions: MEWS, LVEDD, and sST2 increase as the NYHA cardiac function grade of HF patients increases and LVEF decreases, which can reflect the severity of the disease to a certain extent. Additionally, the nomogram model established based on this has a high predictive value for the long-term prognosis of patients and can formulate effective intervention measures for quantitative values.