Predicting Six-Month Re-Admission Risk in Heart Failure Patients Using Multiple Machine Learning Methods: A Study Based on the Chinese Heart Failure Population Database

Figure 1. Study design. LASSO regression: least absolute shrinkage and selection operator regression; AUC: area under the curve; CI: confidence interval.

Figure 2. The results of three variable screening methods. RF = random forest; LR = logistic regression; LASSO = least absolute shrinkage and selection operator regression. (A): Venn presents the result of the variable selection. (B): A total of 10 variables were screened by logistic regression and LASSO regression, and eight variables were screened by random forest, six of which were common to the results of the three methods.

Figure 3. (A): The AUC of the 6 machine learning algorithms. (B): The calibration curve of the logistic regression model. The ideal line indicates that the model prediction is exactly the same as the actual situation, which is the ideal situation. Apparent and bias-corrected lines indicate the prediction performance of the LR model. Bias-corrected lines. Corrected over-fitting. (C): ROC analysis between the 12 clinical variables.

Figure 4. The nomogram of the logistic regression model. NYHA classification: New York Heart classification; ICU: intensive care unit. Points can calculate individual scores for each of the patient’s variables, and the total score is the result of adding the individual scores for each variable. Based on the total score, the probability of all-cause re-admission can be determined.