Prediction of cardiovascular event risk within 90 days in patients with hypertension: A study based on random forest and nomogram models.

This study aimed to investigate risk factors for adverse cardiovascular events occurring within 90 days of onset in patients with hypertension. We developed a random forest (RF) predictive model and a nomogram predictive model, and validated the performance of these models. Baseline data were retrospectively collected from patients with hypertension treated at our hospital from January 2024 to December 2025. Disease-related risk factors were identified using univariate and multivariate logistic regression analysis, and both a RF model and a nomogram prediction model were developed. Internal model validation was performed using the Bootstrap method combined with 10-fold cross-validation, and the predictive performance and clinical utility of the models were evaluated using receiver operating characteristic curves, calibration curves, and decision curves. Results: A total of 1256 patients with hypertension were included in this study, of whom 120 experienced a cardiovascular event within 90 days of follow-up, resulting in an event rate of 9.55%. Multivariate analysis revealed that systolic blood pressure ≥140 mm Hg (odds ratio [OR] = 1.70, 95% confidence interval [CI]: 1.23-3.91, P = .031), poor medication adherence (OR = 1.51, 95% CI: 1.10-1.96, P = .022), concomitant diabetes (OR = 1.80, 95% CI: 1.23-2.46, P = .016), lack of regular exercise (OR = 1.47, 95% CI: 1.07-1.78, P = .001), and age ≥65 years (OR = 1.64, 95% CI: 1.15-3.44, P = .012) were independent risk factors for the occurrence of 90-day cardiovascular events in patients. In the internal validation cohort, the area under the curve was 0.812 for the nomogram model and 0.842 for the RF model. Although the RF model showed a numerically higher area under the curve, the difference between the 2 models was not formally tested and should not be interpreted as evidence of statistical superiority. Both models showed acceptable calibration and potential clinical net benefit based on their calibration curves and decision curve analyses. The nomogram prediction model for 90-day cardiovascular events in patients with hypertension, constructed based on variable selection using a RF algorithm, demonstrates reliable predictive accuracy and clinical utility. It can assist clinicians in the early identification of high-risk patients and provide a reference for developing individualized blood pressure management plans.