Abstract
OBJECTIVE: This study aims to develop and validate a nomogram-based predictive model for estimating the risk of adverse drug reactions (ADR) to statins in patients with coronary heart disease (CHD). METHODS: A retrospective cohort study was conducted using clinical data from 351 patients with CHD who received statin therapy in the cardiology department of a tertiary hospital in Anhui Province, China, between February 2021 and January 2022. The dataset was randomly divided into a development cohort (n = 283) and a validation cohort (n = 68) in an 8:2 ratio. Logistic regression analysis was applied in the development cohort to identify independent risk factors for statin-induced ADR. A nomogram was subsequently constructed in R based on the selected predictors, and its clinical utility, discriminative performance, and calibration were evaluated. RESULTS: The overall incidence of statin-associated ADR among the 351 subjects was 24.22%, classified into three categories according to the affected system: musculoskeletal toxicity, hepatic/renal dysfunction, and gastrointestinal reactions. Univariate and multivariate logistic regression analyses in the development cohort identified the following as significant independent risk factors (P < 0.05): age ≥60 years, body mass index ≥23 kg/m(2), disease duration ≥5 years, presence of ≥3 comorbid conditions, dyslipidemia, history of cerebral infarction, high-dose statin use, and concomitant use of multiple medications. A nomogram model was constructed based on these predictors. The model demonstrated strong discriminative performance, with an area under the receiver operating characteristic (ROC) curve of 0.808 (95% CI [0.751-0.865]) in the development cohort and 0.852 (95% CI [0.752-0.951]) in the validation cohort. CONCLUSION: A nomogram-based risk prediction model was successfully developed to estimate the probability of statin-induced ADR in patients with CHD, based on a set of statistically significant clinical risk factors. The model exhibited favorable predictive accuracy and discrimination. It offers a practical tool for clinicians to identify high-risk individuals and implement early preventive or interventional strategies accordingly.