Development and Validation of a Nomogram for Predicting Intravenous Amiodarone-Induced Phlebitis: A Prospective Cohort Study.

Background: Intravenous amiodarone is primarily used to treat hemodynamically unstable atrial or ventricular arrhythmias and prevent atrial fibrillation after cardiac surgery. However, when administered via peripheral intravenous infusion, the incidence of phlebitis is relatively high. This increase in complication rate not only increases patient discomfort and prolongs hospitalization but may also adversely affect treatment efficacy and quality of life. Therefore, early identification of high-risk patients and timely intervention are crucial for reducing the occurrence of phlebitis, enhancing treatment safety, and improving patient prognosis. Aim: This study aimed to develop and validate a nomogram to predict the risk of phlebitis associated with peripheral intravenous injection of amiodarone. Design: This study is a prospective cohort study. Methods: The research was carried out at Guigang People's Hospital in China. A total of 393 cases who underwent intravenous infusion of amiodarone through the peripheral vein between January 2022 and December 2023 constituted the study cohort. The cohort was randomly divided using computer-generated randomization numbers, with stratification maintained for key clinical characteristics (including age, infusion duration, and concentration). A 70:30 ratio allocation was applied to create a training set (n = 275) for model development and a validation set (n = 118) for performance evaluation. Logistic regression was used to construct the model. A scatterplot was used to test the predictive performance of the model. Results: Age, infusion time, infusion concentration, and whether to perform 2 h of alternating indwelling needle were factors influencing the occurrence of phlebitis associated with peripheral intravenous use of amiodarone. A nomogram model of the risk of developing phlebitis with peripheral intravenous injection of amiodarone was successfully constructed. Bootstrap results showed that the calibration curves fitted well with the ideal curves. The areas under the receiver operating characteristic (ROC) curve for the training and validation sets were 0.812 (95% CI: 0.757-0.867) and 0.798 (95% CI: 0.712-0.883), respectively. The results of Hosmer-Lemeshow goodness-of-fit test revealed χ 2 = 3.414 and p = 0.906 for the training set and χ 2 = 11.835 and p = 0.159 for the validation set. Conclusions: The model showed a good predictive effect, and it can be effectively used as a tool for clinically assessing the risk of phlebitis following peripheral intravenous injection of amiodarone.