Search

Background Amiodarone is one of the most commonly used anti-arrhythmic agents. Amiodarone pulmonary toxicity is a potentially fatal adverse effect associated with amiodarone use. Previous studies on the epidemiology and risk factors for amiodarone pulmonary toxicity showed diverse results. Methods A multicenter retrospective cohort study was conducted to identify clinic-epidemiologic markers associated with amiodarone pulmonary toxicity for development of a prediction rule. Patients taking amiodarone who were managed in 3 centres in Hong Kong from 2005 to 2015 were included in this study. Penalized logistic regression was used to model the outcome as it is rare. Results A total of 34 cases with amiodarone pulmonary toxicity were identified among 1786 patients taking amiodarone for at least 90 days from 2005 to 2015. The incidence of amiodarone pulmonary toxicity was estimated to be 1.9%. The risk factors for amiodarone pulmonary toxicity included advanced age (OR 1.047, 95% CI 1.010–1.085, p = 0.013), ventricular arrhythmia (OR 2.703, 95% CI 1.053–6.935, p = 0.039), underlying lung disease (OR 2.511, 95% CI 1.146–5.501, p = 0.021) and cumulative dose of amiodarone (OR 4.762, 95% CI 1.310–17.309 p = 0.018). Conclusions The incidence of amiodarone pulmonary toxicity in Chinese patients in Hong Kong is estimated to be 1.9% in this study. Age, underlying lung disease, ventricular arrhythmia and cumulative dose of amiodarone are associated with the development of amiodarone pulmonary toxicity. A prediction rule was developed to inform the risk of developing amiodarone pulmonary toxicity.

Biomedical data frequently contain imbalance characteristics which make achieving good predictive performance with data-driven machine learning approaches a challenging task. In this study, we investigated the impact of re-sampling techniques for imbalanced datasets in PET radiomics-based prognostication model in head and neck (HNC) cancer patients.Radiomics analysis was performed in two cohorts of patients, including 166 patients newly diagnosed with nasopharyngeal carcinoma (NPC) in our centre and 182 HNC patients from open database. Conventional PET parameters and robust radiomics features were extracted for correlation analysis of the overall survival (OS) and disease progression-free survival (DFS). We investigated a cross-combination of 10 re-sampling methods (oversampling, undersampling, and hybrid sampling) with 4 machine learning classifiers for survival prediction. Diagnostic performance was assessed in hold-out test sets. Statistical differences were analysed using Monte Carlo cross-validations by post hoc Nemenyi analysis.Oversampling techniques like ADASYN and SMOTE could improve prediction performance in terms of G-mean and F-measures in minority class, without significant loss of F-measures in majority class. We identified optimal PET radiomics-based prediction model of OS (AUC of 0.82, G-mean of 0.77) for our NPC cohort. Similar findings that oversampling techniques improved the prediction performance were seen when this was tested on an external dataset indicating generalisability.Our study showed a significant positive impact on the prediction performance in imbalanced datasets by applying re-sampling techniques. We have created an open-source solution for automated calculations and comparisons of multiple re-sampling techniques and machine learning classifiers for easy replication in future studies.