Abstract / Summary
Objective: The objective of this study was to identify the demographic and clinical predictors of statin intolerance among patients prescribed statins for the primary prevention of cardiovascular disease during the first two years following the first recorded prescription, using real-world data. Methods: This study used the Electronic Practice Based Research Network Linked Dataset. An operational definition based on muscle symptoms and creatine kinase levels was applied to identify patients with statin intolerance. Descriptive and multivariate logistic regression analyses were performed using R software. All variables identified a priori as potential predictors were included in the initial model, and backward stepwise selection using the Akaike Information Criterion was performed across the imputed datasets to identify a parsimonious final model. Results: Overall, 4,016 patients accounting for 60,873 visits were included in the analysis. About 3.5% of patients were statin intolerant. After adjusting for all other variables, higher odds of statin intolerance were observed among female or other gender patients (AOR 1.5, 95% CI 1.0-2.2), patients residing in Socio-Economic Indexes for Australia (SEIFA) area based socioeconomic advantage and disadvantage (IRSAD) category 8 or 9 (AOR 3.4, 95% CI 2.0-5.8), and patients with three or more selected comorbidities (AOR 6.4, 95% CI 2.2-18.3). A similar direction of associations was seen in the final model. Conclusion: Gender, IRSAD, and the number of selected comorbidities were associated with statin intolerance. However, these findings should be interpreted in the context of the operational EHR based definition of statin intolerance. Further validation in other populations and settings is required before these predictors can be considered for clinical application.