Towards personalized primary prevention of cardiovascular diseases
Cardiovascular disease (CVD) is the leading cause of death globally. This thesis aimed to generate evidence to support more personalized primary prevention of CVD using antihypertensive and antihyperlipidemic monotherapy with medicines.
In Chapter 2 we validated the use of dispensing data for vitamin K antagonists, platelet aggregation inhibitors, and/or nitrates as proxies for both new and prior hospitalizations due to major adverse cardio-cerebrovascular events (MACCE). In Chapter 3 and Chapter 4 we applied intention-to-treat (ITT) and per-protocol (PP) analyses, respectively, showing that thiazide diuretics were associated with a lower risk of acute cardiac drug therapy (CDT) compared to beta-blockers. In Chapter 5 we found that atorvastatin and pravastatin users had a higher risk of CDT compared to simvastatin users, using both ITT and PP approaches. In Chapter 6 we demonstrated that comparative effectiveness results using traditional statistical models align closely with those from machine learning-based models. Chapter 7 showed that reaching the deductible limit slightly improved adherence to both antihypertensive and antihyperlipidemic medicines, while higher deductible amounts were associated with decreased adherence. In Chapter 8 and Chapter 9 we identified multiple risk factors affecting drug utilization patterns, and developed predictive models among new users of cardiovascular monotherapy.This thesis underscores the importance of personalized strategies in the primary prevention of CVD. Future research should focus on addressing current research limitations and developing new methods in the evolving field of pharmacoepidemiology.