Grant: Project Grant

Ensuring the safety and effectiveness of drugs used to treat or prevent different diseases is essential for improving population health. Yet, some medications have serious unintended side effects or may be less effective than alternative drugs. Thus, an accurate assessment and comparison of the safety and effectiveness of medications currently used by Canadians is among the main priorities for health research, public health agencies and the general population. However, rigorous scientific assessment of real-world drug safety and effectiveness requires large population-based studies, which pose many analytical challenges. For example, patients prescribed a new drug may be more likely to have a more advanced disease and additional risk factors than receivers of a less expensive older drug. Thus, the observed differences in health outcomes between the users of alternative drugs may be due to differences in their characteristics, rather than to drugs' properties, and complex analyses are necessary to reduce bias and avoid incorrect conclusions. Furthermore, studies of drugs used in clinical practice have to account for treatment duration and drug doses, and for the possible treatment non-adherence, where some patients may take lower doses than prescribed or even stop using the drug. In addition, patients' responses to treatment may depend on their sex, age, and/or general health. To address these methodological challenges, we propose to develop and validate several new statistical methods for large population-based studies of medications. The use of the proposed new methods in large Canadian and international pharmacoepidemiology studies will permit more accurate assessments of the real-world safety and effectiveness of drugs used by Canadians to treat arthritis, infections, and cardiovascular diseases, or taken during pregnancy. Our new methods may also help improve drug prescribing, increasing health benefits while reducing the risks of serious adverse events.