Research and scholarly methods: Propensity scores

Area of Science:

• Pharmacoepidemiology
• Biostatistics
• Health Services Research

Background:

• Confounding bias is a significant challenge in pharmacoepidemiologic research.
• Propensity score methods offer a statistical approach to mitigate this bias.
• Accurate control of confounding is crucial for valid study findings.

Purpose of the Study:

• To provide a comprehensive overview of propensity score methods for controlling confounding bias.
• To introduce pharmacists and researchers to the application and evaluation of propensity scores.
• To facilitate informed discussions on propensity score methodology in research.

Main Methods:

• This is a methods review, summarizing existing literature and best practices.
• Key concepts covered include propensity score definition, calculation, and application.
• Methods for assessing and achieving covariate balance are discussed.

Main Results:

• Propensity scores are dimension-reducing balancing scores.
• They enable the creation of comparable treatment and reference groups based on measured covariates.
• Effective evaluation of covariate balance is essential for valid propensity score application.

Conclusions:

• Propensity score methods are valuable tools for reducing confounding bias in pharmacoepidemiologic studies.
• Understanding assumptions and evaluation techniques is key to appropriate application.
• This review equips researchers with foundational knowledge for using and discussing propensity scores.