A comparison of parametric propensity score-based methods for causal inference with multiple treatments and a binary...

Area of Science:

• Health Services Research
• Biostatistics
• Epidemiology

Background:

• Observational studies are crucial for comparative effectiveness research (CER) but are susceptible to confounding bias.
• Propensity score methods are widely used to adjust for confounders in treatment effect estimation.
• Existing literature often focuses on two treatments or continuous outcomes, leaving a gap for multiple treatments with binary outcomes.

Purpose of the Study:

• To describe and compare propensity-based methods for CER with multiple treatments and binary outcomes.
• To evaluate the performance of these methods through simulation studies.
• To apply these methods to a real-world dataset for prostate cancer therapies.

Main Methods:

• Description of propensity score-based methods for comparing more than two treatments with a binary outcome.
• Simulation studies to assess the relative performance of different methods.
• Application of methods to medical and pharmacy claims data for prostate cancer patients.

Main Results:

• The study provides a comparative assessment of various propensity score methods in a complex scenario.
• Simulation results offer insights into the performance characteristics of each method.
• The application demonstrates the utility of these methods in analyzing real-world health insurance data.

Conclusions:

• Propensity score methods can be adapted and compared for comparative effectiveness research involving multiple treatments and binary outcomes.
• The findings contribute to the evolving literature on causal inference in complex observational studies.
• This research aids in understanding treatment effectiveness using large-scale healthcare claims data.