The next generation of regression standardization with the R package stdReg2

Area of Science:

• Epidemiology
• Biostatistics
• Statistical Software Development

Background:

• Regression standardization is a key technique for causal inference in epidemiological studies.
• Existing statistical software may lack user-friendliness or advanced functionalities.
• There is a need for updated tools to facilitate robust causal analysis.

Purpose of the Study:

• To introduce the upgraded R package stdReg2, enhancing causal inference capabilities.
• To detail new features including a double-robust method and survival analysis standardization.
• To provide guidance for researchers on utilizing the improved statistical software.

Main Methods:

• Updating the R package stdReg to stdReg2.
• Implementing a generalized linear model-based double-robust method.
• Adding regression standardization for restricted mean survival time.

Main Results:

• The stdReg2 package offers improved user-friendliness and flexibility over stdReg.
• New functionalities enable advanced causal inference techniques.
• The package supports both standard regression standardization and restricted mean survival analysis.

Conclusions:

• stdReg2 is a valuable update for epidemiological research requiring causal inference.
• The enhanced statistical software facilitates more robust and flexible analysis.
• New and existing users are encouraged to adopt stdReg2 for their research needs.