Matching Cases and Controls Using SAS® Software.

Area of Science:

• Medical research
• Observational studies
• Biostatistics

Background:

• Matching is crucial in observational studies, particularly in medical research, to control for confounding variables.
• Existing SAS matching programs are scarce and often difficult for novice users.
• There is a need for accessible and user-friendly matching tools in SAS.

Purpose of the Study:

• To present a novel, user-friendly matching program for SAS software.
• To provide researchers with a tool for efficient and unbiased control selection in observational studies.
• To facilitate the customization and testing of matching procedures.

Main Methods:

• Development of a SAS matching program with capabilities for multi-variable matching.
• Inclusion of detailed explanations for program customization and usage.
• Implementation of randomized and automated control selection to minimize bias.
• Integration of features for testing incomplete matching.

Main Results:

• The presented SAS program facilitates matching on multiple variables.
• Automated and randomized control selection minimizes potential selection bias.
• The program includes comprehensive documentation and customizable options.
• Tools are provided for assessing the completeness of the matching process.

Conclusions:

• The developed SAS matching program enhances the feasibility and rigor of observational studies.
• It offers a valuable resource for medical researchers, especially those new to SAS.
• The program's features aid in reducing bias and ensuring robust matching outcomes.