Primer for Clinical Researchers on Ordinal Outcomes for Emergency Medicine

Area of Science:

• Clinical Research
• Biostatistics
• Emergency Medicine

Background:

• Ordinal outcomes are increasingly utilized in clinical research.
• They offer advantages over binary outcomes, including increased statistical power and applicability across various illness severities.
• However, understanding and application of ordinal outcomes can be challenging for clinicians and researchers.

Purpose of the Study:

• To provide a practical overview of ordinal outcomes in emergency medicine.
• To discuss the advantages and disadvantages of using ordinal outcomes.
• To describe effect measures and statistical methods for estimating treatment effects with ordinal outcomes.

Main Methods:

• Review of ordinal outcome types, including hierarchical composite endpoints (e.g., Desirability of Outcome Ranking [DOOR]).
• Description of target parameters such as common odds ratio and win ratio.
• Explanation of statistical methods for parameter estimation, including assumptions and limitations.

Main Results:

• Ordinal outcomes provide a flexible, efficient, and nuanced approach to measuring treatment effects.
• Commonly used parameters include the common odds ratio and win ratio.
• Statistical methods for estimation have specific assumptions and limitations that require careful consideration.

Conclusions:

• Ordinal outcomes enhance the measurement of treatment effects in clinical research.
• Effective utilization requires careful planning for estimation and communication of results.
• This review aims to increase accessibility and understanding of ordinal outcomes for clinical researchers.