Graphic Portrayal of Studies With Paired Data: A Tutorial

Area of Science:

• Biostatistics
• Clinical Research Methodology
• Data Visualization

Background:

• Clinical research frequently utilizes paired outcome data, involving two measurements of the same variable within an individual patient over time.
• Effective graphical reporting of such data is crucial for understanding patient journeys and treatment effects.

Purpose of the Study:

• To explore graphical methods for reporting paired outcome data in clinical research.
• To present techniques that simultaneously illustrate individual patient experiences and compare differences between treatment groups.

Main Methods:

• The study emphasizes the utility of hybrid parallel line plots for visualizing paired data.
• Several variations of these plots are examined to demonstrate their flexibility.
• Stata code is provided to facilitate the creation of these graphical representations.

Main Results:

• Hybrid parallel line plots offer a robust approach to visualizing paired outcome data.
• These plots enable clear depiction of individual patient trajectories.
• The method facilitates direct comparison of treatment group effects.

Conclusions:

• Hybrid parallel line plots are recommended for graphically reporting paired outcome data in clinical studies.
• These visualizations enhance the understanding of both individual patient responses and group-level treatment effects.
• The provided Stata code serves as a practical tool for researchers implementing these methods.