How well do commonly used data presentation formats support comparative effectiveness evaluations?

Area of Science:

• Health decision science
• Risk communication
• Data visualization

Background:

• Effective decision-making relies on accurately understanding comparative effectiveness.
• Optimal methods for presenting comparative data are not well-established.

Purpose of the Study:

• To evaluate how well five common data presentation formats convey comparative effectiveness information.
• To identify which formats best support accurate understanding of health outcome probabilities.

Main Methods:

• An internet survey with 279 participants assessed understanding of hypothetical screening test outcomes.
• Five data formats were compared: table, magnified bar chart, risk scale, frequency diagram, and icon array.
• Outcomes measured included judgment accuracy, perceived vs. actual likelihood ratios, response consistency, and clarity.

Main Results:

• No significant differences in judgment accuracy were found across all formats initially.
• However, flowcharts, icon arrays, and tables led to more accurate and consistent judgments than risk scales and bar charts.
• Tables and flowcharts were perceived as clearest, while low subjective numeracy impaired interpretation of icon displays, bar charts, and flow diagrams.

Conclusions:

• No single data presentation format reliably ensures accurate understanding of comparative effectiveness.
• Patients with low subjective numeracy face particular challenges in interpreting comparative data, regardless of format.
• Further research is needed to develop universally effective data visualization strategies for health decisions.