Should we have confidence in radiology papers?

Area of Science:

• Medical Statistics
• Radiology Research

Background:

• Confidence Intervals (CIs) are crucial for interpreting statistical significance.
• Appropriate CI usage enhances the reliability of research findings in clinical publications.

Purpose of the Study:

• To assess the appropriate application of Confidence Intervals in Clinical Radiology.
• To compare CI utilization across Clinical Radiology, Radiology, and the British Medical Journal (BMJ).

Main Methods:

• Analysis of 50 consecutive papers from Clinical Radiology and comparable samples from Radiology and BMJ.
• Examination of the utilization or omission of Confidence Intervals and statistical hypothesis testing.
• Statistical analysis using chi-squared tests and 95% Confidence Intervals for proportion differences.

Main Results:

• The study identified variations in the appropriate use of Confidence Intervals across the analyzed journals.
• Confidence Intervals provide valuable quantitative information, particularly for small sample sizes or subgroups.
• Statistical hypothesis testing was also analyzed in conjunction with CI usage.

Conclusions:

• Appropriate use of Confidence Intervals in medical research requires further attention.
• Confidence Intervals significantly enhance the interpretation of uncertainty in study results.
• The findings underscore the importance of adhering to statistical best practices in radiological publications.