Strategies in adjusting for multiple comparisons: A primer for pediatric surgeons

Area of Science:

• Surgical Research Statistics
• Biostatistics in Medicine

Background:

• Multiple statistical comparisons in pediatric surgery research increase the risk of false positive results (Type I errors).
• High-impact journals increasingly require careful consideration of multiplicity in study designs.

Purpose of the Study:

• To provide surgeons with a guide on statistical approaches for managing multiple comparisons.
• To help surgeons maintain false positive results at an acceptable level in their research.

Main Methods:

• Review of statistical approaches for controlling Type I error rates in multiple comparisons.
• Discussion of methods including Bonferroni correction, False Discovery Rate (FDR), Tukey's, Scheffé's, Holm's, and Dunnett's procedures.

Main Results:

• Illustrates the application of various multiple comparison procedures using a hypothetical ANOVA example in surgical research.
• Emphasizes that the choice of method is situation-dependent and requires collaboration with biostatisticians.

Conclusions:

• Testing multiple hypotheses increases the risk of rejecting true null hypotheses.
• Awareness and application of multiplicity adjustment methods improve study design and data analysis in surgical research.
• Failure to adjust for multiple comparisons can lead to exaggerated findings and misleading evidence.