Scientific rigour in psycho-oncology trials: why and how to avoid common statistical errors

Area of Science:

• Psycho-oncology research
• Health research methodology
• Statistical analysis in clinical studies

Background:

• Statistical and methodological flaws are prevalent in health research, particularly psycho-oncology.
• These flaws can lead to misleading findings, resource wastage, patient harm, and diminished public confidence.

Purpose of the Study:

• To identify and discuss common statistical errors and methodological pitfalls in psycho-oncology.
• To provide guidance on avoiding these issues across study design, analysis, reporting, and conclusion stages.

Main Methods:

• Review and discussion of common statistical and methodological issues in psycho-oncology.
• Explanation of problems and solutions for 14 specific topics.

Main Results:

• Identified 14 key areas of concern including power, clustering, missing data, variable categorization, subgroup analyses, multiple comparisons, interactions, confidence intervals, and p-value interpretation.
• Provided actionable advice for avoiding common pitfalls in statistical analysis and research design.

Conclusions:

• Recommends early involvement of biostatisticians in psycho-oncology studies.
• Emphasizes the need for ongoing education in statistical best practices and reporting standards.