[Statistical aspects of the definition of reference values]

Area of Science:

• Environmental Science
• Biostatistics
• Toxicology

Background:

• Statistical analysis of reference values is heavily influenced by the criteria used for subject selection.
• Confusing studies establishing reference values with those investigating variable relationships can lead to poor statistical outcomes.

Purpose of the Study:

• To emphasize the critical importance of well-defined study scopes in reference value determination.
• To differentiate between studies establishing reference values and those exploring variable relationships.
• To guide precise subject selection and statistical methodology for accurate reference value studies.

Main Methods:

• Defining clear selection criteria to exclude specific subject groups (e.g., patients, individuals in contaminated zones).
• Utilizing reproducible criteria for selecting reference subjects from diverse population subgroups.
• Employing partition criteria to understand characteristic distributions or describe reference values in specific subgroups.

Main Results:

• Strict selection criteria are essential for accurate reference value studies.
• Partition criteria are useful for subgroup analysis but not for relationship studies.
• Consulting a statistician is crucial for subject and variable selection in population reference value studies.

Conclusions:

• Well-defined study scopes and precise subject selection are paramount for reliable reference value determination.
• Appropriate statistical methods and criteria are necessary to avoid misinterpretation of data.
• Careful consideration of selection and partition criteria, guided by statistical expertise, ensures the integrity of reference value studies.