Introduction
-
Reference individual—a subject who meets the inclusion criteria.
-
Reference population—the group comprising all reference individuals who exist, usually an unknown quantity.
-
Reference sample group—the group of reference individuals selected, usually non-randomly, to represent the reference population.
-
Reference value—value of a test parameter measured from a reference individual (Fig. 1).
-
Reference distribution—the frequency of all reference values (Fig. 1). Often this distribution cannot be described by a single mathematical function. It is relatively rare to find a Gaussian distribution, so defining reference limits as the mean ± some standard deviations of the reference values is rarely appropriate and risks systematic misclassifications (see Parametric method).
-
Reference limit—a value derived mathematically from the reference distribution, defined such that a stated fraction (e.g. 2.5%) of the reference values lies above or below it (Fig. 1).
-
Reference interval—the interval considered as healthy, which for two-tailed limits is the interval between and including the two reference limits (Fig. 1) or for one-tailed reference limits, the values equal to or above/below the one reference limit.
Establishing reference intervals: direct sampling
Defining reference individuals
Exclusion criteria for consideration when selecting reference individuals | Potential partitioning factors for consideration |
---|---|
Restricted diet | Age |
Alcohol use | Sex |
Drug use | Affluence/deprivation |
Drug misuse | Ethnicity/pigmentation |
Recent or current illness | Refractive status |
History of premature birth | |
History or family history of ophthalmic or neurological disease | |
History of retinal surgeries, recent other ocular surgeries (e.g. cataract) | |
Indicators of ocular disease such as high intraocular pressure, diabetes, poor cup-to-disc ratios, poor best-corrected visual acuity |