Statistics from Altmetric.com
Clinicians use several approaches in the interpretation of laboratory results. These include comparison with predetermined cut-off values or reference values, or a comparison between two sequential results for a specific analyte.1 Each has its own merits. The latter is the focus of this commentary.
A simple comparison between two sequential results is not as straightforward as it seems. It should be remembered that each result is associated with its own inherent random variation, meaning that each result obtained is, in fact, a dispersion rather than a singular value. This random variation comprises both variation associated with laboratory activity (pre-analytical and analytical variation) and inherent biological variation (intra-individual) (see box 1).
Box 1: Total variation
CVT = (CVP2+ CVA2+ CVI2)1/2
CVP is considered negligible due to standardised collection protocols, therefore:
CVT = (CVA2+ CVI2)1/2
The total variation consists of the sum of all associated variations, where CVP = pre-analytical variation, CVI = intra-individual biological variation, CVA = analytical variation, and CVT = total variation.
Therefore, a numerical change is an expected occurrence in sequential testing.1 Whether this numerical change is significant is a question answered by determining the reference change value (RCV) as proposed by Harris and Brown.2
The RCV, defined as the critical difference that must be exceeded between two sequential results for a significant (or true) change to occur,3 incorporates the total variation associated with both results and is demonstrated by the equation in box 2.
Box 2: Reference change value
RCV = Z × [(CVA2+ CVI2) + (CVA2+ CVI2)]½RCV = 2½× Z × (CVA2+ CVI2)½