coefficient of variation (CV)

The coefficient of variation expresses standard deviation relative to the mean, usually as CV = 100 × standard deviation / mean. It describes relative imprecision for quantitative measurements and allows comparison across some concentration levels or methods. A smaller CV indicates less dispersion relative to the mean, but does not by itself establish accuracy or suitability for clinical use.

State the material, concentration, number of observations, and time period behind a CV. Within-run repeatability and longer-term within-laboratory imprecision answer different questions and should not be confused. CV becomes unstable or unhelpful when the mean is near zero, and percentage comparisons can mislead for quantities without a meaningful ratio scale. Excluding data requires a documented reason rather than a desire for a better-looking figure.

For accreditation, imprecision evidence supports verification, uncertainty estimation, and ongoing control planning. Review performance at clinically relevant levels rather than relying on a single pooled value. A calculated CV needs appropriate acceptance criteria and representative data; a favorable percentage cannot compensate for unacceptable bias, a narrow study design, or uncontrolled pre-examination problems.