RESUMO
An approach was proposed for the estimation of measurement uncertainty for analytical methods based on one-point calibration. The proposed approach is similar to the popular multiple-point calibration approach. However, the standard deviation of calibration was estimated externally. The approach was applied to the estimation of measurement uncertainty for the quantitative determination of ketamine (K) and norketamine (NK) at a 100 ng/mL threshold concentration in urine. In addition to uncertainty due to calibration, sample analysis was the other major source of uncertainty. To include the variation due to matrix effect and temporal effect in sample analysis, different blank urines were spiked with K and NK and analyzed at equal time intervals within and between batches. The expanded uncertainties (k = 2) were estimated to be 10 and 8 ng/mL for K and NK, respectively.