Calculation of the limit of detection, the limit of identification and the limit of quantitation of tetrahydrocannabinol (THC) in serum

For the determination of the THC in serum 7 calibrations measured under practical conditions were evaluated within the whole spectrum of up to 10 ng/ml. The limits of detection calculated according to DIN 32645 included a probability of a false positive result of 1% ((alpha)-error). The LOD was found to be within a span of 0.7 and 1.7 ng/ml for the most and least intensive ion trace respectively. With regard to the practical experience of THC determination these values appear to be relatively high. There is no advantage in using the total ion current as the least intensive ion trace is the key factor and needs to be included in calculation. To enable quantitative determinations at 2 ng/ml the (alpha)-error can be increased to 10 or 20 % as signals from background noise represent independent events in each ion trace. The LOD for the least intensive ion is then 0.6 or 0.4 ng/ml. Limits of quantitation were 1.6 or 2.3 ng/ml respectively. This allows quantitative determination up to the highest calibration level. In order to enable a comparison of characteristics between labs we propose standardised calibration levels, registration of four ion traces and a calculation of the LOD according to DIN 32645 using the least intensive ion trace and an (alpha)- error of 10 to 20 % maximum. Due to measuring imprecision around the limit of detection, the peak surface area is unlikely to coincide with the field of operations. For the calculation of the limit of quantitation the most intensive ion and the relative imprecision of 33.3 % (k = 3) can be used. Under no circumstances is the limit of quantitation to fall below the limit of detection (LOD).