AOAC Gluten Quantitative Validation Guidance-Round 1(Nov 2023)

233

LIMIT OF DETECTION AND LIMIT OF QUANTITATION ESTIMATION 234 Limits of detection and quantification should be estimated using methods that account for the 235 relationship between concentration and variance commonly observed with immunoassays (i.e., where 236 variance increases with concentration), as described in IUPAC recommendations (Currie 1999) 237 Limit of Detection (LOD) 238 Data Required: mean observed concentration and intermediate precision standard deviation from 239 analysis of at least four analyte levels of each claimed matrix, including blank/zero 240 Calculations: 241 • Calculate mean concentration and intermediate precision standard deviation (S i ) for 242 each test material. 243 • Plot S i versus observed mean concentration. 244 • Perform linear regression (ordinary least square estimate or weighted least square 245 analysis) 246 • Calculate the LOD according to the following formula (where ̅ (0) = calculated mean 247 result from blank samples, S i(0) = intermediate precision standard deviation of blank 248 samples, and slope is the slope from the linear regression above): 249 = ( ̅ ( 0 ) +3.3× ( 0 ) ) (1 − 1.65× ) 250 251 Values used as estimates of ̅ (0) and S i(0) cannot be negative. If the S i(0) from the linear regression (i.e., the 252 intercept value) is negative, use the observed S i(0) from blank matrix samples. If the observed S i(0) also 253 seems unacceptable, use S i from the lowest concentration test material. 254 Multiple Matrices: 255 • Plot S i versus observed mean concentration for all matrices. It is generally expected that the 256 relationship between S i and concentration will be sufficiently similar across matrices to conduct 257 a single regression analysis for the combined data from all matrices. 258 Values used as estimates of ̅ (0) and S i(0) cannot be negative. If the S i(0) from the linear regression (i.e., the 259 intercept value) is negative, use the observed S i(0) from blank matrix samples. If the observed S i(0) also 260 seems unacceptable, use S i from the lowest concentration test material. 261 For estimation of ̅ (0) , calculate the mean observed concentration across all blank matrices. 262 If the relationship between standard deviation and concentration appears to be substantially different 263 between matrices, consult with the AOAC Statistics Committee and Expert Review Panel.

264 265 266 267 268 269 270 271

Example Data and Calculations:

Table 1:EXAMPLE DATA TABLE

0 ppm

0.5 ppm 1.0 ppm 2.5 ppm