*3.1. Reliability of the Analytical Method*

The method used in the scope of the study was validated and approved by the Association of Official Agricultural Chemists (AOAC) as a standardized method [20]. Therefore, the method was verified at our laboratory conditions with a minimum requirement to obtain the provision of objective evidence that a given item fulfils specified requirements [23].

The standard analysis method applied in the scope of the study for the pesticide residue analyses was verified at 10 μg/kg to 60 μg/kg levels for high acid and watercontaining foods and foods with high sugar and low water activity in accordance with the European SANTE/11312/2021 Guidance Document [24]. The following parameters were evaluated in the scope of the verification study: linearity, limit of quantification (LOQ), trueness (in terms of recovery), precision (repeatability and interim precision-intralaboratory reproducibility), and measurement uncertainty.

The quantification and linearity of the method were evaluated during method verification study using matrix-matched calibration curves. Non-treated control orange samples were extracted as blank samples according to the extraction method described previously and the extracts were fortified with multi-standard working solutions at seven concentrations in a range between 2 μg/L and 80 μg/L. Moreover, during the analysis of the processed samples in the scope of the study (Table 1), multi-level calibration curves were constructed to cover the response of the pesticide residue in the sample at concentrations ranging between 10 μg/kg and 4000 μg/kg. The concentrations of pesticide active substances in samples were calculated in μg/kg using the weighted linear calibration curve functions with regression coefficients (R2) > 0.9999.

As a result, mean recovery (%), repeatability (% CVr), and intra-laboratory reproducibility (% CVi) of the method ranged between 98.54–107.60%, 0.82–14.06%, and 1.56– 16.23%, respectively (Figure 2). In addition, a blank sample was spiked with pesticides at the level of 10 μg/kg for each batch of analysis (*n* = 9), and the average recovery ratio was obtained in the range of 91% to 108%. LOQ was 10 μg/kg. The method performance parameters were compatible with the criteria set in the SANTE guideline [24].

*Foods* **2022**, *11*, 3918


**1.**Effectofhouseholdoftheconcentrationofpesticideresidueandfactors

164

 LOQ: limit of

determination,

 0.01 mg/kg, residue amount (B) is taken as 0.010 mg/kg when calculating the processing factor for these processes (Scholtz vd., 2017).

**Figure 2.** Method verification results (SANTE 2021).

Uncertainty is defined as a non-negative parameter characterizing the dispersion of the quantity values being attributed to a measurement based on the information used. There are two basic approaches to estimate measurement uncertainty, namely the bottom-up and top-down approaches. The top-down approach can be based on the method validation/verification and inter-quality-control laboratory dataset obtained within a laboratory. Combined measurement uncertainty arising from laboratory operations (*u*Laboratory) can be estimated from the square root of the quadratic sum of the random component (*u*Rw) and the systematic component (*u*Bias):

$$
\mu\_{\text{Laborator}} = \sqrt{\mu\_{\text{RW}}^2 + \mu\_{\text{Bias}}^2} \tag{3}
$$

*u*Rw is the coefficient of variation of within-laboratory reproducibility which reflects all variation under routine conditions. *u*Bias is the root mean square of the individual bias values and obtained from recovery studies [25,26]. As shown in Figure 3, the relative standard combined uncertainty arises from laboratory operations ranging between 0.1% and 19.21%, which corresponds to 0.2–38.42% of the expanded combined uncertainty at a 95% confidence level (k = 2). Those values comply with the maximum default relative expanded measurement uncertainty set as 50% in the European SANTE/11312/2021 Guidance Document [24].
