199 / 220
Single Lab Validation of a Method for the Determination of β‐
Galactooligosaccharides in Infant Formula & Adult Nutritionals / Sean Austin (NRC,
Lausanne)
25 Jul 2016
CONFIDENTIAL
∙ This document may not be reproduced or disclosed to third parties without prior authorization
Page 16 / 35
5.3. Method Accuracy
Four different blank matrices were spiked at four levels with two different GOS ingredients (as
described in Table 4). Recoveries were calculated by comparing the spiked amount vs the measured
amount (Table 6). Performance is poorest at the lowest spike level (0.2 g/100g) with spike
recoveries ranging from 85‐114%, at the other three levels the recoveries are in the range 86 – 94
%, but in most cases the recoveries are outside the 90 – 110% range specified in the SMPR, and in
most cases the recoveries are slightly low.
Table 6: GOS Recoveries
Nestlé
Code
Recovery
Level 1
Level 2
Level 3
Level 4
Spike
(g/100g)
Rec
(%)
RSD
(%)
Spike
(g/100g)
Rec
(%)
RSD
(%)
Spike
(g/100g)
Rec
(%)
RSD
(%)
Spike
(g/100g)
Rec
(%)
RSD (%)
KBI‐00045
0.211
84.8
4.9
0.980
86.9
2.6
2.47
89.3
1.6
4.91
85.6
2.9
KBI‐00046
0.211
91.5
13
0.990
90.7
5.1
2.45
89.2
3.3
4.97
86.6
1.6
KBI‐00047
0.203
97.6
5.0
0.980
93.6
3.9
2.45
89.4
3.4
4.96
93.9
3.6
LB3‐00037
0.202
114
2.6
1.11
88.7
4.2
2.36
93.1
3.3
4.46
89.2
3.9
5.4.
Limit of Quantification
Based on the spike/recovery data at 0.2 g/100g the method is capable of determining GOS at 0.2
g/100g as specified in the SMPR. However, two of the samples in the SPIFAN matrix kit appear to
contain lower GOS contents (around 0.15 g/100g), and at this level the method still appears to be
performing well. We would thus estimate the LoQ of the method to be below 0.15 g/100g.
VALIDATION REPORT
FOR ERP USE ONLY
DO NOT DISTRIBUTE