Level of Detection (LOD50) of Campylobacter Is Strongly Dependent on Strain, Enrichment Broth, and Food Matrix

The detection of thermotolerant Campylobacter in food may be difficult due to the growth of extended-spectrum β-lactamase (ESBL)-producing Enterobacteriaceae during enrichment, resulting in false-negative samples. Therefore, the ISO protocol (ISO 10272-1:2017) suggests that, next to Bolton broth (BB), Preston broth (PB) is used as an enrichment broth to inhibit competitive flora in samples with suspected high levels of background microorganisms, such as ESBL-producing bacteria. However, the application of the strains used for validation of this ISO was not clearly characterized. This study examined the LOD50 (level of detection, the concentration where the probability of detection is 50%) of the validation strains (three C. jejuni and two C. coli strains) in BB and PB using different food matrices, namely, raw milk, chicken skin, frozen minced meat, and frozen spinach. The LOD50 was calculated by inoculating multiple portions with at least two inoculum levels. For each reproduction, eight test portions were used for each inoculum level and the test portion size was 10 g (chicken skin, frozen minced meat, and frozen spinach) or 10 mL (raw milk). Furthermore, the effect of artificially inoculated ESBL-producing E. coli on the LOD50 was examined to mimic the presence of ESBL-producing background microorganisms in the food matrices, namely, raw milk and chicken skin. In BB, the LOD50 of all strains tested in raw milk, chicken skin, and frozen spinach was rather low (0.4–37 CFU/test portion), while the LOD50 in frozen minced meat was higher and much more variable (1–1,500 CFU/test portion), depending on the strain. Generally, enrichment in PB resulted in higher LOD50 than in BB, especially for C. coli. Co-inoculation with ESBL-producing E. coli increased the LOD50 in BB, while PB successfully inhibited the growth of this competitive microorganism. In conclusion, food matrix and enrichment broth may have a large influence on the LOD50 of different Campylobacter strains. Therefore, it is not possible to give an unequivocal advice on when to use which enrichment broth, and this advocates the use of both methods in case of doubt. Furthermore, this study indicates specific strains that would be a good choice to use for Campylobacter method verification as described in ISO 16140-3:2021.