Based on the need for a scientific basis for existing requirements in EU legislation on freezing of meat or for its possible amendment, the opinion compares microbial growth of relevant pathogenic, spoilage and indicator microorganisms within five scenarios of chilling, storage and defrosting of bovine, ovine and porcine meat, using predictive microbiology models that considered various conditions of temperature and, where possible, pH and aw. Results obtained were compared to a reference scenario: storing meat at 7°C, aerobically, until 15 days post-slaughter. Storage of meat for 6 weeks, vacuum-packed immediately after stabilisation or 15 days post-slaughter, resulted in more growth of at least some of the bacteria assessed compared to the reference scenario, both at 3°C (certainty level 66%-90%) and at 7°C (certainty level 95%-99%). Predictions allowed estimating time at which equivalent microbial growth (i.e. ≤ 0.5 log10 difference) to the reference scenario is reached ('equivalence time'), assuming different initial contamination levels of relevant spoilage bacteria. When storing meat at 7°C, vacuum-packed immediately after stabilisation, equivalence time was determined by Salmonella and reached in 5-6 days of post-slaughter storage (certainty level 66%-90%). When storing meat at 3°C, equivalence time was determined by spoilage lactic acid bacteria and reached in 29-30 days post-slaughter (certainty level 66%-90%). However, when initial contamination with spoilage bacteria was high (e.g. 5 log10 CFU/cm2), predicted spoilage levels of 7 log10 CFU/cm2 were reached after 15-16 days. When considering also expected growth during post-defrosting storage at 4°C for 7 days, equivalence times were of 5-6 days (unchanged) and 13-16 days, respectively, though meat would have to be frozen immediately after stabilisation when initial contamination with spoilage bacteria is high. Predicted levels of indicator microorganisms for verification are provided for different assumed initial contamination levels, representing examples to be further adjusted based on actual measurements in practical settings.
(BIOHAZ) et al. (Thu,) studied this question.