The most effective strategy for verifying the absence of sanitizer contamination in milk depends on the sanitizer type used in fluid milk processing facilities

Contamination of fluid milk with chemicals used for equipment sanitation is a safety and quality concern that has led to incidents of consumers being exposed to milk contaminated with sanitizer. In this study, we first surveyed dairy processors to determine common practices regarding the use of sanitizers in fluid milk processing facilities and the suitability of different preventive and verification practices as part of comprehensive best practices to prevent unintentional sanitizer contamination. Because the results of the survey showed that sensory evaluation is a common practice (73% of processors), we performed a series of sensory experiments to evaluate these practices. We determined the odor detection thresholds of the 2 most commonly used sanitizers, peroxyacetic acid- (PAA) and sodium hypochlorite-based sanitizers, in skim milk. Additionally, we assessed the impact of sensory training on the olfactory identification of milk with sanitizer contamination. Overall, we determined large variations in the odor detection thresholds for PAA in skim milk: 1.19 ± 47.72 mg/L (n = 11) and 13.63 ± 6.92 mg/L (n = 18) for panelists with and without previous training in sensory evaluation of milk, respectively. The odor detection thresholds for sodium hypochlorite were 6.32 ± 4.29 mg/L (n = 11) and 7.24 ± 6.40 mg/L (n = 16) for panelists with and without prior sensory evaluation of fluid milk experience, respectively. The additional sensory experiments performed during this study indicated that an individual's inherent ability may have a larger impact than sensory training on their ability to detect sanitizer in milk. Panelists could not consistently detect 7.2 mg/L sodium hypochlorite in skim milk by smell, whereas consistent odor detection was observed for some individuals at 20 mg/L sodium hypochlorite. Therefore, the use of acuity screening may improve the chances of sanitizer detection in milk by a human sensory panel. Other methods, such as pH measurement, alizarol tests, and the ferric reducing antioxidant power assay, were not able to distinguish between samples with added and without added sanitizers, making these methods unsuitable for detecting sanitizer contamination in fluid milk. The effect of added sanitizer on the freezing point depression of milk samples did not differ from the effect of the same amount of added water. Commercial peroxide test strips were found to be most appropriate for detecting the presence of PAA sanitizer in skim milk and could detect concentrations as low as 2 mg/L PAA. None of the evaluated commercial test strips were found to be sensitive enough to detect 20 mg/L sodium hypochlorite in milk. The insight gained from this work can be used by fluid milk processors to develop manufacturing and testing protocols that will minimize the risk of delivering fluid milk contaminated with sanitizer to consumers.