RESUMO
Adulteration of milk is a common practice that concerns regulatory agencies, industry, and the population. Despite the growing need for checking adulteration, the current methods employed generally have low performance and are highly dependent on manual labor. This study aims to calibrate and validate a compact equipment (MilkoScan FT1) that adopts a Fourier transform infrared spectroscopy methodology to monitor adulteration in raw milk. Almost 2500 milk samples were used for reference spectrum construction and 1650 samples were used to validate the identification of the following five most commonly used adulterants (at three different concentrations each): (1) cornstarch, (2) sodium bicarbonate, (3) sodium citrate, (4) formaldehyde, and (5) saccharose, plus the additions of two levels of water or whey. To define the calibration with the best performance in milk adulteration identification, 12 calibrations involving 8, 10, 12, 14, 16, or 18 factors, with one or two outlier eliminations, were developed. The results of sensitivity and specificity analyses, as well as Kruskal-Wallis and Dunn multiple comparison tests, revealed that the calibration that best identified the adulterants was the one involving 14 factors, with a single elimination of outliers, exhibiting for all adulterants simultaneously, 84% sensitivity and 100% specificity. The calibration showed excellent sensitivity to cornstarch (>98%), sodium bicarbonate (100%), sodium citrate (99%), and formaldehyde (>84%), indicating that this calibration has good capacity for adulteration detection. Thus, this methodology is a viable option for the dairy industry to identify adulteration of raw milk.
RESUMO
The objective of this study was to compare the efficiency of washing and trimming broiler carcasses to reduce bacterial contamination. At the postevisceration site, 100 broiler carcasses were collected during 4 visits to a slaughterhouse in Santa Catarina State, Brazil. Birds were from the same flock, age, and approximately 2.4 kg of weight. Groups were as follows: group 1, with fecal contamination; group 2, without fecal contamination; group 3, with fecal contamination and trimmed; group 4, with fecal contamination and washed; group 5, with fecal contamination, and washed and trimmed. Carcass washings were performed with at least 1.5 L/bird of potable water (0.5 to 1 mg/kg of residual chlorine) at room temperature (20-25°C) using spray cabinets with 44 spray nozzles distributed into 2 chambers (pressure of 2 kgf/cm(2) and 4 kgf/cm(2)). Washed carcasses (trimmed or not) showed significantly (P < 0.05) lower counts of aerobic mesophiles (plate count agar) on the third evaluation, and even lower (P < 0.01) counts for total coliforms (CT) and fecal coliforms (Escherichia coli). Trimmed carcasses showed significantly lower counts (P < 0.05) for plate count agar; however, we observed higher counts for E. coli (P < 0.05). The association of both treatments (washing and trimming) showed significantly higher (P < 0.05) counts for coliforms (CT and E. coli). We can conclude that the washing method is overall more efficient than the trimming method to decontaminate chicken carcasses at the postevisceration site. Hopefully, our findings can help poultry companies to minimize production costs by applying the washing method for carcass decontamination.