Typescript (photocopy). / Digitized by Kansas Correctional Industries

Call number: LD2668 .T4 1985 S364 / Master of Science

No description available.

Master of Science / Food Science Institute / Valentina Trinetta / Salmonella is an important pathogen of public health concern. Each year, Salmonella costs the food industry approximately $2.3 billion. In recent years, the number of cases of Salmonella linked to pork products has also increased in the United States (US). Although pork has the lowest association with human foodborne illness when compared to beef and chicken, it is the most consumed meat in the world. Therefore, Salmonella is a significant food safety concern for the American swine industry. This pathogen can be present along all the food production chain from farm to fork and recent studies reported the isolation of Salmonella enterica serovar Typhimurium (ST) and its monophasic variant 4,[5],12:i:- (STM) in feed and feed ingredients. The occurrence of these pathogens in the pre-harvest environment can translate to entry and contamination of the human food chain. Nevertheless, little is known about Salmonella incidence and association with these types of environments. Hence, the purpose of this study was to investigate the presence and seasonal prevalence of Salmonella spp., ST and STM in selected feed mills, among the major US swine feed production areas. Eleven swine feed mills in eight different states were selected. Six mills produced only mash feed, while the other five facilities produced both mash and pelleted feed. Visits were conducted during fall 2016, early spring 2017 and summer 2017. Twelve environmental samples were collected within each facility and season, representative of the production flow, from receiving of ingredients to the finished product, including floor surfaces, equipment dust, workers’ shoes, and finished feed. Samples were analyzed following the USDA-FSIS guidelines and culture positive samples were analyzed by PCR. A multiplex PCR assay was also performed to differentiate Salmonella Typhimurium and Salmonella I 4,[5],12:i:- from the other serotypes. Associations between mill, season, mill type, sample site and Salmonella prevalence were analyzed using generalized linear mixed models (P < 0.05). From the 383 samples collected, 49 (12.8%) were identified as Salmonella spp.; two (5.1%) were identified from feed, while the other 47 (13.7%) originated from equipment or surfaces. Two samples were positive for ST and three for STM by multiplex PCR. Mill (P = 0.003) and season (P = 0.006) were statistically associated with the presence of Salmonella, with higher prevalence in fall and summer (13.2%) as compared to spring (3.6%). These findings demonstrate the seasonal prevalence of Salmonella spp., ST and STM in feed mills across the US, highlighting the potential role of the feed mill environment as a microbial entry route into the human food chain. The data presented can be also used as a tool to assist in the implementation of mitigation strategies for pre-harvest food safety.

No description available.

A presença de Salmonella é uma das mais importantes barreiras sanitárias à exportação de alimentos. A ração contaminada representa potencial fonte de introdução de Salmonella nos rebanhos suínos, além do risco indireto de infecção ao consumidor. Um estudo transversal conduzido em quatro fábricas de ração teve como objetivos avaliar a frequência de isolamento de Salmonella e coliformes totais nestas unidades, verificar a presença desses agentes nas diversas etapas do processo de produção, correlacionar grupos clonais de Salmonella sp. obtidos pela análise de macrorestrição associada à eletroforese em campo pulsado (PFGE), além de avaliar a concordância entre o escore obtido na aplicação de roteiro de inspeção da Instrução Normativa (IN) 4 do Ministério da Agricultura e Abastecimento e os níveis de coliformes totais encontrados. De 1.269 amostras analisadas, sessenta e três (4,96%) apresentaram Salmonella sp. e 38,53% (n=489) apresentaram presença de enterobactérias. As freqüências de contaminação por Salmonella nas quatro fábricas avaliadas (A, B, C e D) foram, respectivamente, 3,5% (n=11/317), 1,7% (n=5/289), 7,1% (n=23/308) e 7% (n=25/355). Já para coliformes totais foram, respectivamente, 40,7% (129/317), 30,1% (87/289), 52,6% (162/308) e 31,3% (111/355). Nas fábricas A e D, Salmonella foi detectada em amostras de produto final. Estirpes apresentando genótipos semelhantes foram identificadas nos sorovares Orion, Montevideo, Worthington e Agona. Foi possível verificar que o sorovar Montevideo obteve o maior número de grupos clonais, apresentando pulsotipos distribuídos entre ingredientes, poeira, equipamentos e ração. O isolamento de Salmonella foi significativamente mais frequente (p =0,002) em amostras com a presença (36/489; 7,36%) do que com ausência de coliformes totais (27/780; 3,46%). As contagens logarítmicas médias encontradas de coliformes totais, considerando todas as amostras analisadas por fábrica, foram: A: 0,97 (IC 95%: 0,81-1,13); B: 0,78 (IC 95%: 0,61-0,94); C: 1,32 (IC 95%: 1,16-1,49) e D: 0,91 (IC 95%: 0,75-1,06). Constatou-se elevada variabilidade no número de coliformes totais em todos os pontos amostrados nas quatro fábricas, e não houve diferenças significativas entre os pontos amostrados e entre as fábricas (p=0,174). A regressão logística tendo como variável resposta a presença de coliformes totais apontou maiores razão de chance (OR) de isolamento no transportador (OR=3,67; IC 95%: 2,28-5,89), dosagem (OR=9,51, IC 95%: 4,43-20,41), moagem (OR=7,1; IC 95%: 3,27-15,40), mistura (OR=4,08; IC 95%: 2,04-8,17), poeira (OR=3,50; IC 95%:2,10-5,84), resíduos (OR=6,22; IC 95%: 3,88-9,95) e fábrica C (OR=2,43; IC 95%: 1,68-3,53). Da mesma forma, o transportador (OR= 4,43; IC95%: 2,43-8,09) foi o local com maior probabilidade de isolamento de Salmonella, seguidos da poeira coletada nas dependências da fábrica (OR=2,88; IC95%: 1,41-5,88). Comparadas à fábrica B, as unidades C e D apresentaram, respectivamente, 2,74 e 2,83 mais chances de isolamento de Salmonella. Ao analisar os 128 itens necessários no Roteiro de Inspeção da IN4, a fábrica B obteve o menor número de não conformidades. Nas demais fábricas, as maiores inconformidades foram encontradas quanto à estrutura da área interna e externa das fábricas. As fábricas A e D foram as que apresentaram o maior número de itens não conformes relacionados à limpeza e higienização de equipamentos. Entre os 29 itens imprescindíveis na Avaliação de Estabelecimento e Procedimentos Operacionais Padrão, oito não foram cumpridos em pelo menos uma das fábricas. A fábrica B apresentou todos os itens em conformidade, ao passo que as fábricas A e D não cumpriram os oito itens. A fábrica C apresentou três itens não conformes, relacionados à limpeza de piso e parede, recepção e armazenamento de matérias-primas. Não houve concordância entre os escores obtidos no roteiro de avaliação e a média de coliformes totais encontrada nas fábricas, demonstrando que apenas a adequação à legislação pode não garantir a inocuidade do alimento produzido. A partir disso, conclui-se que é necessário implantar programas de monitoramento e controle microbiológico ao longo da linha de produção, independente da situação da fábrica em relação à legislação vigente. A elaboração e produção de equipamentos de fácil limpeza e com mínimo acúmulo de poeira e resíduos deve ser incluída nos programas de controle de Salmonella concomitante às demais medidas de controle preconizadas. / The presence of Salmonella is one of the most important sanitary barriers to food exports. The contaminated feed is a potential source of introduction of Salmonella in pig herds, and may also represent a risk to the consumer. A cross-sectional study conducted in four feed mills was carried out to evaluate the frequency of isolation of enterobacteria and Salmonella at various stages of feed production, to identify clonal groups of Salmonella sp. Obtained by macrorestriction analysis associated with pulsed field gel electrophoresis (PFGE), and to verify the association between the score obtained by feed mills after the application of the checklist of the Normative 4 (IN4) of the Ministry of Agriculture and Food Supply and the mean of total coliforms found in the unit. Among 1,269 samples analysed, sixty-three (4.96%) showed the presence of Salmonella. The four feeds mills evaluated (A, B, C and D) presented, respectively, 3,5% (n=11/317), 1.7% (n=5/289), 7,1% (n=23/308), and 7% (n=25/355) of positive samples. Total coliforms were isolated from 40.7% (129/317), 30.1% (87/289), 52.6% (162/308) e 31.3% (111/355) samples taken at unit A, B, C and D, respectively. In the units A and D, Salmonella was detected in samples of feed. Strains showing similar genotypes have been identified as Serovar Orion, Montevideo, Worthington and Agona. Serovar Montevideo presented the highest number of clonal groups, with common pulsotypes distributed among ingredients, dust, equipment and feed. The isolation of Salmonella was significantly higher (p=0,002) in samples with the presence of total coliforms (36/489; 7.36%) than in the negative ones (27/780; 3.46%). The mean logarithmic counts of total coliforms considering all samples were determined for each plant: A: 0.97 (95% CI: 0.81 to 1.13), B: 0.78 (95% CI: 0.61 to 0.94), C: 1.32 (95% CI: 1.16 to 1.49) and D: 0.91 (95% CI: 0.75 to 1.06). A high variability on the number of total coliforms in all the sampled points was found in all feed mills, and there was no statistical diference between the sampling sites and units (p=0.174). The logistic regression, with isolation of total coliforms as the response variable, showed a high odds ratio (OR) of isolation from conveyors (OR = 3.67, 95% CI: 2.28 to 5.89), dosing (OR = 9.51, 95% CI: 4.43 to 20.41), grinding (OR = 7.10, 95% CI = 3.27 to 15.40), mixing (OR = 4.08, 95 % CI: 2.04 to 8.17), dust (OR = 3.50, 95% CI: 2.10 to 5.84), crusts (OR = 6.22, 95% CI: 3.88 to 9.95) and unit C (OR = 2.43, 95% CI: 1.68 to 3.53). The conveyors (OR = 4.43, 95% CI: 2.43-8.09) were the most likely sites of isolation of Salmonella, followed by the dust collected on the premises of the plant (OR = 2.88, 95% CI: 1.41-5.88). Feed mills C and D showed, respectively, 2.74 and 2.83 higher odds for Salmonella isolation than feed mill B. The verification of the 128 items addressed on the Guidelines for Inspection of Normative 4 (IN 4) of the Brazilian Ministry of Agriculture and Food Supply pointed out that factory B had the lowest number of nonconformities. The other three feed mills, presented most faults related to the physical structure of the facilities. Feed mills A and D presented the largest number of nonconformities related to cleaning and sanitizing of equipment. Among the 29 mandatory items in IN4, eight were not met in at least one of the factories. Feed mill B presented all items accordingly, while plants A and D did not meet any of the eight items. Feed mill C had three non-compliant items related to the floor and wall cleaning, reception and storage of raw materials. There was no association between the score obtained on the inspection checklist and the level of total coliforms found in the sampled units, demonstrating that only the compliance to the legislation cannot guarantee the safety of feed produced. It was concluded that programs for monitoring and control microbial contamination throughout the production line are needed, regardless of the status of the plant in the relation to the current legislation.The design of equipment that allow easy cleaning and with minimal accumulation of dust and debris should be added to other measures in Salmonella control programs.

A presença de Salmonella é uma das mais importantes barreiras sanitárias à exportação de alimentos. A ração contaminada representa potencial fonte de introdução de Salmonella nos rebanhos suínos, além do risco indireto de infecção ao consumidor. Um estudo transversal conduzido em quatro fábricas de ração teve como objetivos avaliar a frequência de isolamento de Salmonella e coliformes totais nestas unidades, verificar a presença desses agentes nas diversas etapas do processo de produção, correlacionar grupos clonais de Salmonella sp. obtidos pela análise de macrorestrição associada à eletroforese em campo pulsado (PFGE), além de avaliar a concordância entre o escore obtido na aplicação de roteiro de inspeção da Instrução Normativa (IN) 4 do Ministério da Agricultura e Abastecimento e os níveis de coliformes totais encontrados. De 1.269 amostras analisadas, sessenta e três (4,96%) apresentaram Salmonella sp. e 38,53% (n=489) apresentaram presença de enterobactérias. As freqüências de contaminação por Salmonella nas quatro fábricas avaliadas (A, B, C e D) foram, respectivamente, 3,5% (n=11/317), 1,7% (n=5/289), 7,1% (n=23/308) e 7% (n=25/355). Já para coliformes totais foram, respectivamente, 40,7% (129/317), 30,1% (87/289), 52,6% (162/308) e 31,3% (111/355). Nas fábricas A e D, Salmonella foi detectada em amostras de produto final. Estirpes apresentando genótipos semelhantes foram identificadas nos sorovares Orion, Montevideo, Worthington e Agona. Foi possível verificar que o sorovar Montevideo obteve o maior número de grupos clonais, apresentando pulsotipos distribuídos entre ingredientes, poeira, equipamentos e ração. O isolamento de Salmonella foi significativamente mais frequente (p =0,002) em amostras com a presença (36/489; 7,36%) do que com ausência de coliformes totais (27/780; 3,46%). As contagens logarítmicas médias encontradas de coliformes totais, considerando todas as amostras analisadas por fábrica, foram: A: 0,97 (IC 95%: 0,81-1,13); B: 0,78 (IC 95%: 0,61-0,94); C: 1,32 (IC 95%: 1,16-1,49) e D: 0,91 (IC 95%: 0,75-1,06). Constatou-se elevada variabilidade no número de coliformes totais em todos os pontos amostrados nas quatro fábricas, e não houve diferenças significativas entre os pontos amostrados e entre as fábricas (p=0,174). A regressão logística tendo como variável resposta a presença de coliformes totais apontou maiores razão de chance (OR) de isolamento no transportador (OR=3,67; IC 95%: 2,28-5,89), dosagem (OR=9,51, IC 95%: 4,43-20,41), moagem (OR=7,1; IC 95%: 3,27-15,40), mistura (OR=4,08; IC 95%: 2,04-8,17), poeira (OR=3,50; IC 95%:2,10-5,84), resíduos (OR=6,22; IC 95%: 3,88-9,95) e fábrica C (OR=2,43; IC 95%: 1,68-3,53). Da mesma forma, o transportador (OR= 4,43; IC95%: 2,43-8,09) foi o local com maior probabilidade de isolamento de Salmonella, seguidos da poeira coletada nas dependências da fábrica (OR=2,88; IC95%: 1,41-5,88). Comparadas à fábrica B, as unidades C e D apresentaram, respectivamente, 2,74 e 2,83 mais chances de isolamento de Salmonella. Ao analisar os 128 itens necessários no Roteiro de Inspeção da IN4, a fábrica B obteve o menor número de não conformidades. Nas demais fábricas, as maiores inconformidades foram encontradas quanto à estrutura da área interna e externa das fábricas. As fábricas A e D foram as que apresentaram o maior número de itens não conformes relacionados à limpeza e higienização de equipamentos. Entre os 29 itens imprescindíveis na Avaliação de Estabelecimento e Procedimentos Operacionais Padrão, oito não foram cumpridos em pelo menos uma das fábricas. A fábrica B apresentou todos os itens em conformidade, ao passo que as fábricas A e D não cumpriram os oito itens. A fábrica C apresentou três itens não conformes, relacionados à limpeza de piso e parede, recepção e armazenamento de matérias-primas. Não houve concordância entre os escores obtidos no roteiro de avaliação e a média de coliformes totais encontrada nas fábricas, demonstrando que apenas a adequação à legislação pode não garantir a inocuidade do alimento produzido. A partir disso, conclui-se que é necessário implantar programas de monitoramento e controle microbiológico ao longo da linha de produção, independente da situação da fábrica em relação à legislação vigente. A elaboração e produção de equipamentos de fácil limpeza e com mínimo acúmulo de poeira e resíduos deve ser incluída nos programas de controle de Salmonella concomitante às demais medidas de controle preconizadas. / The presence of Salmonella is one of the most important sanitary barriers to food exports. The contaminated feed is a potential source of introduction of Salmonella in pig herds, and may also represent a risk to the consumer. A cross-sectional study conducted in four feed mills was carried out to evaluate the frequency of isolation of enterobacteria and Salmonella at various stages of feed production, to identify clonal groups of Salmonella sp. Obtained by macrorestriction analysis associated with pulsed field gel electrophoresis (PFGE), and to verify the association between the score obtained by feed mills after the application of the checklist of the Normative 4 (IN4) of the Ministry of Agriculture and Food Supply and the mean of total coliforms found in the unit. Among 1,269 samples analysed, sixty-three (4.96%) showed the presence of Salmonella. The four feeds mills evaluated (A, B, C and D) presented, respectively, 3,5% (n=11/317), 1.7% (n=5/289), 7,1% (n=23/308), and 7% (n=25/355) of positive samples. Total coliforms were isolated from 40.7% (129/317), 30.1% (87/289), 52.6% (162/308) e 31.3% (111/355) samples taken at unit A, B, C and D, respectively. In the units A and D, Salmonella was detected in samples of feed. Strains showing similar genotypes have been identified as Serovar Orion, Montevideo, Worthington and Agona. Serovar Montevideo presented the highest number of clonal groups, with common pulsotypes distributed among ingredients, dust, equipment and feed. The isolation of Salmonella was significantly higher (p=0,002) in samples with the presence of total coliforms (36/489; 7.36%) than in the negative ones (27/780; 3.46%). The mean logarithmic counts of total coliforms considering all samples were determined for each plant: A: 0.97 (95% CI: 0.81 to 1.13), B: 0.78 (95% CI: 0.61 to 0.94), C: 1.32 (95% CI: 1.16 to 1.49) and D: 0.91 (95% CI: 0.75 to 1.06). A high variability on the number of total coliforms in all the sampled points was found in all feed mills, and there was no statistical diference between the sampling sites and units (p=0.174). The logistic regression, with isolation of total coliforms as the response variable, showed a high odds ratio (OR) of isolation from conveyors (OR = 3.67, 95% CI: 2.28 to 5.89), dosing (OR = 9.51, 95% CI: 4.43 to 20.41), grinding (OR = 7.10, 95% CI = 3.27 to 15.40), mixing (OR = 4.08, 95 % CI: 2.04 to 8.17), dust (OR = 3.50, 95% CI: 2.10 to 5.84), crusts (OR = 6.22, 95% CI: 3.88 to 9.95) and unit C (OR = 2.43, 95% CI: 1.68 to 3.53). The conveyors (OR = 4.43, 95% CI: 2.43-8.09) were the most likely sites of isolation of Salmonella, followed by the dust collected on the premises of the plant (OR = 2.88, 95% CI: 1.41-5.88). Feed mills C and D showed, respectively, 2.74 and 2.83 higher odds for Salmonella isolation than feed mill B. The verification of the 128 items addressed on the Guidelines for Inspection of Normative 4 (IN 4) of the Brazilian Ministry of Agriculture and Food Supply pointed out that factory B had the lowest number of nonconformities. The other three feed mills, presented most faults related to the physical structure of the facilities. Feed mills A and D presented the largest number of nonconformities related to cleaning and sanitizing of equipment. Among the 29 mandatory items in IN4, eight were not met in at least one of the factories. Feed mill B presented all items accordingly, while plants A and D did not meet any of the eight items. Feed mill C had three non-compliant items related to the floor and wall cleaning, reception and storage of raw materials. There was no association between the score obtained on the inspection checklist and the level of total coliforms found in the sampled units, demonstrating that only the compliance to the legislation cannot guarantee the safety of feed produced. It was concluded that programs for monitoring and control microbial contamination throughout the production line are needed, regardless of the status of the plant in the relation to the current legislation.The design of equipment that allow easy cleaning and with minimal accumulation of dust and debris should be added to other measures in Salmonella control programs.

A presença de Salmonella é uma das mais importantes barreiras sanitárias à exportação de alimentos. A ração contaminada representa potencial fonte de introdução de Salmonella nos rebanhos suínos, além do risco indireto de infecção ao consumidor. Um estudo transversal conduzido em quatro fábricas de ração teve como objetivos avaliar a frequência de isolamento de Salmonella e coliformes totais nestas unidades, verificar a presença desses agentes nas diversas etapas do processo de produção, correlacionar grupos clonais de Salmonella sp. obtidos pela análise de macrorestrição associada à eletroforese em campo pulsado (PFGE), além de avaliar a concordância entre o escore obtido na aplicação de roteiro de inspeção da Instrução Normativa (IN) 4 do Ministério da Agricultura e Abastecimento e os níveis de coliformes totais encontrados. De 1.269 amostras analisadas, sessenta e três (4,96%) apresentaram Salmonella sp. e 38,53% (n=489) apresentaram presença de enterobactérias. As freqüências de contaminação por Salmonella nas quatro fábricas avaliadas (A, B, C e D) foram, respectivamente, 3,5% (n=11/317), 1,7% (n=5/289), 7,1% (n=23/308) e 7% (n=25/355). Já para coliformes totais foram, respectivamente, 40,7% (129/317), 30,1% (87/289), 52,6% (162/308) e 31,3% (111/355). Nas fábricas A e D, Salmonella foi detectada em amostras de produto final. Estirpes apresentando genótipos semelhantes foram identificadas nos sorovares Orion, Montevideo, Worthington e Agona. Foi possível verificar que o sorovar Montevideo obteve o maior número de grupos clonais, apresentando pulsotipos distribuídos entre ingredientes, poeira, equipamentos e ração. O isolamento de Salmonella foi significativamente mais frequente (p =0,002) em amostras com a presença (36/489; 7,36%) do que com ausência de coliformes totais (27/780; 3,46%). As contagens logarítmicas médias encontradas de coliformes totais, considerando todas as amostras analisadas por fábrica, foram: A: 0,97 (IC 95%: 0,81-1,13); B: 0,78 (IC 95%: 0,61-0,94); C: 1,32 (IC 95%: 1,16-1,49) e D: 0,91 (IC 95%: 0,75-1,06). Constatou-se elevada variabilidade no número de coliformes totais em todos os pontos amostrados nas quatro fábricas, e não houve diferenças significativas entre os pontos amostrados e entre as fábricas (p=0,174). A regressão logística tendo como variável resposta a presença de coliformes totais apontou maiores razão de chance (OR) de isolamento no transportador (OR=3,67; IC 95%: 2,28-5,89), dosagem (OR=9,51, IC 95%: 4,43-20,41), moagem (OR=7,1; IC 95%: 3,27-15,40), mistura (OR=4,08; IC 95%: 2,04-8,17), poeira (OR=3,50; IC 95%:2,10-5,84), resíduos (OR=6,22; IC 95%: 3,88-9,95) e fábrica C (OR=2,43; IC 95%: 1,68-3,53). Da mesma forma, o transportador (OR= 4,43; IC95%: 2,43-8,09) foi o local com maior probabilidade de isolamento de Salmonella, seguidos da poeira coletada nas dependências da fábrica (OR=2,88; IC95%: 1,41-5,88). Comparadas à fábrica B, as unidades C e D apresentaram, respectivamente, 2,74 e 2,83 mais chances de isolamento de Salmonella. Ao analisar os 128 itens necessários no Roteiro de Inspeção da IN4, a fábrica B obteve o menor número de não conformidades. Nas demais fábricas, as maiores inconformidades foram encontradas quanto à estrutura da área interna e externa das fábricas. As fábricas A e D foram as que apresentaram o maior número de itens não conformes relacionados à limpeza e higienização de equipamentos. Entre os 29 itens imprescindíveis na Avaliação de Estabelecimento e Procedimentos Operacionais Padrão, oito não foram cumpridos em pelo menos uma das fábricas. A fábrica B apresentou todos os itens em conformidade, ao passo que as fábricas A e D não cumpriram os oito itens. A fábrica C apresentou três itens não conformes, relacionados à limpeza de piso e parede, recepção e armazenamento de matérias-primas. Não houve concordância entre os escores obtidos no roteiro de avaliação e a média de coliformes totais encontrada nas fábricas, demonstrando que apenas a adequação à legislação pode não garantir a inocuidade do alimento produzido. A partir disso, conclui-se que é necessário implantar programas de monitoramento e controle microbiológico ao longo da linha de produção, independente da situação da fábrica em relação à legislação vigente. A elaboração e produção de equipamentos de fácil limpeza e com mínimo acúmulo de poeira e resíduos deve ser incluída nos programas de controle de Salmonella concomitante às demais medidas de controle preconizadas. / The presence of Salmonella is one of the most important sanitary barriers to food exports. The contaminated feed is a potential source of introduction of Salmonella in pig herds, and may also represent a risk to the consumer. A cross-sectional study conducted in four feed mills was carried out to evaluate the frequency of isolation of enterobacteria and Salmonella at various stages of feed production, to identify clonal groups of Salmonella sp. Obtained by macrorestriction analysis associated with pulsed field gel electrophoresis (PFGE), and to verify the association between the score obtained by feed mills after the application of the checklist of the Normative 4 (IN4) of the Ministry of Agriculture and Food Supply and the mean of total coliforms found in the unit. Among 1,269 samples analysed, sixty-three (4.96%) showed the presence of Salmonella. The four feeds mills evaluated (A, B, C and D) presented, respectively, 3,5% (n=11/317), 1.7% (n=5/289), 7,1% (n=23/308), and 7% (n=25/355) of positive samples. Total coliforms were isolated from 40.7% (129/317), 30.1% (87/289), 52.6% (162/308) e 31.3% (111/355) samples taken at unit A, B, C and D, respectively. In the units A and D, Salmonella was detected in samples of feed. Strains showing similar genotypes have been identified as Serovar Orion, Montevideo, Worthington and Agona. Serovar Montevideo presented the highest number of clonal groups, with common pulsotypes distributed among ingredients, dust, equipment and feed. The isolation of Salmonella was significantly higher (p=0,002) in samples with the presence of total coliforms (36/489; 7.36%) than in the negative ones (27/780; 3.46%). The mean logarithmic counts of total coliforms considering all samples were determined for each plant: A: 0.97 (95% CI: 0.81 to 1.13), B: 0.78 (95% CI: 0.61 to 0.94), C: 1.32 (95% CI: 1.16 to 1.49) and D: 0.91 (95% CI: 0.75 to 1.06). A high variability on the number of total coliforms in all the sampled points was found in all feed mills, and there was no statistical diference between the sampling sites and units (p=0.174). The logistic regression, with isolation of total coliforms as the response variable, showed a high odds ratio (OR) of isolation from conveyors (OR = 3.67, 95% CI: 2.28 to 5.89), dosing (OR = 9.51, 95% CI: 4.43 to 20.41), grinding (OR = 7.10, 95% CI = 3.27 to 15.40), mixing (OR = 4.08, 95 % CI: 2.04 to 8.17), dust (OR = 3.50, 95% CI: 2.10 to 5.84), crusts (OR = 6.22, 95% CI: 3.88 to 9.95) and unit C (OR = 2.43, 95% CI: 1.68 to 3.53). The conveyors (OR = 4.43, 95% CI: 2.43-8.09) were the most likely sites of isolation of Salmonella, followed by the dust collected on the premises of the plant (OR = 2.88, 95% CI: 1.41-5.88). Feed mills C and D showed, respectively, 2.74 and 2.83 higher odds for Salmonella isolation than feed mill B. The verification of the 128 items addressed on the Guidelines for Inspection of Normative 4 (IN 4) of the Brazilian Ministry of Agriculture and Food Supply pointed out that factory B had the lowest number of nonconformities. The other three feed mills, presented most faults related to the physical structure of the facilities. Feed mills A and D presented the largest number of nonconformities related to cleaning and sanitizing of equipment. Among the 29 mandatory items in IN4, eight were not met in at least one of the factories. Feed mill B presented all items accordingly, while plants A and D did not meet any of the eight items. Feed mill C had three non-compliant items related to the floor and wall cleaning, reception and storage of raw materials. There was no association between the score obtained on the inspection checklist and the level of total coliforms found in the sampled units, demonstrating that only the compliance to the legislation cannot guarantee the safety of feed produced. It was concluded that programs for monitoring and control microbial contamination throughout the production line are needed, regardless of the status of the plant in the relation to the current legislation.The design of equipment that allow easy cleaning and with minimal accumulation of dust and debris should be added to other measures in Salmonella control programs.