The present study considered the risk factors associated with slaughtered cattle variables and the stratification of the property of origin of animals presenting suspicious lesions for bTB in the state of Mato Grosso. The univariate analysis concerning sex indicated a statistically significant difference. However, when performing the bivariate analysis, this variable became non-statistically different, demonstrating that samples with suspicious lesions collected by the bTB surveillance service do not display an associated positive diagnostic relationship to sex. This corroborates the results described by several authors (Biffa et al. 2011; Clegg, Good and More 2016; Okeke et al. 2014; Pascual-Linaza et al. 2017).
Regarding the univariate analysis, lesions obtained from zebu cattle were statistically different compared to other breeds. However, the bivariate analysis indicated no significant association between zebu breeds and other breeds, most likely due to the fact that meat production in the state of Mato Grosso is mainly of zebu cattle, which is more representative in the sampling due to of the amount of cattle sent for slaughter during the study period. Similarly, studies carried out in Nigeria indicated no statistically significant differences concerning positive M. bovis diagnoses associated with breed, thus corroborating the results obtained herein (Akinseye et al. 2018; Okeke et al. 2014).
A statistically significant difference in the univariate analysis concerning suspicious lesions associated with age group was observed only between 12–24 months and 24–36 months, and no statistically significant difference between these two age groups were observed in relation to > 36 months. However, the bivariate analysis demonstrated that the age group variable was not significantly associated to lesion positivity for M. bovis, similar to that reported by Okeke et al. (2014) in Nigeria. Other authors described statistically significant difference, pointing to a higher risk of positive bTB lesions associated with older animals (Akinseye et al. 2018; Biffa et al. 2011; Boukary et al. 2012; Pascual-Linaza et al. 2017). Other reports associated young animals to a higher frequency of severe bTB lesions (Biffa et al. 2012), demonstrating that lesions obtained from older animals are less likely to test positive for M. bovis (Shittu et al. 2013).
The univariate analysis regarding livestock circuits indicated no statistically significant difference in the frequency of injuries obtained from CP 01 and CP 02, and a statistically significant difference between the CP 03 and CP 04 circuits. In the bivariate analysis, however, no significant difference was observed among the livestock circuits, despite the constant frequency of bTB outbreaks in CP 02 (Leite) and CP 04 (Breeding). The bTB outbreaks were associated with dairy properties, where the health problem was attributed to greater animal density and long production cycles, corroborating Néspoli et al. (2016).
The health slaughter variable was statistically different concerning injuries obtained from animals positive for bTB (07/105) and animals commercially sent for slaughter (98/105) in the univariate analysis. In the bivariate analysis, slaughter was the only statistically different (p < 0.05) variable, indicating that animals with a positive diagnosis for bTB (positive for in vivo diagnosis) were 2.82 - fold more likely to present suggestive lesions confirmed as being caused by bTB by the nested q-PCR test.
Regarding the descriptive statistical analysis, 42.86% (07/03) of the injuries originating from slaughtered animals (07/105), were confirmed as positive for the molecular diagnosis of M. bovis. Cazola et al. (2015) reported that, from a total of 13 cattle positive in the intradermal test, 84.6% (11/13) were confirmed for M. bovis by the bacterial culture method (Cazola et al. 2015), demonstrating a high rate of disease confirmation. Fráguas et al. (2008) indicated that, from a total of 97 bovine animals reactive to the intradermal test, only 11.34% (11/97) were confirmed as M. bovis in bacterial cultures and postulated that this low bacterial recovery rate may be related to the difficulty of carrying out M. bovis cultures, which requires a minimum amount of viable bacilli, in addition to a decontamination technique applied during the process that can render this microorganism unfeasible.
Zarden et al. (2013) followed the slaughter of 24 cattle reactive to the tuberculin skin test, with 29.16% (n = 7) presenting suspicious lesions for bTB. When submitting samples without lesions for laboratory analysis, it was found that 73.16% (n = 19) of the animals tested positive when using bacterial culture, while 75.16% (n = 18) tested positive by the molecular method. The authors concluded that the combination of these methods increased sensitivity and specificity, with 95.8% (n = 23) of the animals testing positive for M. bovis, reinforcing that inconclusive tuberculinization results presented greater diagnosis chances through suspicious lesions. Figueiredo et al. (2009) performed a pool of samples of individual macroscopic lesions from 34 animals reactive to the tuberculin skin test and obtained a 50% (n = 17) positive diagnosis in the bacterial culture method, with 88.24% (n = 15) of the samples confirmed through multiplex-PCR.
Araújo et al. (2014a) collected 80 lesion samples from animals reactive to the intradermal tuberculin skin test, and observed presence of M. bovis DNA in 61.25% (n = 49) by the nested-PCR technique, with a greater number of animals testing positive for the pathogenic agent when compared to bacterial culture in the group of animals without information or without tuberculin tests (p < 0.05), that is, animals that were commercially sent to slaughter and observed as presenting lesions in the sanitary inspection line.
Risk factors may vary between surveys, due to regional differences, such as bTB prevalence, climatic variation, breed characteristics, management, production cycle, degree of property technification and strains. Each region may present intrinsic risk factors that must be considered when applying risk-based surveillance strategies, providing better resource application and promoting improvements in local agricultural defense management.
Taking into account that sex, breed, age group and livestock circuit were not significantly different or associated with factors that influence the diagnosis of bTB in the state of Mato Grosso and that slaughter sanitary conditions were statistically different (p < 0.05) in confirming the diagnosis of bovine tuberculosis, a greater investment in the in vivo diagnosis of the disease is suggested, as a way of increasing the chances of identifying and tracking new bovine tuberculosis foci in the state of Mato Grosso.
Given the above, the combination of intradermal and molecular tuberculin diagnostic tests potentiate the identification of M. bovis and the location of remaining disease foci, suggesting that the adopted strategy is appropriate for maintaining the health profile of the state of Mato Grosso.