The five animals whose bacteriological results and DFA model classification did not agree have their test profiles presented in Table 4.
In all 5 animals whose bacteriological results and DFA model classification did not concur (WB13, 127, 128, 130, 171) M. bovis was cultured but the model classified them as negative.
The association of C-reactive protein, white blood cell count, vaginal sample bacteriological results, and a prediction model at admission, for early-onset neonatal infection and maternal chorioamnionitis were analyzed by comparing areas under the receiver operating characteristic curves and specificity.
We should also consider that results from the DFA model are concordant with our bacteriological data.
Overall, we successfully developed/optimized an aerosol neonatal macaque model that mimics clinical and bacteriological characteristics of early Mtb-infection as seen in human newborns/infants.
RT-PCR positivity for IS6110 gene along with high IFN-γ response in the eyes of infected animals as compared to non-infected mice without any bacteriological evidence indicate the establishment of a paucibacillary TB uveitis mouse model, a model mimicking the human TB uveitis condition.
The rate of misclassification (4.5%) is extremely low, taking into consideration that bacteriological culture results were not inputted into the model.
As a result, two prognostic models may be developed, with clinical and bacteriological cure after one week serving as dependent variables, respectively.
Statistical analyses of SCS and bacteriological titres were performed with ANOVA applied to mixed models (SAS® v.9.1), and with a non-parametric Wilcoxon test for viability and milk cell population.
DFA produced a model of TB status based on results from diagnostic tests other than bacteriological culture, which was then used to classify animals as presumably infected or not.
