Application of quantitative analytic methods for the epidemiological study of zoonotic diseases
PhD Thesis defense by Mª Luisa de la Cruz Conty at the VISAVET Centre of the Complutense University of Madrid
December 19th, 2018
The present doctoral thesis explored the usefulness of the application of quantitative analytic methods for the study of the epidemiology of diseases of relevance in animal and public health.
Most epidemiological studies are based on the classification of individuals from a given population as infected or uninfected based on the results obtained in a diagnostic test.
However, often the accuracy of the test is not formally evaluated, or estimates obtained in a different (and non-applicable) context are merely extrapolated, what can lead to important biases. In this thesis, Bayesian models were applied in two very different scenarios in order to estimate the performance of several ante-mortem diagnostic tests. First, the accuracy of two diagnostic tests traditionally used for diagnosis of leishmaniasis in other species, the immunofluorescence antibody test (IFAT) and the nested PCR, was evaluated in rabbits and hares, that were recently identified as competent reservoirs of Leishmania infantum. Next, the performance of a recently commercialized test for diagnosis of bovine tuberculosis (bTB) based on the specific detection of gamma-interferon in the blood of infected animals that had been authorized for use as part of the Spanish bTB Eradication Program, the IDvet kit, was assessed.
Results of the first study estimated that the sensitivity and specificity of the IFAT were around 70-80%, thus suggesting the potential use of the IFAT as a screening tool for diagnosis of leishmaniasis in rabbits and hares. In contrast, the limited estimated sensitivity of the nested PCR (<30%) would discourage its use in this context. In the second study the estimated sensitivity of the IDvet test at the cut-off recommended by the manufacturer was 36.7% (14.7-78.8), significantly lower than estimates previously obtained for the Bovigam test, that had been used in the National bTB Eradication Program for over 10 years. These results point out the need of adjusting the IDvet cut-off for its application in the Spanish epidemiological setting.
Once the accuracy of the diagnostic test has been correctly determined, its results can be used as the basis for the evaluation of the spatio-temporal patterns of a disease. In the third study of this thesis, spatial analytic methods were used in combination with bacterial molecular characterization techniques to identify clusters of persistence of bTB in the region of Madrid in the period 2010-2012 to help recognize factors impairing the progress of the National Eradication Program in this region. Results demonstrated a heterogeneous distribution of the disease in Madrid, with a significant cluster of positive herds in the north of the region that was consistently identified throughout the study period. The high genetic homogeneity of the isolated strains and the spatial autocorrelation of the estimated transmission coefficients suggested the existence of local transmission events between neighboring farms and/or the effect of local factors shaping the epidemiology of the disease in the region.
The next step in an epidemiological study is the integration of the previous information in order to design and implement surveillance and/or control strategies. The constant evaluation of the efficacy of these strategies is fundamental; therefore, the following two studies in this thesis were focused on the evaluation of surveillance and control strategies for endemic diseases using different approaches. As suggested by the results of the second study, the success of the Spanish bTB Eradication Program is limited due in part to the fact that currently available diagnostic tests do not correctly determine the existence of an infection with the causative agent of bTB [members of the Mycobacterium tuberculosis complex (MTC), primarily M. bovis], what may lead to the misclassification of bTB infected herds as officially tuberculosis free (OTF) due to the absence of reactors in the single intradermal tuberculin (SIT) test. Therefore, the fourth study had the objective of estimating the herd sensitivity of the current diagnostic scheme at the herd level, defined as the probability of finding at least one reactor in a diagnostic test applied in a truly infected population, in OTF herds in the region of Castilla y León using scenario tree modeling. Model results revealed that the SIT test is able to detect three out of four newly infected herds in an initial herd-test and that, after two consecutive herd-tests carried out in the same herds, this number rises to nine out of ten. These results highlight the importance of the routine testing of farms even when considered free of disease in an area of high incidence. Regarding the bacteriological analysis, according to the model it was able to confirm the presence of MTC infection in approximately two out of three newly infected herds detected through the SIT test. This underlines the importance of limiting cattle movement originating from SIT positive herds even when the causative agent has not been isolated until the presence of infection can be conclusively ruled out. On the other hand, infection with Salmonella enterica is endemic in pig herds in Spain and, because consumption of improperly processed or cooked pork is considered one of the sources of foodborne salmonellosis, large efforts are invested in strategies to reduce its prevalence at the herd level, among which vaccination is one of the main ones. However, the great heterogeneity in vaccines and vaccination protocols complicates determining its true efficacy on the field; therefore, the fifth study of this thesis aimed to clarify this issue through a systematic review and meta-analysis of the peer-reviewed scientific literature. The information extracted and analyzed indicated that vaccination against Salmonella leads to a significant although moderate (~ 30%) reduction in the prevalence of colonization and/or excretion of S. Typhimurium and S. Choleraesuis serotypes in swine, regardless of the vaccine type (attenuated or inactivated) used.
|
