Abstract
Foot and mouth disease (FMD) is a highly contagious viral disease of cloven hooved animals. The causal agent FMD virus (FMDV) is a picornavirus, which is globally distributed and has seven serotypes, namely serotype O, A, C, Asia 1, SAT 1, SAT 2 and SAT 3. FMD is endemic in Bangladesh and serotypes O, A and Asia 1 are currently circulating in the country. Many farmers vaccinate their cattle against FMD. The present study focused on epidemiologic modelling of FMD in Bangladesh. Historical data on FMD outbreaks over a period of one year (July 2012 to June 2013) and data on different putative risk factors were collected through a cross-sectional survey from three geographical regions, namely Shahjadpur Upazila (Sirajgonj district), Savar Upazila (Dhaka district) and Ghatail Upazila (Tangail district), having different cattle demography and husbandry practices. A logistic risk model was developed via univariate and multivariate analyses. Subsequently, active surveillance was conducted in the three regions during the period from July 2013 to June 2015. Virological and serological samples were collected from the outbreaks for the detection and serotyping of the virus by conventional and real time RT-PCR and detection of antibodies to structural and non-structural proteins (NSP) of FMDV by ELISA. RT-PCR products were sequenced and subjected to phylogenetic analysis. Spatial distribution of the outbreaks were analyzed in terms of Ripley's K function and spatial autocorrelation. An SIR (susceptible-infected-recovered) model was developed. Finally a more complex FMD simulation model was adapted to predict the impact of FMD control strategies. The initial survey revealed that about 22% of small and medium-scale farms or households (103 out of 458) experienced FMD outbreaks over a period of one year with 10% prevalence at the cattle population level. In addition, in a large farm (Savar Dairy Farm) only 1% (14 out 1224) cattle were affected. The logistic risk model showed that contact with feed sellers increased the chances of outbreaks 9.36 times; proximity of farms/households to inter district link road (< 1 kilometer) increased the chances of outbreaks 1.76 times and introduction of new animals in the farm/household increased the risk of FMD outbreaks two times. Vaccination was associated with a 74% reduction in risk of FMD outbreaks. During active surveillance 21 outbreaks (involving 164 farms or households) were recorded. Spatial analysis revealed that outbreaks in Shahjadpur region were dispersed or randomly distributed. In contrast, outbreaks in Savar and Ghatail regions were clustered in pattern. FMDV could be detected in 25 out of 37 samples, which included serotype O (Ind2001 lineage, ME-SA topotype), serotype A (genotype VII, Asia topotype) and serotype Asia 1 (genetic group VII) viruses. About 75% of serum samples had antibodies to NSP indicating exposure to field virus, which included both affected (44 out of 52) and in- contact (14 out of 25) cattle. Antibodies to FMDV structural proteins were detected against serotype O, A and Asia 1 in 44%, 47% and 72% serum samples, respectively, which were due to either exposure to field virus or vaccination. A total of 31 cattle developed clinical FMD despite vaccination probably due to short-lived or inadequate vaccinal immunity. The average basic reproduction number (Ro) of FMD was 8.26. A simulation model of FMD was successfully adapted using the transmission kernel and basic reproduction number determined from field data collected in the present study. The practical application of such a simulation model in evaluating FMD control strategies was demonstrated.