To analyse the disease dynamics in goat herds and to study the effect of control measures, a Q fever transmission model was developed by CVI. With such a model we can study questions like ‘What strategy can lower the incidence of Q fever in goats or even lead to disease extinction?’ and ‘Is vaccination a necessary component of a control strategy?’. For that purpose we first derived a deterministic model for the spread of Coxiella burnetii within a Dutch dairy goat herd housed in a deep litter stable. This model consists of a system of Ordinary Differential Equations (ODEs). Parameter values were based on literature and on-going experimental work, and knowledge gaps were identified. This deterministic model was used to study the Q fever dynamics in the herd without control measures. A sensitivity analysis of the model was performed to study which parameters are most important for the model outcome. Subsequently, a stochastic version of the model was developed to study the effect of separate and combined control and management measures in detail. This was necessary because stochastic processes become important when infected animals in a herd are controlled to very low numbers. The output of that model was subsequently used by LEI to perform a cost-benefit analysis of the control and management strategies. The most important results of the Q fever modelling study are as follows: • Due to the synchronised lambing season in combination with most C. burnetii excretion at partus and abortion, Q fever prevalence shows a saw-tooth pattern within a year. From year to year, a peak is observed in number of infected animals, abortions and infection pressure in year 1-2 after introduction of C. burnetii in the herd. • The epidemiological analyses show that vaccination is effective in quickly reducing the prevalence in a dairy goat flock. Other control strategies have more impact on flock management (more replacement of animals); (re-)infections can than lead to greater outbreaks due to the higher number of susceptible animals in the flock. • When looking at the average time to extinction of the infection and at the infection pressure in a goat flock, the best control strategy is “Preventive vaccination” (i.e. yearly), followed by the reactive vaccination strategies “Vaccination after abortion storm” and “Vaccination after BTM (bulk tank milk) positive”. • When aiming at a fast reduction of the Q fever infection pressure to the environment (like bacterial spores in dust), the strategy “Culling of pregnant animals after abortion storm” is effective. Then, no peak is observed anymore in number of infected animals, abortions and infection pressure in the goat herd. However, Q fever will not go extinct by this measure. • Infected lactating goats excrete C. burnetii bacteria intermittently, and in different amounts. A Search&Destroy method by PCR of milk with a sensitivity of 50% of detecting and culling infected goats, will not result in extinction of Q fever in the flock. This control measure was the worst of the six evaluated strategies. • Next to the reactive strategies, Q fever can be controlled by manipulating the frequency of pregnancy (and thus lambing) of goats. The herd management style “Every year pregnant” is always worse than the other two herd management styles “Every two year pregnant” and “First two years pregnant”. • C. burnetii can survive from breeding season to breeding season as intracellular bacterium in the goat (persistent infection). According to the model, these persistently infected goats are sufficient for the bacterium to survive from year to year in the flock. • The average lifetime of a goat in a Dutch dairy goat flock is 2.7 year. If vaccination will be stopped, within a few years most of the animals in the flock are without immunity against C. burnetii. This susceptible population is very vulnerable for (re-)introductions of Q fever. • During the modelling study, gaps in knowledge were identified, and assumptions had to be made. ¿ The transmission rate, the fraction of infected pregnant animals to become persistently infected and the fraction of C. burnetii excretion targeted to dust could not be estimated with certainty from literature data. Therefore, the infection dynamics in the model was studied for a range of values for these parameters. The parameter values which achieved a high abortion peak in the second season and a lower disease incidence in later seasons were used in the default model. ¿ Infection pressure caused by non-pregnant goats. It is unknown what the effect is of an infection in non-pregnant goats and how many and how long C. burnetii bacteria will be excreted to the environment. Also unknown is the fraction of these goats which becomes immune after acute infection. ¿ Infection pressure caused by spores of bacteria. C. burnetii can survive from breeding season to breeding season as bacterial spore. However, the fraction of surviving spores after one year is unknown, and so is their contribution to the infection pressure from stable environment to the goat flock. De conclusions of the economic analysis are: • As the dairy goat vaccination programme continues, future expenses in maintaining the current protected status are relatively low. • For Q-fever infected goat farms in the Netherlands preventive vaccination is always the preferred control strategy compared to other analysed control strategies. This also holds for a partial analysis if only on-farm costs and benefits are accounted for. • Averted human health costs depend to a large extend on the number of infected human cases per infected farm or animal. However, much is yet unknown with respect to goat-human transmission rates. • Preventive vaccination on Q-fever free farms should also be considered if the probability of re-infection is estimated to be more frequent than once per 15 to 20 years. • Freedom of Q-fever disease in which the agent is absent in livestock and environment would enable a return to non-vaccinated herds. However, much is yet unknown with respect to the probability and mechanisms of re-infection.
|Place of Publication||Lelystad|
|Publisher||Central Veterinary Institute of Wageningen UR|
|Publication status||Published - 2013|
|Publisher||Central Veterinary Institute of Wageningen UR|
Bontje, D. M., Backer, J. A., Roest, H. I. J., van Asseldonk, M. A. P. M., Bergevoet, R. H. M., & van Roermund, H. J. W. (2013). Control measures against Q fever in Dutch dairy goat herds: epidemiological and economical evaluation. (13/CVI0070). Central Veterinary Institute of Wageningen UR.