[Possibilities and limits of serological diagnosis in pig stocks in the case of Mycoplasma hyopneumoniae infection].

When conducting their investigations to diagnose infectious diseases in swine, practitioners are often forced to use reduced numbers of animals in their samples in order to minimize costs for farmers. A cross-sectional study was conducted approximating such field conditions to show the limits of interpretation with reduced sample sizes in case of Enzootic Pneumonia. Compared with other respiratory pathogens, Mycoplasma hyopneumoniae, the etiologic agent of Enzootic Pneumonia spreads very slowly, mainly when animals are in direct contact. Furthermore, the interpretation of serological results is difficult because several weeks must usually pass for serological reactions to become apparent. Serological testing is normally used to confirm a clinical diagnosis by detecting an increase in antibodies against the etiologic agent. Samples are collected at the beginning of disease and four to six weeks later. An increasing number of serological positive animals in a herd is usually interpreted as spread of infection. The ,,true" prevalence we observed in our investigation was used to make a statistical analysis describing the probability of detecting an increasing prevalence from 0.07 to 0.33 with a reduced sample size. We showed that the probability of detecting an increase of two seropositive animals was 44% if 5 samples per group were analysed. When only 3 samples were analysed per group, this probability decreases to 21%. Compromise must be found between epidemiological necessary and financially acceptable sample size; this could be a minimum of 10 samples per age group.