Detection of Mycoplasma hyopneumoniae infection in live pigs is a critical component to measure the success of disease control or elimination strategies. However, in vivo diagnosis of M. hyopneumoniae is difficult and the imperfect sensitivity of diagnostic tools has been deemed as one of the main challenges. Here, the sensitivity of laryngeal swabs and deep tracheal catheters for detection of M. hyopneumoniae early and late after infection was determined using inoculation status as a gold standard in experimentally infected pigs and a Bayesian approach in naturally infected pigs. Three-hundred and twenty 8-week old seeder pigs were intra-tracheally inoculated with M. hyopneumoniae strain 232 and immediately placed with 1920 contact pigs to achieve a 1:6 seeder-to-contact ratio. A subset of seeders and contacts were longitudinally sampled at 7, 28, 97, and 113 days post-inoculation (dpi) and at 28, 56, 84, and 113 days post-exposure (dpe), respectively, using laryngeal swabs and deep tracheal catheters. Samples were tested for M. hyopneumoniae by a species-specific real-time PCR. The sensitivity of deep tracheal catheters was higher than the one obtained in laryngeal swabs at all samplings (seeders: 36% higher than laryngeal swabs at 7 dpi, 29% higher at 97 dpi, and 44% higher at 113 dpi; contacts: 51% higher at 56 dpe, 42% higher at 84 dpe, and 32% higher at 113 dpe). Our study indicates that deep tracheal catheters were a more sensitive sample than laryngeal swabs. The sensitivity of both sample types varied over time and by exposure method, and these factors should be considered when designing diagnostic strategies.
Keywords: Bayesian modeling; Deep tracheal catheter; In vivo; Laryngeal swab; Sensitivity.
Copyright © 2019 Elsevier B.V. All rights reserved.