Introduction

Swine producers and their veterinarians are always looking for better ways to detect, treat and prevent diseases in pigs. On indi- vidual farm sites, earlier more precise detection of clinical disease enables earlier more targeted treatment; which can lead to a faster response/recovery, improved performance and reduced econom- ic loss. Also, for production system areas, flows and networks; earlier more precise detection of clinical disease enables more proactive adaptation of prevention protocols – further enabling preserved targeted performance and profitability.

Clinical disease detection is typically the direct responsibility of growing pig owned and contract site managers, and is a function of skill, experience and time spent in the farm. However, detec- tion of clinical disease onset can be problematic due to variation in experience, time spent in the barn and distractions at certain times of the year like planting and harvesting. Continuous sound monitoring systems hold the potential to detect the onset of clinical respiratory disease earlier with greater consistency and reliability.

Materials and methods

Cough monitors (SOMO+ Respiratory Distress Monitor, SoundTalks NV, Leuven, Belgium) were obtained and installed in a 2400 head wean-to-finish barn. Pigs were placed into site per normal practice. During the first turn, 11 devices were installed, with four devices over the middle of the pens on each side of the building spaced equidistant from each other and three in the cen- tral alleyway spaced equidistant from each other. For subsequent turns, eight devices were installed, with four devices over the middle of the pens on each side of the building spaced equidis- tant from each other.

A respiratory distress index (RDI) was continuously generated from recorded sound files and uploaded to a cloud database.

RDI’s were continuously monitored and alerts were automatically sent to pre-determined personnel when a significant rise in RDI

was detected by the system. When an RDI alert was generated, diagnostic samples were collected and tested by PCR for PRRS, IAV-S, Mycoplasma hyopneumoniae, PCV2 and parainfluenza.

Also, the RDI data were charted and patterns of cough were cate- gorized. For each RDI episode, diagnostic samples via oral fluids, tracheal swabs and/or serum samples were collected and tested by PCR for PRRS, IAV-S, Mycoplasma hyopneumoniae, PCV2 and parainfluenza. RDI episodes were aligned with their correspond- ing diagnostic results and the resulting aggregate cough patterns were characterized.

Results and discussion

RDI episodes were detected at the site, including: IAV-S (H1N1), IAV-S (H3N2), and Mycoplasma hyopneumoniae. Dif- ferences in patterns of cough were observed between IAV-S and Mycoplasma hyopneumoniae.

Two distinctive RDI patterns were detected across the three farm sites, one associated with IAV-S (H1N1 or H3N2), and another associated with Mycoplasma hyopneumoniae. IAV-S associated RDI patterns had a distinctive bi-modal shape, whereas the pat- tern associated with Mycoplasma hyopneumoniae showed a gradu- al relatively linear rising pattern.

The detection of the respiratory disease episodes by the SOMO+ Respiratory Disease Monitor ranged from an estimated 3-5 days earlier than detection by farm personnel.

The ability to detect clinical respiratory disease earlier and clas- sify cough patterns according to primary etiology is useful at the individual site, flow and system levels. With this information, local site managers can better adjust and respond with more time- ly, appropriate diagnostic sampling and treatment. Further, those responsible for flows/systems and areas/networks can better as- sess larger scale behavior of specific disease agents and the clinical impact of intervention and control protocols.