Analysis of the Performance of the Animal Health Surveillance System in the Outbreak of Swine Vesicular Disease in the State of Santa Catarina - Brazil

Background: The occurrence of vesicular disease associated with Senecavirus A in a pig-producing region of Santa Catarina increased in 2015, reflected by the number of syndromic notifications to the official animal health service. In view of the recurrence of this event in 2018, the objectives of this study were to analyze the official data related to cases of suspected vesicular disease in pigs and to evaluate whether the experience in conducting the investigations of 2015 was incorporated into the years subsequent to 2015. We addressed this goal by analysis of the performance parameters of the state animal health surveillance system. Materials, Methods & Results: Descriptive analyses of data from official investigations of suspected vesicular disease in swine in different regions were carried out, and statistical models were used to: i) test the effect of the year on the age of the investigated injuries; ii) assess whether there was an association between the year and the type of outcome of the official investigation (discarded case or a probable case of vesicular disease, which resulted in the collection of samples for laboratory diagnosis and interdiction of the affected properties); iii) evaluate whether there was an association between the year and the detection of Senecavirus A RNA among the molecular analyses carried out after a case was classified as probable vesicular disease. From 05/22/2015 to 03/28/2019, there were 2093 notifications of suspected vesicular disease in pigs to the official service of Santa Catarina, with 1538 (73.5%) occurring in 2015 and 555 (26.5%) in subsequent years. After 2015, when compared to the base year, the chances of detecting late vesicular lesions (>3 days) were similar (increased 1.11 times, but there was no statistically significant association), in view of a panorama in which 55.29% of cases had the lesions classified as late throughout the analyzed period. The variation in the odds was relatively homogeneous among the regional units (Intraclass Correlation Coefficient [ICC] = 10.9%), but in the Sao Miguel do Oeste unit, it was significantly lower than the average. It was also lower than the average, in regional units with a high ratio of properties with pigs attended by the official veterinarian. The chances of cases being considered probable increased 32.3 times, but the descriptive analysis of the average interdiction period decreased. The estimated ICC was 34.9%, and in Campos Novos and Cacador, the chances of cases being probable were significantly higher than the average, while in the Canoinhas unit, the chances were significantly lower. The prevalence of cases with molecular detection of Senecavirus A was 78% lower after 2015. Discussion: Despite the increased sensitivity of the surveillance process for suspected vesicular events in the years after 2015, the specificity of the molecular diagnosis of Senecavirus A decreased. This is likely, because there was difficulty in sampling lesions with vesicular fluid and/or when the lesion epithelium had recently been disrupted, conditions most conducive to the identification of viral genetic material. In this context, the western region of the state, the main pig producer, continued to be the most accurate in confirming cases of Senecavirus A. Even with the improvements of laboratory diagnosis and the reduction of the interdiction period, it was important to standardize the differentiation of a late infectious lesion from a traumatic injury, as well as any association with other sample analyses, in order to minimize the disorders linked to the mischaracterization of a manifestation of Senecavirus A, without compromising the syndromic focus of its vesicular character.