Challenges of investigating outbreaks of infectious salmon anaemia in eastern Canada

The first known case of mortality due to infectious salmon anaemia (ISA) virus in Canada occurred in New Brunswick in the late summer of 1996. In an observational study of risk factors associated with ISA virus (ISAV) outbreaks, mortality patterns were used to define cases at the cage level and the proportion of case-cages at a site was used to define problem sites. The small number of sites involved in the original outbreak resulted in a reduced ability to detect site level associations and the need to account for the clustering of disease within a site when analysing cage level associations. Group identification was more commonly attached to the physical locations of net pens rather than to grouping of fish stocks. Obtaining mortality data from individual sites provided several challenges, due to inconsistencies in data recording frequency and data storage on the farms in the study. Diagnostic criteria which were unreliable and inconsistently applied in the early stages of the outbreak resulted in the need to define outbreak case retrospectively, based on unexplained mortality rates. Inconsistent or missing data on health and productivity at the cage level limited the ability to evaluate the important risk factors. Whenever possible, more reliable data were used as surrogate measures, such as noting the frequency of sea lice treatments to reflect lice burdens. Separating cause from effect for some factors, such as mortality dive frequency, was challenge due to the retrospective nature of this observational study. Direct versus indirect effects could not be differentiated in relation to such factors as moist feed. Despite these difficulties, this study detected several management factors which modified the relative risk of experiencing mortalities due to ISAV. Reduced risk was apparent with the increased health and survival of smolt in the first summer in seawater and with feeding moist feed during the winter and spring. Aggressive lice control and lower initial stocking numbers reduced the probability of ISA outbreaks. Multiple-year class sites, previous ISA cases at the site, divers visiting multiple sites and multiple-site companies were important site level risk factors for ISA outbreaks..