Use of regulatory data to describe and identify risk factors for salmonid rickettsial septicaemia in Chile, 2011–2017

The bacterial disease salmonid rickettsial septicaemia (SRS) has caused substantial economic losses in the salmonid aquaculture industry, particularly in Chile. Increasing understanding of the epidemiology of this disease is important for improving the productivity and sustainability of the industry: by identifying important influences on disease occurrence, interventions can be identified and optimised to minimise disease occurrence and its effect on the industry. This study aimed to describe SRS-attributed mortality across production species, time and space in the Chilean salmonid industry; identify risk factors for mortality attributed to SRS; and generate additional hypotheses for future investigations of SRS epidemiology in the Chilean salmonid industry. Regulatory data held in the Sistema de Fiscalización de la Acuicultura and Informe Ambiental para la Acuicultura databases of Servicio Nacional de Pesca y Acuicultura (‘Sernapesca’, the Chilean National Fisheries and Aquaculture Service) were analysed using the information-theoretic approach for risk factor analysis. Post hoc modelling generated hypotheses for future investigations. Analyses determined associations between the outcomes of SRS-attributed mortality incidence risk and rate across salmonid production cycles with a range of putative risk factors. Results indicate that salmonid production species is a key driver of the risk of a Chilean salmonid production cycle experiencing SRS-attributed mortality, and the SRS-attributed mortality rate amongst affected cycles. Compared to Atlantic salmon, rainbow trout and coho salmon production cycles were less likely to report SRS-attributed mortality (rainbow trout: incidence risk ratio = 0.17, 95% CI 0.07–0.40, p < .001; coho salmon: incidence risk ratio 0.05, 95% CI 0.02–0.11, p < .001). Amongst affected production cycles, coho salmon had a relatively reduced incidence rate of SRS-attributed mortality (incidence rate ratio = 0.32, 95% CI 0.20–0.49, p < .001) whilst rainbow trout had a relatively increased rate (incidence rate ratio = 3.60, 95% CI 2.51–5.17, p < .001). The incidence rate of SRS-attributed mortality was considerably lower in one region (Region XII: incidence rate ratio = 0.000003, 95% CI 0.0000003–0.00002; p < .001) compared to other regions (Regions X and XI). Further hypotheses for investigation generated in this study include associations between aspects of vaccination strategy and improved SRS-attributed mortality outcomes. This study highlights specific considerations for further epidemiological investigations using routinely collected Chilean aquaculture regulatory data.