Mortality attributable to bluetongue virus serotype 8 infection in Dutch dairy cows

In 2007, bluetongue virus serotype 8 (BTV-8) re-emerged in the Netherlands and a large number of farmers notified morbidity and mortality associated with BTV-8 to the authorities. All dead cows in the Netherlands are registered in one of the three age classes: newborn calves <3 days, calves 3 days to 1 year, and cows >1 year. These registrations result in a complete data set of dead cattle per herd per day from 2003 until 2007. In this study, the mortality associated with BTV-8 for the Dutch dairy industry was estimated, based on this census data. Default, mortality associated with BTV-8 was estimated for the confirmed notification herds. Moreover, an additional analysis was performed to determine if mortality associated with BTV-8 infection occurred in non-notification herds located in BTV-8 infected compartments. A multivariable population-averaged model with a log link function was used for analyses. Separate analyses were conducted for the three different age groups. Confirmed notification herds had an increased cow mortality rate ratio (MRR) (1.4 (95% CI: 1.2-1.6)); calf MRR (1.3 (95% CI: 1.1-1.4)); and newborn calf MRR (1.2 (95% CI: 1.1-1.3)). Furthermore, in non-notification herds in BTV-8 infected compartments, mortality significantly increased 1.1 times (95% CI: 1.1-1.1) in cows, 1.2 times (95% CI: 1.2-1.2) in calves and 1.1 times (1.1-1.1) in newborn calves compared with BTV-8 non-infected months. Using objective census data over a 5-year period, the MRRs indicated increased mortality associated with BTV-8 infection not only in herds of which the farmer notified clinical signs but also in non-notification herds in infected compartments.