Spatial and temporal epidemiology of pseudorabies virus infection.

OBJECTIVE To examine the pattern of pseudorabies virus (PRV) infection in Pennsylvania and identify the area factors associated with herd quarantine status. SAMPLE POPULATION 123 PRV-quarantined commercial swine herds identified between 1986 and 1993 were selected as cases, and 162 uninfected herds were selected as controls. PROCEDURE Herd location, demographics, and temporal vaccination and quarantine data for a case-control study were obtained from producer questionnaires and state records, using a database of swine herds from 2 Pennsylvania counties. Any herd that was on quarantine as of Jan 1, 1991, or quarantined subsequent to this date, was defined as a case. A herd was defined as a control if it had never been quarantined for PRV. Controls were group matched to cases by year. Study herds were centered in a circle, or buffer zone, with a 1.61-km (1-mile), 3.22-km (2-mile), or 6.44-km (4-mile) radius, and densities of operation types, quarantined herds, nonquarantined herds, and vaccinated herds in the buffer zone were compared. The analytical outcome was the probability of a herd being quarantined, conditional on the buffer zone density of herds quarantined, herds not quarantined, and herds in which a PRV vaccine was used. These density variables were categorized into high, medium, and low, or just high and low categories. Confounding by year was assessed in the analysis. Analysis was performed, using unconditional logistic regression. RESULTS Decreased density of PRV-quarantined herds in the study region was associated with reduced risk of a herd becoming quarantined, whereas increased density of nonquarantined, presumably uninfected herds was associated with decreased probability of a herd becoming quarantined. Decreased density of vaccinated herds was associated with increased probability of a herd becoming quarantined. In addition, being a farrow-to-finish study herd was associated with increased probability of becoming quarantined, compared with being a feeder pig producer study herd. CONCLUSIONS Associations with quarantine status and area densities of vaccinated, nonquarantined, and quarantined herds indicate the importance of area spread in PRV control. These effects are seen most strongly at a 3.22-km (2-mile) radius, but also are seen at a 6.44-km (4-mile) radius.