Table 1.

Predictors of M. bovis infection prevalence in adult badgers from 10 RBCT proactive culling areas, based on logistic regression

PredictorOdds ratio (95% C.I.)χ2dfP
Base model covariates
    Trial area253.498* <0.001
    Gender36.861<0.001
        Male vs. female1.54 (1.34–1.76)
    Tooth wear18.3150.003
        2 vs. 10.98 (0.54–1.78)
        3 vs. 11.24 (0.69–2.23)
        4 vs. 11.21 (0.67–2.19)
        5 vs. 11.64 (0.89–3.01)
        Not recorded vs. 10.45 (0.11–1.85)
    Carcass storage3.8610.050
        >7 days vs. ≤7 days0.64 (0.41–1.01)
    Necropsy laboratory26.699<0.001
    Culture laboratory1.5520.460
    Cull sequence 10.72710.394
    Cull sequence 252.611<0.001
Interannual variation
    FMD18.191<0.001
        2002 vs. other years1.70 (1.33–2.16)
Boundary permeability
    Permeability × cull sequence-1 interaction 3.1610.076
    Permeability × cull sequence-2 interaction 14.991<0.001
Capture location within trial area
    Capture location 9.4910.002
    Capture location × cull sequence-1 interaction 6.5810.010
  • C.I., confidence interval.

  • *The variable representing 10 trial areas is associated with only eight degrees of freedom because inclusion of the boundary permeability variable (a characteristic of each trial area) accounted for one of the trial area degrees of freedom.

  • Odds ratios associated with the cull sequence variables and their interaction terms are in Fig. 2. Fitted values are in Fig. 1d.