Table 1.

Herd-level, episode- specific and bovine tuberculosis (bTB) management variables considered for inclusion in a Cox proportional hazards regression model of factors influencing bTB recurrence interval

Variable category	Variable name	Details
Herd-level variables	Disease Control Area (DCA)	Four geographical DCAs within New Zealand; 1 = northern North Island, 2 = southern North Island, 3 = northwestern South Island, 4 = southern South Island
	Vector Risk Area (VRA)	Location of herd with regard to VRAs; 1 = vector risk area, 0 = vector-free area
	Herd type	Herd type; 1 = dairy, 2 = dairy dry, 3 = beef breeding, 4 = beef dry, 5 = deer breeding
	Herd size	Herd size, defined as the number of animals in the index whole-herd test after detection of infection; includes all animals aged ⩾6 weeks present on the property
	Prior episodes	The number of infected episodes prior to the index episode in the study period
Episode-specific variables	Whole herd test	The number of skin test-positive animals at the first whole-herd test at or following the detection of the index bTB episode
	Skin test positive	The total number of skin test-positive reactions during the index episode
	Slaughter test-positives	The total number of animals slaughtered as test-positives during the index episode
	Lesions	The total number of animals detected with lesions during the index episode (includes test-positive post-mortem and meat inspection of non-test-positives)
bTB management variables	Parallel testing	Whether a parallel blood (IFN-γ) test was used at any time during the index episode (1 = yes, 0 = no)
	Clear	The combination of tests used at the final clearance test of the index episode (1 = skin test only all clear; 2 = skin test-positive animal(s) cleared on negative serial IFN-γ; 3 = skin or IFN-γ positive animal(s) cleared on slaughter with no visible lesions
	rxyn	Whether any test-positive animals were found by skin or IFN-γ testing at the final clearance test of the index episode (1 = yes, 0 = no)