Residency time and other risk factors for large bovine tuberculosis breakdowns in Irish cattle herds

Residency time and other risk factors for large bovine tuberculosis breakdowns in Irish cattle herds | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Residency time and other risk factors for large bovine tuberculosis breakdowns in Irish cattle herds Miriam Casey-Bryars, Jamie A. Tratalos, Andrew Byrne, Simon More, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8181407/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 10 You are reading this latest preprint version Abstract Bovine tuberculosis (bTB) poses challenges to sustainable livestock production, and to animal and human health globally. Its prevalence increased in Irish cattle herds between 2016 and 2023. Over the same period, bTB breakdowns (periods of trade restriction and enhanced surveillance within cattle herds due to bTB detection) in dairy herds saw an increase in the average and variability of bTB case numbers. To help inform Irish national control policy, we investigated the risk factors for large bTB breakdowns. Specifically, we set out to characterise the relationship between cattle residency times and case counts in Irish herds. For breakdowns ending between 2012 and 2023 (N = 37,176 breakdowns in 24,730 herds) we describe the association between herd size, cattle residency times, breakdown history, initiating test type, year, inward movements, neighbourhood burden and bTB breakdown case counts of greater than one, four and ten cases, using logistic regressions. For a subpopulation of 15,834 skin test initiated breakdowns, we investigated the same risk factors for increasing standard skin reactor counts using zero-truncated negative-binomial regression. We show that increasing herd size and cattle residency times are the strongest predictors of increased case counts. Herd management and initiating test type are also important but play smaller roles. Dairy and fattener herd types, and risk-based skin tests are associated with the largest breakdowns. Slaughterhouse-initiated breakdowns were associated with single cases. Interferon-γ surveillance policy evolved over the study period. When cases detected by this test type were excluded, year did not account for much variation within the estimated models. Herd- or animal-level measures of neighbourhood burden added the least explanatory ability to our models. Our results demonstrate that, while herd size and residency times are important factors driving large case counts, metrics of neighbourhood risk are not. This suggests that, following initial introduction of infection, cattle-to-cattle transmission within herds plays a more important role in amplifying infections rather than further introduction of infection from the neighbourhood. Targeting larger breeding herds for prevention of introduction of infection may assist with the control of bTB. The causes and consequences of larger breakdowns in fattener herds warrant further research. Bovine tuberculosis Mycobacterium bovis Residency time Herd size Statistical modelling Within-herd transmission Risk-factors Figures Figure 1 Figure 2 Figure 3 Introduction Mycobacterium bovis ( M. bovis ) is a zoonotic bacterium which causes tuberculosis in cattle and other animals. Due to its impact on human and bovine health, and to maintain access to export markets for cattle products, there has been a national eradication programme in place in Ireland since the 1950’s ( 1 , 2 ). After a large initial reduction in M. bovis burden in cattle during the earlier decades of the programme, progress plateaued. Between 2016 and 2024, the prevalence of bovine tuberculosis (bTB, specifically referring to tuberculosis in cattle) has increased in Ireland ( 3 ). This has caused challenges for the sustainability of Irish farming and high costs to tax-payers and farmers. In 2024, the programme cost the Irish exchequer €100.6 million, with €54.4 million allocated to compensating farmers for bTB-related cattle culling ( 4 – 6 ). Ireland’s lowest recorded annual herd-level incidence occurred in 2016, with 3.30% of herds reporting new breakdowns. However, incidence has since increased, reaching 6.04% in 2024 ( 4 , 6 , 7 ). Because of its complex multispecies epidemiology, including evidence for circulation of infection between cattle and wildlife ( 8 – 12 ), there is robust debate about what is driving the recent increase in prevalence and the potential effectiveness of interventions. Cattle herd management systems have evolved over the past decade and mixed and dairy herds have grown larger ( 3 , 13 ). There have also been increased numbers of fattener herds, dairy herds which purchase their replacement heifers from separate enterprises, and the associated heifer rearing store herds ( 13 ). Although there has been a badger BCG vaccination campaign in place in recent years, there has been a concurrent increase in the number of badgers culled. Between 2014 and 2017, 24,121 badgers were culled (~ 6,030 per annum) whereas between 2020 and 2023, 21,393 were culled (~ 7,131 per annum) ( 14 ). Extensive efforts and policy changes have been implemented to enhance detection of infected cattle, including training and auditing to improve quality of skin testing by private veterinary practitioners ( 2 , 3 , 15 – 17 ). However, as of mid 2025, no major changes have been applied in terms of cattle movement controls ( 7 ). Answering the question of why, despite extensive and expensive controls, the prevalence of bTB is increasing, is critical for the Irish livestock industry. As well as herd-level breakdown occurrence, breakdown case counts (the total number of bTB positive cattle detected during single breakdowns) increased between 2016 and 2024, in particular amongst dairy and mixed herd management systems. There has also been an increase in lesions detected per attested bovine sent for routine slaughter, suggesting a genuine increase in infection levels rather than only increased ante-mortem surveillance sensitivity ( 3 ). In this paper, we focus on quantifying drivers of case counts disclosed during Irish bTB breakdowns, capitalising on rich datasets spanning more than two decades on cattle movement, demographics and bTB testing. For clarity, we highlight that we are not addressing risk-factors for bTB breakdown occurrence, which have been previously well reported ( 8 , 18 , 19 ). Rather, we are investigating risk factors to explain the larger case-counts within breakdowns, which have been occurring in dairy and mixed herd types over recent years. A previous study in Ireland, Clegg et al. (2018), investigated risk factors for case counts of greater than 13 versus case counts of between 2 and 4 in breakdowns beginning in 2014 or 2015 ( 20 ). This study found that the year the breakdown started, increasing herd-size, previous exposure to bTB, increased bTB incidence in the local area, the presence of an animal with a bTB lesion and a bTB breakdown in a herd with shared staff / management / facilities were risk factors associated with larger breakdowns. Further studies have used breakdown duration ( 21 – 23 ) or additional cases detected within the breakdown after the initiating test ( 24 , 25 ) as indirect evidence of M. bovis transmission, with many risk factors identified, in particular herd size and lesioned ante-mortem reactors, overlapping with those of Clegg et al. (2018) ( 20 ). We build on this previous work, considering a broader range of breakdown case counts, a wider time window, and incorporating a range of extra metrics of movement and neighbourhood burden. Furthermore, we incorporate a time-window comprising substantial demographic changes in the Irish cattle industry ( 3 , 13 ). For the first time, we consider cattle residency times (the amount of time an animal resided in a herd before being sent to another herd or for slaughter) within herds as an explanatory variable for case counts. For infectious diseases, the potential for transmission increases with the duration of contact between infectious and susceptible pairs ( 26 ). As a proxy time for contact time between individuals, we would therefore expect residency time to also be an important correlate for risk of transmission. Within such a heavily managed test-and-cull system, the observed patterns of case counts are driven by both the dynamics of transmission of infection and removal of infection through the surveillance system. Understanding how to control bTB in Ireland depends on understanding the interaction of these two, which requires mechanistic models of transmission ( 27 – 30 ). As a first step towards the construction of such models for Ireland, and to improve our understanding of the factors underlying the distribution of case counts within herds, we carried out the descriptive analysis and risk factor study presented here. Potential drivers of detection of infected cattle in the herd, and subsequent case counts, include retained infected cattle from previous breakdowns, introduction of infection from the neighbourhood, inward movement of infected cattle and cattle-to-cattle transmission within the herd. Surveillance system characteristics, including diagnostic test characteristics, policy and implementation practices also influence case counts. Breakdowns are initiated by bTB case detections. These comprise lesion detections in attested cattle going to slaughter for human consumption, or alternatively through Ireland’s ante-mortem bTB surveillance programme based on the comparative intra-dermal skin test (skin test hereafter) ( 9 ). In breakdowns with larger case counts, interferon gamma testing may be implemented to detect additional infected cattle ( 15 , 31 ). Within the managed system of bTB control in Ireland, it is challenging to differentiate between the contribution of surveillance intensity and true within-herd prevalence of infection on the observed patterns of incidence. For example, fattener herds have more intensive slaughterhouse surveillance compared to dairy herds as they have greater rates of movement to slaughter ( 32 ). To address this, as well as considering the full population of breakdowns (Full population A : total cases), where the response variable comprises count of all bTB cases, we consider a sub-population of breakdowns (Sub-population B : standard skin cases). In this sub-population, breakdowns initiated by slaughterhouse cases are excluded. Further variation in case count due to heterogeny in surveillance effort arises from evolving interferon-γ testing policy over our study period ( 15 ), and to a lesser degree, that for application of severe skin test interpretation ( 33 ). Therefore, for our sub-population, we consider count of standard skin test reactors as our response variable, as policy relating to identifying these has not changed. Finally, although the specificity of the skin test is likely to be high ( 34 – 36 ), we exclude the possibility of breakdowns due to false positive case detections within our “Sub-population B : standard skin cases,” considering only those with at least one ante-mortem reactor which went on to have a lesion detected at slaughter. Based on this approach, our study is based on two populations, for which we present separate, but complementary analyses. For our full, unfiltered population of breakdowns initiated by any method (“Full population A : total cases”) all methods of bTB case detection contribute to our case count response variable. Our sub-population of breakdowns (“Sub-population B : standard skin cases”) are filtered, with the rationale above, to exclude slaughterhouse initiated breakdowns and those with no bTB lesion reported upon post-mortem examination ante-mortem reactors. The response variable for this sub-population is count of standard skin test reactors. With these analyses, we aim to take an initial step towards quantifying the relative roles of potential drivers of M. bovis transmission within herds. Methods Data Cattle movement and bTB surveillance data were available from between 2000 and 2023 ( 31 , 37 ). Detailed herd management data, integrated with cattle movements and demographics, were available from 2008 onwards ( 13 , 37 ). Up-to-date data on herd location was available through the Irish Land Parcel Identification system ( 38 ). We describe data generation and processing in detail in our Supplementary Methods but give a brief overview here. Full population A : total cases “Full population A : total cases” was defined to address our primary question examining a response variable defined by the total count of cases within bTB breakdowns ending between 2012 and 2023. This total case count includes skin test reactors at both the standard and severe interpretation, interferon-γ test reactors and slaughterhouse cases. The choice of time period was informed by the requirement of two- and three-year windows (respectively) in advance of the year of breakdown start to calculate our cattle movement and herd-level neighbourhood burden metrics (Supplementary Methods Table 2 ). Therefore, using the more complete movement data from 2008 onwards and allowing for our average three year in degree metric (average count of unique source herds for inward moving cattle in the three years preceding the breakdown start), meant excluding breakdowns beginning before 2010. To avoid selection of only shorter breakdowns in earlier years, we excluded breakdowns ending in 2010 and 2011. This choice was supported by only 0.5% of our full breakdown population having a duration longer than two years. We used year of breakdown end for our analyses to avoid right censoring issues. After applying these constraints we are left with a study population of breakdowns ending between 2012 and 2023, and starting after 2009. Our residency time calculations required herds which comprised at least a single animal each year between 2008 and 2023. We also excluded the 2% of breakdowns in herds which comprised ten cattle or fewer, as their practices may be more aligned with hobby-herd management rather than the broader cattle industry. Sub-population B : standard skin cases From “Full population A : total cases”, we selected the subset of skin-test initiated breakdowns which we termed “Sub-population B : standard skin cases”. With the rationale, as described in our Introduction, of reducing variation due to heterogeneity in surveillance effort, and focussing on drivers of M. bovis transmission, we excluded breakdowns initiated by slaughterhouse surveillance, and included only those with at least one lesioned skin test reactor. To reduce potential bias from combining multiple surveillance mechanisms which evolved over time, the outcome variable for “Sub-population B : standard skin cases” was counts of standard skin test reactors only. Explanatory variables Explanatory variables included herd size (as calculated by averaging the herd sizes profiled from AIM movement data on the first of January, May and September each year), mean residency time of cattle leaving the herd the year of the breakdown start, initiating test type of the breakdown, herd breakdown history (categorised as between 0 and 5 + years of clear surveillance in advance of the breakdown start), in degree (count of unique source herds for inward moving cattle averaged for the year of breakdown start and the two preceding years), year of breakdown end, and a metric of neighbourhood burden. Only 1% of our breakdowns were initiated by pre-movement testing, so, similarly to Clegg et al. (2018) ( 20 ), we merged this category with annual skin testing. Both herd and animal level metrics have been previously used to measure neighbourhood burden of bTB ( 8 , 20 , 39 , 40 ). Adapting from Tratalos et al. (2023) ( 8 ), our herd level metric was the proportion of herds within six kilometres of the herd of interest centroids, with a breakdown history within the past three years. We also investigated the animal level metrics explored by Clegg et al (2018) ( 20 ) and Houtsma et al. (2018) ( 41 ). These were, for the year before the breakdown start, skin reactors per 1000 cattle tested, or skin reactors per square kilometre, in the electoral division of the herd. The mean size of an Irish electoral division is 20.6 km 2 ( 8 ). Only one neighbourhood related variable was considered at a time in our models. After initial univariable and bivariable explorations, all continuous explanatory variables were categorised into quintiles. Statistical analyses The distributions of case counts were plotted. The fit of different statistical distributions to the data were compared using Akaike’s Information Criterion (AIC). We trialled zero-truncated Poisson (zt-P), zero-truncated negative binomial (zt-nb) and zeta distribution fits. We attempted to fit distributions to total case counts in “Full population A : total cases”, and separately to standard skin reactors in “Sub-population B : standard skin cases”. Where models did not converge or exhibited evidence of poor fit, case counts were binned, and a series of logistic regression models fitted to the data to explain case counts greater than one, four and ten. We used multivariable generalised linear models to explore the associations between the explanatory variables and measures of case counts. As 32.8% of herds in our dataset contributed data from more than one breakdown, all of our models included a herd-level random effect on the intercept. Candidate variables for fixed effects comprised all potential explanatory variables justified biologically and from descriptive analyses ( 42 ). Likelihood ratio testing was used to describe explanatory ability added by each variable to the models. Multicollinearities between explanatory variables were investigated by calculation of variance inflation factors. All analyses were conducted in the R Statistical Environment ( 43 ). The R packages “VGAM” ( 44 ) and “glmmTMB” ( 45 ) were used for model implementations. Results Descriptive analyses Our full population comprised 37,176 breakdowns ending between 2012 and 2023, in 24,730 herds. Of these, 40.8% were beef-breeding, 26.9% dairy, 19.3% fattener, 9.0% mixed, 3.2% store and 0.7% trader (classified with reference to ( 32 )). In this population, there were 106,984 standard skin reactors, 32,433 severe skin reactors, 37,557 interferon-γ test reactors and 13,457 slaughterhouse cases. Amongst reactors, 44.2% of standard interpretation skin, 17.5% of severe interpretation skin and 19.7% of interferon-γ reactors had lesions detected at post-mortem examination. Table 1 and Fig. 1 show case count summary statistics stratified by candidate explanatory variables. Broad descriptive analyses are reported in Supplementary Results 1. Case count increased with herd size (Fig. 1 A), residency time (Fig. 1 B) and neighbourhood burden (Supplementary Fig. 1.9). In the case of in-degree (inversely related to residency time, Fig. 1 A), the highest mean and 75th percentile case counts were associated with an in-degree category of 1, compared to an in degree of 0, or higher in-degree (Table 1 ). The lowest case counts were associated with the highest in-degree quintile. Dairy herds had the highest case counts, followed by mixed, beef breeding, fattener, store and trader management types in deceasing order of case count (Supplementary Fig. 1.4A, Table 1 ). In our full population, 41.0% of breakdowns were initiated by annual skin testing/pre-movement testing, 24.6% by a slaughterhouse lesion detection,18.1% by contiguous skin testing, 11.0% by post-derestriction skin testing and 5.3% by risk-based skin testing. Breakdowns initiated by lesion detection at routine slaughterhouse inspection had the lowest case counts (mean 4, median 1), followed by annual skin tests (mean 5, median 2). Risk-based, contiguous and post-derestriction skin test-initiated breakdowns had higher mean case counts and 75th percentiles (Table 2 , Fig. 1 C). Herds which have been without a breakdown for four or more years had lower case counts compared to breakdowns with more recent breakdown history. Year of breakdown end is not shown in Table 2 but summarised in Supplementary Fig. 1.3 and considered in detail by Madden et al. (2025) ( 3 ). Table 1 Summary statistics for case count stratified by potential explanatory variables. *“Sub-population B : standard skin cases” was restricted to breakdowns initiated by skin tests (excluding slaughterhouse initiated breakdowns) and breakdowns with at least one ante-mortem reactor which went on to have lesions detected at slaughter, and therefore had “NA” values in the corresponding table rows for these characteristics. Characteristic Full population A : total cases N = 37,176 1 Sub-population B : standard skin cases N = 15,834 1 Total cases Mean (sd) Total cases Median (IQR) Standard skin positives Mean (sd) Standard skin positives Median (IQR) At least one lesioned skin test reactor present No 2.2 (3.1) 1.0 (1.0, 2.0) NA* NA* Yes 8.9 (16.4) 3.0 (2.0, 9.0) 4.7 (7.5) 2.0 (1.0, 5.0) Herd size category [ 11, 43) 2.6 (3.3) 1.0 (1.0, 3.0) 2.8 (2.8) 2.0 (1.0, 3.0) [ 43, 78) 3.6 (5.6) 1.0 (1.0, 3.0) 3.7 (4.5) 2.0 (1.0, 4.0) [ 78, 126) 4.7 (8.1) 2.0 (1.0, 4.0) 4.7 (6.2) 2.0 (1.0, 5.0) [126, 208) 6.3 (11.5) 2.0 (1.0, 6.0) 5.7 (7.8) 3.0 (1.0, 7.0) [208,1932] 9.3 (21.1) 2.0 (1.0, 7.0) 7.6 (12.7) 3.0 (2.0, 8.0) Mean residency time in years category [ 0.02, 1.07) 2.7 (6.3) 1.0 (1.0, 2.0) 3.0 (5.2) 2.0 (1.0, 3.0) [ 1.07, 1.60) 4.2 (9.0) 2.0 (1.0, 4.0) 3.8 (5.7) 2.0 (1.0, 4.0) [ 1.60, 2.10) 5.4 (12.1) 2.0 (1.0, 5.0) 4.5 (6.7) 2.0 (1.0, 5.0) [ 2.10, 2.80) 6.4 (12.8) 2.0 (1.0, 6.0) 5.4 (8.6) 3.0 (1.0, 6.0) [ 2.80,18.00] 7.9 (16.4) 3.0 (1.0, 8.0) 6.2 (9.0) 3.0 (2.0, 7.0) Herd management type Beef breeding 4.4 (8.5) 2.0 (1.0, 4.0) 4.1 (5.6) 2.0 (1.0, 5.0) Dairy 8.5 (17.7) 3.0 (1.0, 8.0) 6.6 (10.6) 3.0 (2.0, 7.0) Fattener 2.9 (6.6) 1.0 (1.0, 2.0) 3.6 (5.6) 2.0 (1.0, 4.0) Mixed 6.2 (13.0) 2.0 (1.0, 6.0) 5.5 (8.1) 3.0 (1.0, 6.0) Store 2.5 (3.5) 1.0 (1.0, 2.0) 2.5 (2.8) 1.0 (1.0, 3.0) Trader 1.7 (2.3) 1.0 (1.0, 2.0) 2.1 (1.5) 2.0 (1.0, 2.0) Initiating test of breakdown Annual skin 5.0 (10.5) 2.0 (1.0, 4.0) 4.3 (6.9) 2.0 (1.0, 4.0) Contiguous 6.5 (11.4) 3.0 (1.0, 7.0) 4.9 (6.9) 3.0 (1.0, 5.0) Post-derestriction 7.0 (14.3) 3.0 (1.0, 7.0) 5.6 (8.9) 3.0 (1.0, 6.0) Risk based 7.9 (15.0) 2.0 (1.0, 8.0) 6.0 (9.5) 3.0 (1.0, 6.0) Slaughterhouse 3.5 (12.3) 1.0 (1.0, 2.0) NA* NA* Proportion of herds within 6km with breakdowns in 3 years preceding breakdown start [0.00,0.08) 4.5 (10.1) 1.0 (1.0, 3.0) 4.3 (6.7) 2.0 (1.0, 5.0) [0.08,0.12) 4.9 (12.0) 1.0 (1.0, 4.0) 4.5 (8.3) 2.0 (1.0, 5.0) [0.12,0.16) 5.3 (11.9) 2.0 (1.0, 4.0) 4.7 (7.3) 2.0 (1.0, 5.0) [0.16,0.21) 5.6 (12.6) 2.0 (1.0, 5.0) 4.9 (7.6) 2.0 (1.0, 5.0) [0.21,0.70] 6.0 (12.5) 2.0 (1.0, 5.0) 5.1 (7.3) 3.0 (1.0, 6.0) Skin reactors per 1000 cattle tested in district electoral division in year preceding breakdown start. 0 4.7 (11.5) 1.0 (1.0, 4.0) 4.5 (8.0) 2.0 (1.0, 5.0) [0.0404, 0.4705) 5.4 (12.9) 2.0 (1.0, 4.0) 5.0 (8.3) 2.0 (1.0, 5.0) [0.4705, 2.0822) 5.3 (11.7) 2.0 (1.0, 4.0) 4.8 (7.4) 2.0 (1.0, 5.0) [2.0822,258.0645] 5.7 (12.0) 2.0 (1.0, 5.0) 4.8 (7.0) 2.0 (1.0, 5.0) Standard reactors per square kilometer in district electoral division in year preceding breakdown start. 0 4.8 (11.6) 1.0 (1.0, 4.0) 4.4 (7.2) 2.0 (1.0, 5.0) [0.00805,7.29e-02) 4.8 (10.7) 2.0 (1.0, 4.0) 4.6 (7.2) 2.0 (1.0, 5.0) [0.07289,2.54e-01) 5.2 (11.8) 2.0 (1.0, 4.0) 4.7 (7.8) 2.0 (1.0, 5.0) [0.25429,1.05e + 02] 5.8 (12.7) 2.0 (1.0, 5.0) 4.9 (7.5) 2.0 (1.0, 5.0) In-degree category (count unique source herds per annum averaged for two years preceding breakdown and year of breakdown start) 0 5.7 (11.6) 2.0 (1.0, 5.0) 4.7 (6.8) 2.0 (1.0, 5.0) 1 6.6 (14.8) 2.0 (1.0, 6.0) 5.4 (9.2) 3.0 (1.0, 6.0) [ 2, 4) 6.0 (12.0) 2.0 (1.0, 5.0) 5.1 (7.7) 3.0 (1.0, 6.0) [ 4, 12) 5.2 (12.8) 2.0 (1.0, 4.0) 4.4 (6.8) 2.0 (1.0, 5.0) [12,5621] 2.9 (6.9) 1.0 (1.0, 2.0) 3.6 (6.3) 2.0 (1.0, 3.0) Years without bTB breakdowns prior to breakdown start 0 years clear 6.2 (13.3) 2.0 (1.0, 6.0) 5.5 (8.4) 3.0 (1.0, 6.0) 1 year clear 5.8 (13.5) 2.0 (1.0, 5.0) 5.2 (8.7) 3.0 (1.0, 5.0) 2 years clear 6.0 (12.3) 2.0 (1.0, 5.0) 5.4 (8.2) 3.0 (1.0, 6.0) 3 years clear 6.0 (16.1) 2.0 (1.0, 5.0) 5.2 (10.5) 3.0 (1.0, 5.0) 4 years clear 5.6 (13.3) 2.0 (1.0, 4.0) 5.0 (8.5) 2.0 (1.0, 5.0) 5 + years clear 4.8 (10.6) 2.0 (1.0, 4.0) 4.4 (6.6) 2.0 (1.0, 5.0) Dairy and mixed herds were largest in size (Supplementary Fig. 1.4). Residency time distributions varied by herd type (Table 2 , Supplementary Figs. 1.5 and 1.6). In dairy and mixed herds, 42.9% and 17.1% (respectively) of residencies were of short duration (< 60 days). This was driven by the movement of calves off farms shortly after birth. Breeding herds (dairy, mixed, beef breeding) had sub-groups of cows with relatively longer residency times (23.8% dairy, 27.4% mixed and 22.9% beef breeding animals had residency times longer than 2 years, Table 2 ). Store and fattener herds had shorter residency times compared to the breeding herds (Table 2 ). Traders were defined by Brock et al. to have at least 50% of their residency times shorter than 30 days, and therefore had the shortest residency times in our population (Table 2 , Supplementary Figs. 1.5 and 1.6). Figure 1 D shows that the in-degree and residency time measures were negatively correlated (Pearsons ρ = -0.53). Supplementary Figs. 1.7 and 1.8 summarise neighbourhood burden, in degree and initiating test type by herd management system. Alongside their shorter residency times (Table 2 ) trader herds had the highest in-degree, followed by fattener herds, and, as would be expected, these also had the highest proportion of breakdowns initiated by slaughterhouse cases. Store and mixed herds had the next highest in-degree and dairy and beef breeding had the lowest in-degree. Dairy, mixed and fattener herds had higher neighbourhood burden measures compared to beef breeding, store and trader herds. Table 2 A summary of 66,620,507 animal level residency times ending between 2008 and 2023, based on a movement dataset from between 2000 and 2023, stratified by herd management practice. All herds eligible for residency time estimates are included here, not only breakdown herds. “All residency times” considers all animals, whereas “Residency times > 60 days” only animals with residency time > 60 days. We report “>60 days” summaries to exclude the large numbers of very short calf residencies in dairy herds. All residency times Residency times > 60 days Herd type Median (IQR) years Mean years % 2 years Median (IQR) years Mean years % >2 years Dairy 0.4 (0.07–1.94) 1.60 42.9% 23.8% 1.7 (0.84–3.70) 2.75 41.8% Mixed 1.3 (0.37–2.07) 1.82 17.1% 27.4% 1.6 (0.9–2.22) 2.18 33.0% Store 0.9 (0.47–1.41 ) 1.06 9.7% 8.0% 1.0 (0.59–1.46 ) 1.17 8.9% Fattener 0.8 (0.42–1.33 ) 0.98 7.1% 8.1% 0.9 (0.52–1.38 ) 1.05 8.8% Beef breeding 1.1 (0.61–1.92 ) 1.81 7.0% 22.9% 1.2 (0.69–1.99 ) 1.94 24.6% Unknown 0.5 (0.34–0.88 ) 0.88 10.1% 6.8% 0.6 (0.4–0.93 ) 0.97 7.6% Trader 0 (0.01–0.04 ) 0.07 90.5% 0.2% 0.3 (0.22–0.5 ) 0.50 2.2% All 0.8 (0.15–1.64 ) 1.39 25.7% 17.7% 1.1 (0.64–1.96 ) 1.85 23.8% Distributions of case counts Full population A : total cases The total case count distribution was right skewed with mean and standard deviation of 5.3 and 12.1, respectively. The median was 2 and the interquartile range (IQR) was 1–4 cases. The maximum case count in the dataset was 396 cases. Of all the breakdowns, 46.7% had only a single case. In 78.7% of breakdowns, all cases were disclosed in the initial test. If only standard skin reactors and slaughterhouse cases were considered, the mean and standard deviation were 3.3 and 6.1 cases. The median was 1 case, the IQR was 1–3 cases and the maximum was 228 cases. The highly skewed, fat-tailed distribution, with relatively high counts of single case breakdowns, made fitting an appropriate count distribution to the full dataset challenging. Of the three distribution fits trialled, the zeta distribution had the lowest AIC (164,466). The zero-truncated negative binomial distributions (zt-nb) and zero truncated Poisson (zt-P) distributions were associated with convergence challenges due to inflation of single case counts and had a higher AIC (167,992 for zt-nb and 475,610 for zt-P). As estimating and interpreting coefficients from a regression based on the zeta distribution can potentially present challenges ( 46 ), and convergence issues and non-random residual distributions indicated that the negative binomial and Poisson distributions were not appropriate, we had to develop another method of analysis. To this end we reframe the problem by discretising the case distribution and developing a series of logistic regression models for increasing sizes of breakdown (> 1 case, >4 cases and > 10 cases). Sub-population B : standard skin cases In “Sub-population B : standard skin cases” (excluding breakdowns initiated by a slaughterhouse case and restricted to breakdowns with at least one lesion confirmed in an ante-mortem reactor), mean and standard deviation of standard skin reactor counts were 4.7 and 7.5 respectively. The median was 2 and the IQR was 1–5. The maximum was 227 standard skin reactors. It was possible to fit all three distributions to “Sub-population B : standard skin cases” without convergence issues. The AIC of the zeta, zt-nb and zt-P fits were 77,023, 73,762 and 139,337 respectively. We selected the zt-nb based on having the lowest AIC value and no evidence of a systematic lack of fit (randomly distributed residuals). Logistic regression models (Full population A : total cases) Figure 2 summarises the odds-ratios (ORs) of the explanatory variables in the set of three logistic regressions (> 1 case, >4 cases and > 10 cases) in “Full population A : total cases”. Herd size and residency time were consistently associated with increasing total case counts. Herd management, initiating test type and year also helped explain the data. Slaughterhouse initiated breakdowns were strongly associated with single case breakdowns. The effect of year may have been influenced by changing interferon-γ testing policy and implementation over time. The neighbourhood metric had a small but statistically significant effect in explaining cases > 1, but little effect in explaining case counts > 4 or > 10. More detailed results from the model building process, alternative models, likelihood ratio testing and random effects are reported in Supplementary Results 2. Zero-truncated negative binomial regression model (Sub-population B : standard skin cases) In “Sub-population B : standard skin cases” (excluding breakdowns initiated by a slaughterhouse case and restricted to breakdowns with at least one lesion confirmed in an ante-mortem reactor), the root-mean squared error (RMSE) of an intercept only model was 7.5 units. The addition of the herd-level random effect on the intercept reduced this to 6.7 units, and the full model with fixed and random effects reduced it to 6.4 units. The model with fixed effects but excluding random effects had an RSME of 7.2 units. Likelihood ratio testing indicated that herd size and mean residency time added the most explanatory power to the model (Supplementary Results 2). Herd management type and initiating skin test type also improved model fit, but to a lesser degree. Year of breakdown end added only minor explanatory power whereas there was no justification for retention of any of the neighbourhood burden metric variables in the model. Figure 3 summarises the effect of the explanatory variables on the percentage change in mean predicted count. Detailed results for the zt-nb regression are reported in Supplementary Results 2. In our zt-nb regression, the outcome variable (mean count of standard skin reactors) was log-transformed as is routine within its generalised linear model construct. Given this log-transformation, Fig. 3 suggests a non-linear relationship between herd-size and mean residency time and count of standard reactors. Supplementary Fig. 2.6 shows a marginal effects plot from extra analyses, where we replace herd size category with log of herd size, to allow for better visual characterisation of this non-linear relationship. We also present a similar marginal effects plot for log of mean residency time in Supplementary Fig. 2.7. Discussion Main findings Our study is the first to integrate cattle residency times with a suite of other risk factors to explain variation in case counts. Our results demonstrate that, while herd size and residency times are important factors driving large case counts, metrics of neighbourhood risk are not. This suggests that, following initial introduction of infection, cattle-to-cattle transmission within herds plays a more important role in amplifying infections rather than further introduction of infection from the neighbourhood. Residency times Duration of contact between individuals is an essential component of modelling transmission of infection ( 26 ). Our finding that mean residency times in herds are important predictors of case counts fits with this transmission biology. Breeding herds have subsets of cows with relatively longer residency times compared to store, fattener or trader herds. As part of national risk-based approaches to disease control, there is much to be gained from avoiding the introduction of infection into breeding herds, given the increased risk of transmission within these herds once they become infected. As well as allowing more time for cattle-to-cattle transmission to happen, longer residency times allow more time for detection of herd level infection, if present, with the diagnostic testing system. Herds with shorter residencies, fewer breakdowns and lower case counts (for example, store herds engaged in contract rearing replacement heifers for breeding herds) may still be relevant for dispersal of infection to other herds, even if the cattle are not present in the herd of origin for long enough to be detected. Residency time could also be a proxy for duration of exposure to a common extrinsic force of infection, such as wildlife in the neighbourhood. However, metrics of neighbourhood burden added little additional explanatory value to our models, making the cattle-to-cattle transmission hypothesis more plausible. Our description of individual animal residencies, and variation in their distributions in the different herd management systems, highlights the potential for future studies to integrate these into M. bovis transmission models. Furthermore, whilst mean residency time is the best proxy of contact times we have available for the current study deign, it is not a perfect measure of contact time for all animals in a herd. That is, it does not account for the heterogeneity of cattle management (particularly the long-term management animals in smaller epidemiological units/cohorts) within each specific herd. Better characterisation of within-herd contact patterns will enhance future modelling efforts. Herd size Case count increased with herd size. Clegg et al. (2018) also reported an association between increasing herd size and large bTB breakdowns (comparing breakdowns with at least 13 reactors to those with 2–4 reactors) ( 20 ). In Northern Ireland, Wright et al. (2013) similarly reported larger outbreaks with increasing herd size ( 47 ). Herd size is also associated with the herd-level risk of having a bTB breakdown ( 19 , 48 ). Cited reasons for this include indirect association with other risk factors for introduction of infection. For example, larger herds may buy in more cattle or be dispersed over more land parcels, have higher production intensity operations correlating with physiological or nutritional stress risk factors and aggregation ( 19 , 37 , 49 – 51 ). These indirect associations could also arguably be the case for increased case counts within herds, but our study suggests that this is not the case, for the following reasons. Dairy herds are largest in size, raising the question whether herd size is associated with a subset of longer residencies which explain transmission. However, beef breeding herds have a similar proportion of longer cow residencies, but lower case counts. Larger herds may also occupy a greater geographical area and distinct land parcels and therefore have more contact with extrinsic forces of infection ( 52 ). The true area occupied by cattle in a herd is challenging to quantify as land parcels associated with a herd may be used for silage or crop production ( 49 ). Despite this challenge, the lack of effect of any of our neighbourhood burden measures on case count in our study suggests that herd size may be associated with a separate case count driver to farm footprint. The combination of large herd size and increasing emphasis on diagnostic test sensitivity over specificity, as case count increases within a breakdown, may be associated with increased false positives in larger herds. However, the strong effect of herd size remains in our subset analysis, where we exclude the possibility of false positive breakdowns and use only standard skin reactors as our response variable. Downs et al. (2016) ( 53 ) highlight that a lower proportion of dairy cattle skin reactors have lesions at slaughter compared to beef breed skin reactors, which is consistent with Irish data ( 3 ). However, Madden et al. (2025) highlight increased lesioned skin reactors in dairy and mixed management types in recent years, suggesting a genuine increase in infection levels ( 3 ). This combination of evidence suggests that false positives, or increased bTB test sensitivity in dairy cattle, do not fully explain increasing case counts with herd size. Dairy and mixed herds have increased in size over the past decade in Ireland ( 3 , 13 ). Production of milk solids per cow has also increased ( 54 ). These developments may be associated with increased cattle-to-cattle transmission. Increasing herd size may be associated with increased intensification, and extra physiological stress on cattle. For example, high yielding cattle in large and intensive dairy herds may be under increased nutritional stress, and more susceptible to infection, especially in the peri-partum period ( 55 ). They may need to compete for cubicles or other resources in recently expanded herds which have not yet updated their housing facilities. Optimum group size for expression of typical social behaviour may be as low as between three and five cattle ( 56 ), whilst behavioural studies of dairy herds suggest that peer-to-peer (“kindergarten”) bonds, developed at a young age between calves managed in the same group, shape cattle’s behaviour ( 57 , 58 ). Dairy herd size related increases in contract rearing of replacement heifers ( 13 ), may increase behavioural stress due to new introductions of cattle, loss of early-life peer-to-peer bonds and group sizes far above the optimum. In our study, although case count increases, proportion of cases relative to whole herd size drops as herd size increases (Supplementary Fig. 1.10). This non-linear increase in case count with herd size in this study may be consistent with a transmission pattern somewhere between no impact of herd size on effective contact rate ( 59 , 60 ), termed frequency dependence, and some degree of scaling of transmission with herd size, which can be termed non-linear density dependence ( 27 , 28 ). Both frequency and density dependence are mathematical generalisations of heterogenous biological processes influencing effective contact rates between animals ( 61 ). In the future, synthesis of nuanced understanding of herd management, high resolution data representing within-herd contact patterns and pathobiology related to infectiousness, with modelling, may remove much of the dichotomy between density and frequency dependent transmission of infection. Neighbourhood burden In studies of herd-level breakdown risk, metrics of neighbourhood burden are important predictors of infection being disclosed ( 8 , 19 ). In contrast to this, and in the different context of considering case counts once breakdowns are already initiated, we did not detect a substantial role for neighbourhood burden. In “Full population A : total cases”, neighbourhood burden had a weak positive effect on the probability of case counts greater than one, but less impact on larger case counts. In the “Sub-population B : standard skin cases,” its retention in the model was not justified by likelihood ratio testing. This may be consistent with the hypothesis that, after initial introduction of infection, cattle-to-cattle transmission, rather than extrinsic force of infection, is the dominant amplifier of infection and associated case counts. Contiguous herd tests are implemented at four month intervals in herds adjacent to a herd undergoing a breakdown with two or more cases ( 9 ), and therefore may indirectly reflect neighbourhood burden. Compared to annual skin testing, contiguous testing was associated with increased probability of higher case counts, but to a lesser degree than risk-based testing. However, the effect of neighbourhood burden on case counts was similar even if initiating test type was excluded from the model (Supplementary Figs. 2.4 and 2.5), further evidencing the lack of neighbourhood effect. Application of local veterinary epidemiological assessments may explain why contiguous herd tests were positively associated with case count (compared to annual skin tests). The local government veterinary inspector decides to implement contiguous testing based on risk assessment of the index breakdown and consideration of neighbourhood land usage by cattle ( 62 ). Therefore, contiguous testing may be considered a special case of risk-based testing. In the case of total case count in the full population, a further explanation could be the immediate implementation of severe skin test interpretation in contiguous tests. This cannot explain the positive influence of contiguous testing on the count of standard skin reactors in the subpopulation, but it is possible that risk-based enhanced vigilance in interpreting skin readings in contiguous herd tests, may also play a role. The minimal effect of any of our metrics of neighbourhood burden on case counts in our study contrasts with that reported by Clegg et al (2018), who reported increased odds of a case count of 13 or more (compared to 2–4) cases per logged unit of reactors per km 2 in the local electoral division ( 20 ). As well as our herd-level metric of neighbourhood burden, to allow better comparison with Clegg et al. (2018), we included animal-level metrics of neighbourhood burden as alternative explanatory variables. Despite this, none of our neighbourhood burden metrics added substantial explanatory ability to our models. The reasons for this difference with Clegg et al. (2018) are unknown, and require further investigation. We considered whether it may potentially be associated with the more extensive demographic and movement data available for the current study. We had access to the nuanced herd management categories developed by Brock et al. (2021, 2022) ( 13 , 32 ) and incorporated mean residency time. Reactors per km 2 may correlate with cattle per km 2 , which in turn is associated with the “dairy belt” in the southern area of Ireland and the largest herd sizes ( 37 ). However, even when we omit herd management and residency time related variables, the effect neighbourhood burden remains insignificant (results not shown). Clegg at al. (2018) reported that number of badgers captured over the previous 10 years within 1 km of the farm was not associated with the risk of larger breakdowns, when controlling for covariates ( 20 ). However, due to the targeted nature of badger culling around breakdown herds, such metrics are confounded by cattle bTB related badger culling effort. Milne et al. (2023) similarly reported a lack of conclusive evidence for estimates of local badger density as an explanatory variable for breakdown duration in Northern Ireland ( 23 ), despite other risk factor analyses finding positive associations with the probability of a breakdown in both Northern Ireland ( 63 ) and in Ireland ( 64 ). Our work does not directly address force of infection from badgers in the neighbourhood. However, based on a previous study which showed that bTB cases in cattle herds are sentinels for M. bovis infection in badgers ( 35 ), we believe that measures of cattle bTB in the neighbourhood are correlated with badger bTB levels. Indeed, national scale testing of badgers from 2007–2012 in Ireland found significant spatial contemporaneous and lagged associations between badger TB status and local cattle TB prevalence at scales ranging from < 250m to 1km around capture sites ( 65 ). Therefore, in the current study, lack of evidence for neighbourhood burden driving case counts within breakdowns is consistent with previous work showing lack of badger related effects on breakdown size and duration ( 20 , 23 ). Given these results, our conceptual model is one of case counts being primarily driven by within herd dynamics and less by what is happening in the wider neighbourhood via direct or indirect routes. Initiating test type In comparison to annual skin testing, risk-based testing had the greatest positive effect on case count, supporting the current policy of enhanced risk based surveillance ( 9 ). A similar result was reported from Northern Ireland, specifically with backward contact tracing related testing ( 47 ). Risk based tests in our study included forward- and backward- contact tracing check tests, tests to regain free trading status after an inconclusive slaughterhouse lesion result, miscellaneous checks or because of an association with another herd with bTB cases, for example through shared land, facilities or management. As well as increased risk of infection, these tests may be associated with increased vigilance for extra cases, due to epidemiological linkages with other bTB cases. The finding of Clegg et al. (2018) ( 20 ) that being under the same management as a herd with a bTB breakdown is a predictor of a large breakdown is consistent with our finding. This increased vigilance may also apply to contiguous testing of herds adjacent to bTB breakdowns and post-derestriction testing at 6, 12 and 18 months after the end of a breakdown. These initiating test types had a positive effect (but smaller than risk-based testing) on case counts, relative to annual skin testing. However, they had no effect on the probability the largest breakdowns with > 10 cases, consistent with what Clegg et al. (2018) reported ( 20 ). This may suggest that the more intensive post-breakdown and contiguous surveillance is detecting infection before it has the chance to be amplified to produce a large breakdown. Slaughterhouse initiated breakdowns were strongly associated with single case counts during the present study. This is consistent with the previous work in the Republic of Ireland ( 66 , 67 ) which highlighted that 19.7% and 19%, respectively, of breakdowns initiated by lesions at routine slaughter went on to have further cases, which is lower than the overall proportion of breakdowns with greater than one case (53.3% in this study). Wright et al. (2013) reported a similar finding in Northern Ireland ( 47 ). The underlying biology of this finding may be partially explained by the association between trader and fattener herds which have relatively larger proportions of movements to slaughter ( 32 ), and therefore more cattle from the herd subjected to routine PM inspection for lesions (increasing the probability of early detection), and shorter residency times (decreasing the probability of within herd transmission and time window for detection). Similarly, Clegg et al. (2018) reported that herds that introduced more cattle had smaller bTB breakdowns ( 20 ), consistent with the negative association we found between the highest three in degree quintiles and case count. This is likely to be due to the positive correlation between rates of inward and outward movements ( 37 ) and consistent with the negative correlation we found between residency time and in degree (which in turn is also correlated with inward movements). Slaughterhouse initiated breakdowns, high rates of inward and outward movements, high in degree and low case counts are all associated with fattener and trader herds. Store herds also have high in-degree and shorter residency times, but fewer movements to slaughter. This may also explain the smaller case counts in store herds, in the absence of a high proportion of slaughterhouse cases. To summarise, multiple factors that are associated with shorter residency times are also associated with lower case counts. Two previous Irish studies, ( 66 , 67 ), reported that, when slaughterhouse lesions were detected in breeding herds (dairy or beef breeding), there were larger counts of skin test reactors subsequently, compared to non-breeding herds, which is consistent with the hypothesis above. Our description of case counts stratified by both initiating test type and breeding herd status (Supplementary Fig. 1.12) similarly shows higher mean case count in breeding compared to non-breeding herds. Herd type Our multivariable analyses suggest that dairy herds have larger breakdowns compared to beef breeding herds, even after herd size and residency time are accounted for. Dairy management is recognised as a risk factor for breakdown probability in many studies ( 8 , 18 , 19 , 68 ). As we discussed in the context of increased herd size, potential mechanisms for the positive effect of dairy management on case count may include physiological pressures (parturition, energy balance, nutrition), genetic susceptibility or management practices, and this is an area for future research. In “Sub-population B : standard skin cases”, when slaughterhouse initiated breakdowns were excluded, fattener herd type was associated with larger breakdowns. This was also the case for breakdowns with > 10 cases in “Full population A : total cases”, but not for the smaller case count categories. Controlled-finishing-units (CFUs, fattener style herds which cannot sell to other herds, are required to have increased biosecurity requirements, a subject to reduced testing requirements and often have prolonged breakdown status) were excluded from our study population, precluding CFU status as an explainer of larger case counts in fattener herds. Therefore, this subset of fattener herds with larger breakdowns warrants further investigation. In Northern Ireland, such herds were associated with small case counts (measured as the number of reactors), but tended to have higher strain diversity, indicative of multiple pathway introductions, for example via trade ( 69 , 70 ). It is notable that, in Northern Ireland, long-term genetic typing of infection suggested that dairy herd breakdowns were dominated by single genotype breakdowns, indicative of within-herd local spread of infection or ‘microepidemics’, when compared to smaller sized beef fattening enterprises, where higher strain richness occurred, linked to more buying-in of animals ( 70 ). We found that a subset of fattener herds have large breakdowns. Synthesising our finding of a subset of fattener herds with larger breakdowns with the Northern Irish findings of increased genetic diversity of M. bovis in fattener herds ( 69 , 70 ), highlights the risk that fattener herds may be a source of new M. bovis variants into the neighbourhood, even though they are sending cattle to slaughter rather than to other herds. It is unknown, but important to find out, whether the cattle bought in to fatten, and sent relatively quickly to slaughter, are introducing new M. bovis transmission chains which become established in the neighbourhood. Therefore, investigating the proximity to such fattener herds as a risk factor for bTB breakdowns may be an important area for future research in the spatial epidemiology of M. bovis . Year Our data descriptions, in line with the in-depth temporal analyses of ( 3 ), show that the mean and variation of case counts in dairy and mixed herd management types have increased over recent years. However, once evolving interferon-γ testing policy and implementation is set aside (by considering only standard skin reactors in “Sub-population B : standard skin cases”), in our multivariable regression models, year does not add much explanatory ability to our multivariable regression models. Our and Madden et al. (2025) ’s work ( 3 ) demonstrates that higher case counts represent a true increase in infection levels, and are not only due to increased surveillance effort, as there is an increase in lesioned skin reactors, slaughterhouse cases and standard skin test reactors. Given the minimal effect of year, after excluding interferon-γ testing, our study suggests that changes in herd size and management may account for much of the year related effect through increased cattle-to-cattle transmission of infection. Modelling challenges, limitations and future work Our challenges in fitting statistical distributions to our full dataset are as expected, given the multiple heterogenous processes contributing to the distribution of case counts. We addressed this issue with relatively straightforward analytic approaches. In the full population, we conducted a series of logistic regressions. Separately, based on the hypotheses about the underlying processes, we defined a sub-population which minimised bias due to heterogeny in surveillance effort. An alternative approach in future work may be to capitalise on flexible Bayesian mixture models to explicitly incorporate heterogenous processes. The analyses presented in this paper provides evidence that residency times and herd size are important predictors of case count, and that neighbourhood burden plays a lesser role. Future work will involve mechanistic modelling to better understand these processes and model interventions to reduce transmission. Our study focussed on describing potential drivers of case counts in bTB breakdowns. However, if solely statistical prediction was required, methods such as principal component analyses, elastic net regression or other machine learning approaches may lend themselves to analysis of multiple options for explanatory variables, many with collinearities. Conclusion Our analyses suggest that herd size and residency time are the most important drivers of case count in Irish bTB breakdowns. This supports cattle-to-cattle transmission as an amplifier of infection within herds. After initial introduction of infection, within herd transmission seems to be the main driver of case counts. Protecting large breeding herds from introduction of infection may be particularly beneficial in preventing large scale amplification of M. bovis in such herds. Our study adds to the evidence base on potential drivers of Irish bTB burden and will inform future mechanistic model development to explicitly model the underlying processes. Declarations Acknowledgements We would like to thank Professor Eamonn Gormley of University College Dublin, and Drs Nicola Harvey, Michael Horan and Philip Breslin of the Department of Agriculture, Food and the Marine, for helpful discussions relating to this study. Funding This research received no specific funding. JM, JT (and up to September 2024 MC and SM) work in the Centre for Veterinary Epidemiology and Risk Analysis in University College Dublin which is funded by the Department of Agriculture, Food and the Marine (DAFM). MC is currently funded by a University College Dublin Ad Astra Fellowship. Competing interests None declared. Data availability The datasets analysed during this study are available from Department of Agriculture, Food and the Marine (DAFM), but are subject to data protection regulations and limitations: https://www.gov.ie/en/department-of-agriculture-food-and-the-marine/organisation-information/data-protection/ Ethics declaration The datasets analysed during this study are from pre-existing databases of the Department of Agriculture, Food and the Marine (DAFM), and were analysed in compliance with the General Data Protection Regulation of the European Union. No experimental procedures, animal handling, or interventions were performed. In accordance with the policies of BMC Veterinary Research and institutional guidelines, ethical approval was not required because the study used only secondary data that were collected previously for purposes unrelated to this research. Consent to Publish declaration Consent to Publish declaration: not applicable. This study uses only pre-existing, anonymized data and involves no animal or human subjects. Author contributions Conceptualisation: MC, AC, JM; Methodology : MC, AC, JM, JT ; Formal analysis: MC, AC; Investigation: MC, AC, JM; Data curation: MC, JM, JC, Resources: SM, DB, Writing (original draft): MC, Writing (review): All authors; Visualisation: MC; Validation: MC, JM, AC, AB, SM. AC and JM contributed equally to this study. References More SJ, Good M. The tuberculosis eradication programme in Ireland: A review of scientific and policy advances since 1988. Vet Microbiol. 2006;112(2-4 SPEC. ISS.):239–51. Ryan E, Breslin P, O’Keeffe J, Byrne AW, Wrigley K, Barrett D. The Irish bTB eradication programme: combining stakeholder engagement and research-driven policy to tackle bovine tuberculosis. Ir Vet J [Internet]. 2023;76(1):1–13. Available from: https://doi.org/10.1186/s13620-023-00255-8 Madden J, Gormley E, McGrath G, McAloon C, Casey-Bryars M. A new online tool for near real-time exploration of cattle and bovine tuberculosis trends in Ireland [Internet]. 2025. Available from: https://j-madden-m.github.io/Irish_bTB_trends/ DAFM. Bovine TB Action Plan [Internet]. 2025. Available from: https://assets.gov.ie/static/documents/d4cfc18d/7784-DAFM_TB_Action_Plan_LR.pdf DAFM. National bovine tuberculosis statistics. 2025. Dáil Éireann. Dáil Éireann Debate, Wednesday - 14 May 2025 [Internet]. 2025. Available from: https://www.oireachtas.ie/en/debates/question/2025-05-14/160/ More SJ. Eradication of bovine tuberculosis in Ireland: is it a case of now or never? Ir Vet J [Internet]. 2024;77(1):4–9. Available from: https://doi.org/10.1186/s13620-024-00282-z Tratalos JA, Fielding HR, Madden JM, Casey M, More SJ. Can Ingoing Contact Chains and other cattle movement network metrics help predict herd-level bovine tuberculosis in Irish cattle herds? Prev Vet Med [Internet]. 2023;211(April 2022):105816. Available from: https://doi.org/10.1016/j.prevetmed.2022.105816 More SJ, Good M. Understanding and managing bTB risk: Perspectives from Ireland. Vet Microbiol. 2015;176(3–4):209–18. Allen AR, Skuce RA, Byrne AW. Bovine Tuberculosis in Britain and Ireland – A Perfect Storm? the Confluence of Potential Ecological and Epidemiological Impediments to Controlling a Chronic Infectious Disease. Front Vet Sci [Internet]. 2018 Jun 5 [cited 2021 Oct 26];5:109. Available from: https://www.frontiersin.org/article/10.3389/fvets.2018.00109/full Griffin JM, Williams DH, Kelly GE, Clegg TA, O’Boyle I, Collins JD, et al. The impact of badger removal on the control of tuberculosis in cattle herds in Ireland. Prev Vet Med [Internet]. 2005 Mar 1 [cited 2021 Jun 24];67(4):237–66. Available from: https://www.sciencedirect.com/science/article/pii/S0167587704002090?via%3Dihub Chang Y, Hartemink N, Byrne AW, Gormley E, McGrath G, Tratalos JA, et al. Inferring bovine tuberculosis transmission between cattle and badgers via the environment and risk mapping. Front Vet Sci. 2023;10. Brock J, Lange M, Tratalos JA, Meunier N, Guelbenzu-Gonzalo M, More SJ, et al. The Irish cattle population structured by enterprise type: overview, trade & trends. Irish Vet J 2022 751 [Internet]. 2022;75(1):1–11. Available from: https://irishvetjournal.biomedcentral.com/articles/10.1186/s13620-022-00212-x General O of the C and A. Bovine TB eradication programme [Internet]. 2025. Available from: https://www.audit.gov.ie/en/find-report/publications/2025/14-bovine-tb-eradication-programme.pdf Clegg TA, Doyle M, Ryan E, More SJ, Gormley E. Characteristics of Mycobacterium bovis infected herds tested with the interferon-gamma assay. Prev Vet Med [Internet]. 2019;168(April):52–9. Available from: https://doi.org/10.1016/j.prevetmed.2019.04.004 More SJ, Houtsma E, Doyle L, McGrath G, Clegg TA, De La Rua-Domenech R, et al. Further description of bovine tuberculosis trends in the United Kingdom and the Republic of Ireland, 2003-2015. Vet Rec. 2018;183(23):717. Clegg TA, Duignan A, More SJ. The relative effectiveness of testers during field surveillance for bovine tuberculosis in unrestricted low-risk herds in Ireland. Prev Vet Med [Internet]. 2015 Apr 1 [cited 2019 Aug 12];119(1–2):85–9. Available from: https://www.sciencedirect.com/science/article/pii/S0167587715000598?via%3Dihub Skuce RA, Allen AR, McDowell SWJ. Herd-level risk factors for bovine tuberculosis: a literature review. Vet Med Int [Internet]. 2012 [cited 2021 Jun 17];2012:621210. Available from: http://www.ncbi.nlm.nih.gov/pubmed/22966479 Broughan JM, Judge J, Ely E, Delahay RJ, Wilson G, Clifton-Hadley RS, et al. A review of risk factors for bovine tuberculosis infection in cattle in the UK and Ireland. Epidemiol Infect. 2016;144(14):2899–926. Clegg TA, Good M, Hayes M, Duignan A, McGrath G, More SJ. Trends and predictors of large tuberculosis episodes in cattle herds in Ireland. Front Vet Sci. 2018;5(MAY):1–12. Karolemeas K, McKinley TJ, Clifton-Hadley RS, Goodchild AV, Mitchell A, Johnston WT, et al. Predicting prolonged bovine tuberculosis breakdowns in Great Britain as an aid to control. Prev Vet Med [Internet]. 2010 Dec 1 [cited 2021 Sep 8];97(3–4):183–90. Available from: https://www.sciencedirect.com/science/article/pii/S0167587710002564?via%3Dihub Byrne AW, Barrett D, Breslin P, Madden JM, O’Keeffe J, Ryan E. Bovine Tuberculosis (Mycobacterium bovis) Outbreak Duration in Cattle Herds in Ireland: A Retrospective Observational Study. Pathogens [Internet]. 2020 Oct 5 [cited 2021 Sep 27];9(10):815. Available from: https://www.mdpi.com/2076-0817/9/10/815 Milne G, Allen A, Graham J, Lahuerta-Marin A, McCormick C, Presho E, et al. Bovine tuberculosis breakdown duration in cattle herds: An investigation of herd, host, pathogen and wildlife risk factors. PeerJ [Internet]. 2020 [cited 2021 Sep 27];2020(2):e8319. Available from: http://www.ncbi.nlm.nih.gov/pubmed/32117602 Madenci D, Sánchez-Molano E, Madenci D, Tsairidou S, Winters M, Mitchell AP, et al. 700. Detection of genetic variability in cattle infectivity for bovine tuberculosis (bTB). J Dairy Sci [Internet]. 2025;2887–90. Available from: http://dx.doi.org/10.3168/jds.2024-25697 Brooks-Pollock E, Keeling M. Herd size and bovine tuberculosis persistence in cattle farms in Great Britain. Prev Vet Med. 2009;92(4):360–5. Vynnycky E, White R. An introduction to infectious disease modelling. Oxford: Oxford University Press; 2010. Conlan AJK, McKinley TJ, Karolemeas K, Pollock EB, Goodchild A V., Mitchell AP, et al. Estimating the Hidden Burden of Bovine Tuberculosis in Great Britain. Ferguson N, editor. PLoS Comput Biol [Internet]. 2012 Oct 18 [cited 2021 Apr 1];8(10):e1002730. Available from: https://dx.plos.org/10.1371/journal.pcbi.1002730 Conlan AJK, Brooks Pollock E, McKinley TJ, Mitchell AP, Jones GJ, Vordermeier M, et al. Potential Benefits of Cattle Vaccination as a Supplementary Control for Bovine Tuberculosis. PLoS Comput Biol. 2015;11(2):1–26. Conlan AJK, Vordermeier M, de Jong MCM, Wood JLN. The intractable challenge of evaluating cattle vaccination as a control for bovine tuberculosis. Elife. 2018;7:1–39. VanderWaal K, Enns EA, Picasso C, Alvarez J, Perez A, Fernandez F, et al. Optimal surveillance strategies for bovine tuberculosis in a low-prevalence country. Sci Rep [Internet]. 2017 Dec 23 [cited 2021 Oct 21];7(1):4140. Available from: http://www.nature.com/articles/s41598-017-04466-2 Madden JM, O’Donovan J, Casey-Bryars M, Sweeney J, Messam LL, McAloon CG, et al. The impact of changing the cut-off threshold of the interferon-gamma (IFN-γ) assay for diagnosing bovine tuberculosis in Ireland. Prev Vet Med. 2024;224(January). Brock J, Lange M, Tratalos JA, More SJ, Graham DA, Guelbenzu-Gonzalo M, et al. Combining expert knowledge and machine-learning to classify herd types in livestock systems. Sci Rep [Internet]. 2021 Dec 4 [cited 2021 Feb 8];11(1):2989. Available from: http://www.nature.com/articles/s41598-021-82373-3 Griffin J, Aznar I, Breslin P, Good M, Gordon S, Gormley E, et al. What is the scope for existing (including recently developed) diagnostic methods to detect infected cattle which are not currently detected by the existing programme? Food Risk Assess Eur. 2024;1(2). de la Rua-Domenech R, Goodchild AT, Vordermeier HM, Hewinson RG, Christiansen KH, Clifton-Hadley RS. Ante mortem diagnosis of tuberculosis in cattle: A review of the tuberculin tests, γ-interferon assay and other ancillary diagnostic techniques. Res Vet Sci. 2006;81(2):190–210. Murphy D, Gormley E, Collins DM, McGrath G, Sovsic E, Costello E, et al. Tuberculosis in cattle herds are sentinels for Mycobacterium bovis infection in European badgers (Meles meles): The Irish Greenfield Study. Vet Microbiol [Internet]. 2011;151(1–2):120–5. Available from: http://dx.doi.org/10.1016/j.vetmic.2011.02.034 Goodchild A V., Downs SH, Upton P, Wood JLN, De La Rua-Domenech R. Specificity of the comparative skin test for bovine tuberculosis in Great Britain. Vet Rec. 2015;177(10):258. Tratalos J, Madden J, McGrath G, Graham D, Áine Collins, More S. Spatial and network characteristics of Irish cattle movements. Prev Vet Med. 2020;183. Zimmermann J, Fealy RM, Lydon K, Mockler EM, O’Brien P, Packham I, et al. The Irish land-parcels identification system (LPIS) – experiences in ongoing and recent environmental research and land cover mapping. Biol Environ. 2016;116B(1):53–62. Houtsma E, Clegg TA, Good M, More SJ. Further improvement in the control of bovine tuberculosis recurrence in Ireland. Vet Rec. 2018;183(20):622. Madden J, McGrath G, Sweeney J, Murray G, Tratalos JA, More S. Spatio-temporal models of bovine tuberculosis in the Irish cattle population , 2012-2019. Spat Spatiotemporal Epidemiol. 2021;39. Houtsma E, Clegg TA, Good M, More SJ. Further improvement in the control of bovine tuberculosis recurrence in Ireland. Vet Rec. 2018;183(20):622. Bursac Z, Gauss CH, Williams DK, Hosmer DW. Purposeful selection of variables in logistic regression. Source Code Biol Med. 2008;3:1–8. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing [Internet]. Vienna, Austria: R Foundation for Statistical Computing; 2024. Available from: http://www.r-project.org. Yee TW. Vector Generalized Linear and Additive Models: With an Implementation in R. Vector Generalized Linear and Additive Models: With an Implementation in R. 2015. 1–589 p. Brooks ME, Kristensen K, van Benthem KJ, Magnusson A, Berg CW, Nielsen A, et al. Package “glmmTMB” Title Generalized Linear Mixed Models using Template Model Builder. 2022;9(December):378–400. Available from: https://orcid.org/0000-0001-5715-616X Doray LG, Arsenault M. Estimators of the regression parameters of the zeta distribution. 2002;30:439–50. Wright DM, Allen AR, Mallon TR, McDowell SWJ, Bishop SC, Glass EJ, et al. Field-Isolated Genotypes of Mycobacterium bovis Vary in Virulence and Influence Case Pathology but Do Not Affect Outbreak Size. PLoS One. 2013;8(9). Skuce RA, Allen AR, McDowell SWJ. Herd-Level Risk Factors for Bovine Tuberculosis: A Literature Review. Vet Med Int [Internet]. 2012 [cited 2021 Jun 17];2012:1–10. Available from: http://www.hindawi.com/journals/vmi/2012/621210/ McGrath G, More S. Farm fragmentation in Ireland. Vet Ital. 2024;60(4). Humblet MF, Gilbert M, Govaerts M, Fauville-Dufaux M, Walravens K, Saegerman C. New assessment of bovine tuberculosis risk factors in Belgium based on nationwide molecular epidemiology. J Clin Microbiol. 2010;48(8):2802–8. Downs SH, Durr P, Edwards J, Clifton-Hadley R. Trace micro-nutrients may affect susceptibility to bovine tuberculosis in cattle. Prev Vet Med. 2008;87(3–4):311–26. Milne G, Graham J, McGrath J, Kirke R, McMaster W, Byrne AW. Investigating Farm Fragmentation as a Risk Factor for Bovine Tuberculosis in Cattle Herds: A Matched Case-Control Study from Northern Ireland. Pathogens. 2022;11(3). Downs SH, Broughan JM, Goodchild A V., Upton PA, Durr PA. Responses to diagnostic tests for bovine tuberculosis in dairy and non-dairy cattle naturally exposed to Mycobacterium bovis in Great Britain. Vet J [Internet]. 2016;216:8–17. Available from: http://dx.doi.org/10.1016/j.tvjl.2016.06.010 Kelly P, Shalloo L, Wallace M, Dillon P. The Irish dairy industry – Recent history and strategy, current state and future challenges. Int J Dairy Technol. 2020;73(2):309–23. Trevisi E, Amadori M, Cogrossi S, Razzuoli E, Bertoni G. Metabolic stress and inflammatory response in high-yielding, periparturient dairy cows. Res Vet Sci [Internet]. 2012;93(2):695–704. Available from: http://dx.doi.org/10.1016/j.rvsc.2011.11.008 Takeda K ichi, Sato S, Sugawara K. The number of farm mates influences social and maintenance behaviours of Japanese Black cows in a communal pasture. Appl Anim Behav Sci. 2000;67(3):181–92. Marina H, Ren K, Hansson I, Fikse F, Nielsen PP, Rönnegård L. New insight into social relationships in dairy cows and how time of birth, parity, and relatedness affect spatial interactions later in life. J Dairy Sci. 2024;107(2):1110–23. Marina H, Nielsen PP, Fikse WF, Rönnegård L. Multiple factors shape social contacts in dairy cows. Appl Anim Behav Sci. 2024;278(July). de Jong MCM, Diekmann O, Heesterbeek H. How does transmission of infection depend on population size. Vol. 94, Journal of the American Statistical Association. 1995. p. 84–94. Fromsa A, Willgert K, Srinivasan S, Mekonnen G, Bedada W, Gebre S, et al. BCG vaccination reduces bovine tuberculosis transmission, improving prospects for elimination. 2024;3962. Begon M, Bennett M, Bowers RG, French NP, Hazel SM, Turner J. A clarification of transmission terms in host-microparasite models: Numbers, densities and areas. Epidemiol Infect. 2002;129(1):147–53. Good M, Duignan A. Veterinary Handbook for Herd Management in the TB eradication programme [Internet]. Dublin; 2016. 23–25 p. Available from: https://www.researchgate.net/publication/323402319_Veterinary_Handbook_for_herd_management_in_the_bovine_TB_Eradication_Programme Wright DM, Reid N, Montgomery WI, Allen AR, Skuce RA, Kao RR. Herd-level bovine tuberculosis risk factors: Assessing the role of low-level badger population disturbance. Sci Rep. 2015;5(November 2014):1–11. Byrne AW, White PW, Mcgrath G, O’keeffe J, Martin W. Risk of tuberculosis cattle herd breakdowns in Ireland: effects of badger culling effort, density and historic large-scale interventions [Internet]. 2014 [cited 2021 Jun 25]. Available from: http://www.veterinaryresearch.org/content/45/1/109 Byrne AW, Kenny K, Fogarty U, O’Keeffe JJ, More SJ, McGrath G, et al. Spatial and temporal analyses of metrics of tuberculosis infection in badgers (Meles meles) from the Republic of Ireland: Trends in apparent prevalence. Prev Vet Med [Internet]. 2015;122(3):345–54. Available from: http://dx.doi.org/10.1016/j.prevetmed.2015.10.013 Byrne AW, Barrett D, Breslin P, Madden JM, O’Keeffe J, Ryan E. Post-mortem surveillance of bovine tuberculosis in Ireland: herd-level variation in the probability of herds disclosed with lesions at routine slaughter to have skin test reactors at follow-up test. Vet Res Commun. 2020;44(3–4):131–6. Olea-Popelka FJ, Costello E, White P, McGrath G, Collins JD, O’Keeffe J, et al. Risk factors for disclosure of additional tuberculous cattle in attested-clear herds that had one animal with a confirmed lesion of tuberculosis at slaughter during 2003 in Ireland. Prev Vet Med. 2008;85(1–2):81–91. Humblet MF, Boschiroli ML, Saegerman C. Classification of worldwide bovine tuberculosis risk factors in cattle: A stratified approach. Vet Res. 2009;40(5). Milne MG, Graham J, Allen A, Lahuerta-Marin A, McCormick C, Breadon E, et al. Herd characteristics, wildlife risk and bacterial strain genotypes in persistent breakdowns of bovine tuberculosis in Northern Irish cattle herds. In: M.L. Brennan & A. Lin, editor. Proceedings of the Society for Veterinary Epidemiology and Preventive Medicine Annual Meeting [Internet]. Talin, Estonia; 2018. p. 56–67. Available from: https://svepm.org.uk/wp-content/uploads/2023/05/phpcK8JOc20180313220149.pdf#page=56 Milne MG, Graham J, Allen A, McCormick C, Presho E, Skuce R, et al. Variation in Mycobacterium bovis genetic richness suggests that inwards cattle movements are a more important source of infection in beef herds than in dairy herds. BMC Microbiol. 2019;19(1):1–13. Additional Declarations No competing interests reported. Supplementary Files Supplementarymethods.docx Supplementaryresultspart1descriptiveplots.docx Supplementaryresultspart2alternativeanalysesandextraresults.docx Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 08 Apr, 2026 Reviews received at journal 05 Apr, 2026 Reviews received at journal 29 Mar, 2026 Reviewers agreed at journal 15 Mar, 2026 Reviewers agreed at journal 23 Feb, 2026 Reviewers invited by journal 28 Nov, 2025 Editor invited by journal 28 Nov, 2025 Editor assigned by journal 27 Nov, 2025 Submission checks completed at journal 27 Nov, 2025 First submitted to journal 22 Nov, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8181407","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":552252640,"identity":"796f08ce-e597-4dde-b88c-84cb1f6e3be8","order_by":0,"name":"Miriam Casey-Bryars","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1klEQVRIiWNgGAWjYDACZgaGA3DOBwYGxgZmUrQwziBKC4p2HpAWQqp023kfHviZwyDP33/4mLRtm51sAzvzAbxazA6zGxzs3cZgOONGWpp0bluycQMzWwIBLWwMB3i3MSQw3OAxA2o5kNjAzGNAUMvBv0At8ufPf5O2BGvh/0BQy2GQLQYHctikGSG24NUB0SK7TcJw4400Y8uec8nGbcxsBBx2/hjzx7fbbOTlzh9+eONHmZ1sP//hB/itgQAJEMECJtmIUQ8DzPi9PQpGwSgYBSMWAACRqkIxkPJZbAAAAABJRU5ErkJggg==","orcid":"","institution":"University College Dublin","correspondingAuthor":true,"prefix":"","firstName":"Miriam","middleName":"","lastName":"Casey-Bryars","suffix":""},{"id":552252641,"identity":"e6bb4b8e-4b7f-4d01-89cb-6171355808ca","order_by":1,"name":"Jamie A. Tratalos","email":"","orcid":"","institution":"University College Dublin","correspondingAuthor":false,"prefix":"","firstName":"Jamie","middleName":"A.","lastName":"Tratalos","suffix":""},{"id":552252642,"identity":"1014cdf6-33ed-4632-bb1b-30a0342eddab","order_by":2,"name":"Andrew Byrne","email":"","orcid":"","institution":"Department of Agriculture Food and the Marine","correspondingAuthor":false,"prefix":"","firstName":"Andrew","middleName":"","lastName":"Byrne","suffix":""},{"id":552252643,"identity":"b0ea17b1-27dd-46bd-a413-3aad0fa959ff","order_by":3,"name":"Simon More","email":"","orcid":"","institution":"University College Dublin","correspondingAuthor":false,"prefix":"","firstName":"Simon","middleName":"","lastName":"More","suffix":""},{"id":552252644,"identity":"feb69b16-3efd-4e86-b717-af0a8a4de560","order_by":4,"name":"Damien Barrett","email":"","orcid":"","institution":"Department of Agriculture Food and the Marine","correspondingAuthor":false,"prefix":"","firstName":"Damien","middleName":"","lastName":"Barrett","suffix":""},{"id":552252645,"identity":"9e659518-8f23-43fd-8b57-3a3a65d3911b","order_by":5,"name":"Jamie M. Madden","email":"","orcid":"","institution":"University College Dublin","correspondingAuthor":false,"prefix":"","firstName":"Jamie","middleName":"M.","lastName":"Madden","suffix":""},{"id":552252646,"identity":"b6a6b552-9d4c-40e5-8403-b96fedaeb0f6","order_by":6,"name":"Andrew Conlan","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Andrew","middleName":"","lastName":"Conlan","suffix":""}],"badges":[],"createdAt":"2025-11-22 15:38:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8181407/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8181407/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97266350,"identity":"fc6850f9-96b7-4727-9bbd-59303c33664d","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":967632,"visible":true,"origin":"","legend":"","description":"","filename":"ResidencytimesandotherriskfactorsforlargebTBbreakdownsmaintext.docx","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/b250fb3b5d9c30c83a0e5853.docx"},{"id":97266347,"identity":"f6a3610d-6ee0-4a0b-afc6-2d9e57bbe975","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9603,"visible":true,"origin":"","legend":"","description":"","filename":"512225951fce451c916dba9d96f19359.json","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/fff45c1e70969f724716a107.json"},{"id":97266353,"identity":"be20f78b-9a15-45a3-9034-c3a6d77546a9","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":217250,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymethods.docx","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/ec399bb76b3ac8d450a1ed5f.docx"},{"id":97266360,"identity":"f4a021df-f68a-4069-af8c-5bf7b77ac856","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1642552,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryresultspart1descriptiveplots.docx","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/b858bb774c3ebf83255d564f.docx"},{"id":97367272,"identity":"fd2040c6-2fac-4d7d-93f5-61ba491d56c3","added_by":"auto","created_at":"2025-12-03 16:17:55","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2123009,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryresultspart2alternativeanalysesandextraresults.docx","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/2237c8e05b2aae03f6733612.docx"},{"id":97367982,"identity":"bc28352f-3304-4d36-9e70-1a936319b7fe","added_by":"auto","created_at":"2025-12-03 16:21:11","extension":"xml","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":192445,"visible":true,"origin":"","legend":"","description":"","filename":"512225951fce451c916dba9d96f193591enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/cacbd6a94c610235ea63aba1.xml"},{"id":97367858,"identity":"0cfa3bd7-4fe5-4ab6-abe4-c0b966ce06e5","added_by":"auto","created_at":"2025-12-03 16:20:55","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":227336,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/685da0570e6656b2af8e40ff.jpeg"},{"id":97266364,"identity":"04413bd8-d306-45a5-b0c1-f6f89bed9280","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":448679,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/dabe2c2ba029706bc8df8782.jpeg"},{"id":97368196,"identity":"b7a4d18b-a8c6-4bd9-8cec-da16423438e2","added_by":"auto","created_at":"2025-12-03 16:21:46","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":386489,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/49315adce98bc66f756aaeb9.jpeg"},{"id":97266355,"identity":"3ab9b272-3f9e-48ae-8278-8b08f054cafb","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53078,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/8d1a58742b0b7e406d1f1ace.png"},{"id":97368506,"identity":"fc4c2302-428f-40e5-ac7b-787d7a77d001","added_by":"auto","created_at":"2025-12-03 16:22:22","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":92016,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/9cd0bf5d87944dc41b5b58ba.png"},{"id":97266365,"identity":"45870635-0d06-4d34-a449-d4b4831e65aa","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":78752,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/93ce2164e6cba13fc8364ff8.png"},{"id":97266366,"identity":"8dd330eb-9a23-4f87-8d9c-8d6293c9d9b3","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"xml","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":192870,"visible":true,"origin":"","legend":"","description":"","filename":"512225951fce451c916dba9d96f193591structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/b9438cf966ef1adc26db8154.xml"},{"id":97266367,"identity":"6f0bc5d0-3076-441a-be0b-e9b802899d3a","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"html","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":203680,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/570b7c726b918b68589d879e.html"},{"id":97266349,"identity":"832819a6-a938-4f2a-a68c-267e2321100c","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":244363,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePanels A – C show case count stratified by herd size (A), mean herd residency time (B) and initiating test type (C). Triangles represent the mean and boxplots represent median and interquartile range. Panel D shows a negative association between residency time and in degree*. This plot is based on the full population of breakdowns (Full population A : total cases, N = 37,176 breakdowns ending between 2012 and 2023, in 24,730 herds). \u0026nbsp;*A standardised and centred square root transform was used to stabilise variance for in degree and residency time.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/28a91a4c4ccec91fba5584a1.png"},{"id":97266352,"identity":"cadc4a55-52f6-4bf6-b7bf-44a02d6e9d8a","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":505635,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eResults from multivariable logistic regression, with a herd level random effect on the intercept, to explain (left) total case counts greater than one in all breakdowns (n = 37,176); (centre) amongst breakdowns with greater than one case (n = 20,029) , total case counts greater than four; (right) amongst breakdowns with greater than four cases (n= 9,104), total case counts greater than ten. \u0026nbsp;These outputs are based on Full population A : total cases, breakdowns ending between 2012 and 2023.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/4b9fd3b3ce40aef34bd8254f.png"},{"id":97368480,"identity":"1a62544d-d06f-4bdd-8895-9ee2166de891","added_by":"auto","created_at":"2025-12-03 16:22:21","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":414318,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEstimated risk factors from the multivariable zero truncated negative binomial regression model to explain count of standard skin reactors. These outputs are based on Sub-population B : standard skin cases, in breakdowns ending between 2012 and 2023, excluding breakdowns initiated by a slaughterhouse case and restricted to breakdowns with at least one lesion confirmed in an ante-mortem reactor.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/470f02c7fb0584f3a44dc5cc.png"},{"id":98621927,"identity":"033ce0fd-0270-49c4-ab28-a9c82181e7a1","added_by":"auto","created_at":"2025-12-19 16:34:41","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2618461,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/a335da47-de3a-41a7-b249-db8ceacce57a.pdf"},{"id":97266348,"identity":"bb9189f7-850d-499b-9fc1-b0264b47f657","added_by":"auto","created_at":"2025-12-02 14:34:00","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":217250,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymethods.docx","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/40b7d350e530dc0fe26c5be9.docx"},{"id":97367151,"identity":"8ad7c928-1e4e-47d9-806c-b0dc30fe975f","added_by":"auto","created_at":"2025-12-03 16:17:10","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":1642552,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryresultspart1descriptiveplots.docx","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/df75dedbceaad2f6540fdf6a.docx"},{"id":97367175,"identity":"2514e4f9-5af5-4c6d-99f3-8a690705b09e","added_by":"auto","created_at":"2025-12-03 16:17:18","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":2123009,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryresultspart2alternativeanalysesandextraresults.docx","url":"https://assets-eu.researchsquare.com/files/rs-8181407/v1/192379b4dc2c6ee6147c5de1.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Residency time and other risk factors for large bovine tuberculosis breakdowns in Irish cattle herds","fulltext":[{"header":"Introduction","content":"\u003cp\u003e\u003cem\u003eMycobacterium bovis\u003c/em\u003e (\u003cem\u003eM. bovis\u003c/em\u003e) is a zoonotic bacterium which causes tuberculosis in cattle and other animals. Due to its impact on human and bovine health, and to maintain access to export markets for cattle products, there has been a national eradication programme in place in Ireland since the 1950’s (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). After a large initial reduction in \u003cem\u003eM. bovis\u003c/em\u003e burden in cattle during the earlier decades of the programme, progress plateaued. Between 2016 and 2024, the prevalence of bovine tuberculosis (bTB, specifically referring to tuberculosis in cattle) has increased in Ireland (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). This has caused challenges for the sustainability of Irish farming and high costs to tax-payers and farmers. In 2024, the programme cost the Irish exchequer €100.6\u0026nbsp;million, with €54.4\u0026nbsp;million allocated to compensating farmers for bTB-related cattle culling (\u003cspan additionalcitationids=\"CR5\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e–\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). Ireland’s lowest recorded annual herd-level incidence occurred in 2016, with 3.30% of herds reporting new breakdowns. However, incidence has since increased, reaching 6.04% in 2024 (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eBecause of its complex multispecies epidemiology, including evidence for circulation of infection between cattle and wildlife (\u003cspan additionalcitationids=\"CR9 CR10 CR11\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e–\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e), there is robust debate about what is driving the recent increase in prevalence and the potential effectiveness of interventions. Cattle herd management systems have evolved over the past decade and mixed and dairy herds have grown larger (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e). There have also been increased numbers of fattener herds, dairy herds which purchase their replacement heifers from separate enterprises, and the associated heifer rearing store herds (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e). Although there has been a badger BCG vaccination campaign in place in recent years, there has been a concurrent increase in the number of badgers culled. Between 2014 and 2017, 24,121 badgers were culled (~ 6,030 per annum) whereas between 2020 and 2023, 21,393 were culled (~ 7,131 per annum) (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e). Extensive efforts and policy changes have been implemented to enhance detection of infected cattle, including training and auditing to improve quality of skin testing by private veterinary practitioners (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan additionalcitationids=\"CR16\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e–\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e). However, as of mid 2025, no major changes have been applied in terms of cattle movement controls (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAnswering the question of why, despite extensive and expensive controls, the prevalence of bTB is increasing, is critical for the Irish livestock industry. As well as herd-level breakdown occurrence, breakdown case counts (the total number of bTB positive cattle detected during single breakdowns) increased between 2016 and 2024, in particular amongst dairy and mixed herd management systems. There has also been an increase in lesions detected per attested bovine sent for routine slaughter, suggesting a genuine increase in infection levels rather than only increased ante-mortem surveillance sensitivity (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). In this paper, we focus on quantifying drivers of case counts disclosed during Irish bTB breakdowns, capitalising on rich datasets spanning more than two decades on cattle movement, demographics and bTB testing. For clarity, we highlight that we are not addressing risk-factors for bTB breakdown occurrence, which have been previously well reported (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e). Rather, we are investigating risk factors to explain the larger case-counts within breakdowns, which have been occurring in dairy and mixed herd types over recent years.\u003c/p\u003e\u003cp\u003eA previous study in Ireland, Clegg et al. (2018), investigated risk factors for case counts of greater than 13 versus case counts of between 2 and 4 in breakdowns beginning in 2014 or 2015 (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). This study found that the year the breakdown started, increasing herd-size, previous exposure to bTB, increased bTB incidence in the local area, the presence of an animal with a bTB lesion and a bTB breakdown in a herd with shared staff / management / facilities were risk factors associated with larger breakdowns. Further studies have used breakdown duration (\u003cspan additionalcitationids=\"CR22\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e–\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e) or additional cases detected within the breakdown after the initiating test (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e) as indirect evidence of \u003cem\u003eM. bovis\u003c/em\u003e transmission, with many risk factors identified, in particular herd size and lesioned ante-mortem reactors, overlapping with those of Clegg et al. (2018) (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eWe build on this previous work, considering a broader range of breakdown case counts, a wider time window, and incorporating a range of extra metrics of movement and neighbourhood burden. Furthermore, we incorporate a time-window comprising substantial demographic changes in the Irish cattle industry (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e). For the first time, we consider cattle residency times (the amount of time an animal resided in a herd before being sent to another herd or for slaughter) within herds as an explanatory variable for case counts. For infectious diseases, the potential for transmission increases with the duration of contact between infectious and susceptible pairs (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e). As a proxy time for contact time between individuals, we would therefore expect residency time to also be an important correlate for risk of transmission.\u003c/p\u003e\u003cp\u003eWithin such a heavily managed test-and-cull system, the observed patterns of case counts are driven by both the dynamics of transmission of infection and removal of infection through the surveillance system. Understanding how to control bTB in Ireland depends on understanding the interaction of these two, which requires mechanistic models of transmission (\u003cspan additionalcitationids=\"CR28 CR29\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e–\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e). As a first step towards the construction of such models for Ireland, and to improve our understanding of the factors underlying the distribution of case counts within herds, we carried out the descriptive analysis and risk factor study presented here.\u003c/p\u003e\u003cp\u003ePotential drivers of detection of infected cattle in the herd, and subsequent case counts, include retained infected cattle from previous breakdowns, introduction of infection from the neighbourhood, inward movement of infected cattle and cattle-to-cattle transmission within the herd. Surveillance system characteristics, including diagnostic test characteristics, policy and implementation practices also influence case counts.\u003c/p\u003e\u003cp\u003eBreakdowns are initiated by bTB case detections. These comprise lesion detections in attested cattle going to slaughter for human consumption, or alternatively through Ireland’s ante-mortem bTB surveillance programme based on the comparative intra-dermal skin test (skin test hereafter) (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). In breakdowns with larger case counts, interferon gamma testing may be implemented to detect additional infected cattle (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e). Within the managed system of bTB control in Ireland, it is challenging to differentiate between the contribution of surveillance intensity and true within-herd prevalence of infection on the observed patterns of incidence. For example, fattener herds have more intensive slaughterhouse surveillance compared to dairy herds as they have greater rates of movement to slaughter (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e). To address this, as well as considering the full population of breakdowns (Full population A : total cases), where the response variable comprises count of all bTB cases, we consider a sub-population of breakdowns (Sub-population B : standard skin cases). In this sub-population, breakdowns initiated by slaughterhouse cases are excluded. Further variation in case count due to heterogeny in surveillance effort arises from evolving interferon-γ testing policy over our study period (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e), and to a lesser degree, that for application of severe skin test interpretation (\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e). Therefore, for our sub-population, we consider count of standard skin test reactors as our response variable, as policy relating to identifying these has not changed. Finally, although the specificity of the skin test is likely to be high (\u003cspan additionalcitationids=\"CR35\" citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e–\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e), we exclude the possibility of breakdowns due to false positive case detections within our “Sub-population B : standard skin cases,” considering only those with at least one ante-mortem reactor which went on to have a lesion detected at slaughter.\u003c/p\u003e\u003cp\u003eBased on this approach, our study is based on two populations, for which we present separate, but complementary analyses. For our full, unfiltered population of breakdowns initiated by any method (“Full population A : total cases”) all methods of bTB case detection contribute to our case count response variable. Our sub-population of breakdowns (“Sub-population B : standard skin cases”) are filtered, with the rationale above, to exclude slaughterhouse initiated breakdowns and those with no bTB lesion reported upon post-mortem examination ante-mortem reactors. The response variable for this sub-population is count of standard skin test reactors. With these analyses, we aim to take an initial step towards quantifying the relative roles of potential drivers of \u003cem\u003eM. bovis\u003c/em\u003e transmission within herds.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eData\u003c/p\u003e\u003cp\u003eCattle movement and bTB surveillance data were available from between 2000 and 2023 (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e). Detailed herd management data, integrated with cattle movements and demographics, were available from 2008 onwards (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e). Up-to-date data on herd location was available through the Irish Land Parcel Identification system (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e). We describe data generation and processing in detail in our Supplementary Methods but give a brief overview here.\u003c/p\u003e\u003cp\u003eFull population A : total cases\u003c/p\u003e\u003cp\u003e“Full population A : total cases” was defined to address our primary question examining a response variable defined by the total count of cases within bTB breakdowns ending between 2012 and 2023. This total case count includes skin test reactors at both the standard and severe interpretation, interferon-γ test reactors and slaughterhouse cases. The choice of time period was informed by the requirement of two- and three-year windows (respectively) in advance of the year of breakdown start to calculate our cattle movement and herd-level neighbourhood burden metrics (Supplementary Methods Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Therefore, using the more complete movement data from 2008 onwards and allowing for our average three year in degree metric (average count of unique source herds for inward moving cattle in the three years preceding the breakdown start), meant excluding breakdowns beginning before 2010. To avoid selection of only shorter breakdowns in earlier years, we excluded breakdowns ending in 2010 and 2011. This choice was supported by only 0.5% of our full breakdown population having a duration longer than two years. We used year of breakdown end for our analyses to avoid right censoring issues. After applying these constraints we are left with a study population of breakdowns ending between 2012 and 2023, and starting after 2009. Our residency time calculations required herds which comprised at least a single animal each year between 2008 and 2023. We also excluded the 2% of breakdowns in herds which comprised ten cattle or fewer, as their practices may be more aligned with hobby-herd management rather than the broader cattle industry.\u003c/p\u003e\u003cp\u003eSub-population B : standard skin cases\u003c/p\u003e\u003cp\u003eFrom “Full population A : total cases”, we selected the subset of skin-test initiated breakdowns which we termed “Sub-population B : standard skin cases”. With the rationale, as described in our Introduction, of reducing variation due to heterogeneity in surveillance effort, and focussing on drivers of \u003cem\u003eM. bovis\u003c/em\u003e transmission, we excluded breakdowns initiated by slaughterhouse surveillance, and included only those with at least one lesioned skin test reactor. To reduce potential bias from combining multiple surveillance mechanisms which evolved over time, the outcome variable for “Sub-population B : standard skin cases” was counts of standard skin test reactors only.\u003c/p\u003e\u003cp\u003eExplanatory variables\u003c/p\u003e\u003cp\u003eExplanatory variables included herd size (as calculated by averaging the herd sizes profiled from AIM movement data on the first of January, May and September each year), mean residency time of cattle leaving the herd the year of the breakdown start, initiating test type of the breakdown, herd breakdown history (categorised as between 0 and 5 + years of clear surveillance in advance of the breakdown start), in degree (count of unique source herds for inward moving cattle averaged for the year of breakdown start and the two preceding years), year of breakdown end, and a metric of neighbourhood burden. Only 1% of our breakdowns were initiated by pre-movement testing, so, similarly to Clegg et al. (2018) (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e), we merged this category with annual skin testing.\u003c/p\u003e\u003cp\u003eBoth herd and animal level metrics have been previously used to measure neighbourhood burden of bTB (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e). Adapting from Tratalos et al. (2023) (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e), our herd level metric was the proportion of herds within six kilometres of the herd of interest centroids, with a breakdown history within the past three years. We also investigated the animal level metrics explored by Clegg et al (2018) (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e) and Houtsma et al. (2018) (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e). These were, for the year before the breakdown start, skin reactors per 1000 cattle tested, or skin reactors per square kilometre, in the electoral division of the herd. The mean size of an Irish electoral division is 20.6 km\u003csup\u003e2\u003c/sup\u003e (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). Only one neighbourhood related variable was considered at a time in our models.\u003c/p\u003e\u003cp\u003eAfter initial univariable and bivariable explorations, all continuous explanatory variables were categorised into quintiles.\u003c/p\u003e\u003cp\u003eStatistical analyses\u003c/p\u003e\u003cp\u003eThe distributions of case counts were plotted. The fit of different statistical distributions to the data were compared using Akaike’s Information Criterion (AIC). We trialled zero-truncated Poisson (zt-P), zero-truncated negative binomial (zt-nb) and zeta distribution fits. We attempted to fit distributions to total case counts in “Full population A : total cases”, and separately to standard skin reactors in “Sub-population B : standard skin cases”. Where models did not converge or exhibited evidence of poor fit, case counts were binned, and a series of logistic regression models fitted to the data to explain case counts greater than one, four and ten.\u003c/p\u003e\u003cp\u003eWe used multivariable generalised linear models to explore the associations between the explanatory variables and measures of case counts. As 32.8% of herds in our dataset contributed data from more than one breakdown, all of our models included a herd-level random effect on the intercept. Candidate variables for fixed effects comprised all potential explanatory variables justified biologically and from descriptive analyses (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e). Likelihood ratio testing was used to describe explanatory ability added by each variable to the models. Multicollinearities between explanatory variables were investigated by calculation of variance inflation factors.\u003c/p\u003e\u003cp\u003eAll analyses were conducted in the R Statistical Environment (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e). The R packages “VGAM” (\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e) and “glmmTMB” (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e) were used for model implementations.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eDescriptive analyses\u003c/p\u003e\u003cp\u003eOur full population comprised 37,176 breakdowns ending between 2012 and 2023, in 24,730 herds. Of these, 40.8% were beef-breeding, 26.9% dairy, 19.3% fattener, 9.0% mixed, 3.2% store and 0.7% trader (classified with reference to (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e)). In this population, there were 106,984 standard skin reactors, 32,433 severe skin reactors, 37,557 interferon-γ test reactors and 13,457 slaughterhouse cases. Amongst reactors, 44.2% of standard interpretation skin, 17.5% of severe interpretation skin and 19.7% of interferon-γ reactors had lesions detected at post-mortem examination.\u003c/p\u003e\u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e show case count summary statistics stratified by candidate explanatory variables. Broad descriptive analyses are reported in Supplementary Results 1. Case count increased with herd size (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA), residency time (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB) and neighbourhood burden (Supplementary Fig.\u0026nbsp;1.9). In the case of in-degree (inversely related to residency time, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA), the highest mean and 75th percentile case counts were associated with an in-degree category of 1, compared to an in degree of 0, or higher in-degree (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The lowest case counts were associated with the highest in-degree quintile. Dairy herds had the highest case counts, followed by mixed, beef breeding, fattener, store and trader management types in deceasing order of case count (Supplementary Fig.\u0026nbsp;1.4A, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn our full population, 41.0% of breakdowns were initiated by annual skin testing/pre-movement testing, 24.6% by a slaughterhouse lesion detection,18.1% by contiguous skin testing, 11.0% by post-derestriction skin testing and 5.3% by risk-based skin testing. Breakdowns initiated by lesion detection at routine slaughterhouse inspection had the lowest case counts (mean 4, median 1), followed by annual skin tests (mean 5, median 2). Risk-based, contiguous and post-derestriction skin test-initiated breakdowns had higher mean case counts and 75th percentiles (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC). Herds which have been without a breakdown for four or more years had lower case counts compared to breakdowns with more recent breakdown history. Year of breakdown end is not shown in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e but summarised in Supplementary Fig.\u0026nbsp;1.3 and considered in detail by Madden et al. (2025) (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eSummary statistics for case count stratified by potential explanatory variables. *\u0026ldquo;Sub-population B : standard skin cases\u0026rdquo; was restricted to breakdowns initiated by skin tests (excluding slaughterhouse initiated breakdowns) and breakdowns with at least one ante-mortem reactor which went on to have lesions detected at slaughter, and therefore had \u0026ldquo;NA\u0026rdquo; values in the corresponding table rows for these characteristics.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCharacteristic\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e\u003cp\u003eFull population A : total cases N\u0026thinsp;=\u0026thinsp;37,176\u003csup\u003e\u003cem\u003e1\u003c/em\u003e\u003c/sup\u003e\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eSub-population B : standard skin cases N\u0026thinsp;=\u0026thinsp;15,834\u003csup\u003e\u003cem\u003e1\u003c/em\u003e\u003c/sup\u003e\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTotal cases Mean (sd)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eTotal cases Median (IQR)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStandard skin positives Mean (sd)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eStandard skin positives Median (IQR)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eAt least one lesioned skin test reactor present\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.2 (3.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 2.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNA*\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eNA*\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e8.9 (16.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.0 (2.0, 9.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.7 (7.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eHerd size category\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[ 11, 43)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.6 (3.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 3.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e2.8 (2.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 3.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[ 43, 78)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3.6 (5.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 3.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3.7 (4.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[ 78, 126)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4.7 (8.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.7 (6.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[126, 208)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.3 (11.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.7 (7.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 7.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[208,1932]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e9.3 (21.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 7.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e7.6 (12.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (2.0, 8.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eMean residency time in years category\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[ 0.02, 1.07)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.7 (6.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 2.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3.0 (5.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 3.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[ 1.07, 1.60)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4.2 (9.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3.8 (5.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[ 1.60, 2.10)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.4 (12.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.5 (6.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[ 2.10, 2.80)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.4 (12.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.4 (8.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[ 2.80,18.00]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e7.9 (16.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.0 (1.0, 8.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e6.2 (9.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (2.0, 7.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eHerd management type\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBeef breeding\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4.4 (8.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.1 (5.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDairy\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e8.5 (17.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.0 (1.0, 8.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e6.6 (10.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (2.0, 7.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFattener\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.9 (6.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 2.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3.6 (5.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMixed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.2 (13.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.5 (8.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStore\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.5 (3.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 2.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e2.5 (2.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.0 (1.0, 3.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTrader\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.7 (2.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 2.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e2.1 (1.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 2.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eInitiating test of breakdown\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAnnual skin\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.0 (10.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.3 (6.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eContiguous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.5 (11.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.0 (1.0, 7.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.9 (6.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePost-derestriction\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e7.0 (14.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.0 (1.0, 7.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.6 (8.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRisk based\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e7.9 (15.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 8.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e6.0 (9.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSlaughterhouse\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3.5 (12.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 2.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNA*\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eNA*\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eProportion of herds within 6km with breakdowns in 3 years preceding breakdown start\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[0.00,0.08)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4.5 (10.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 3.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.3 (6.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[0.08,0.12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4.9 (12.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.5 (8.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[0.12,0.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.3 (11.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.7 (7.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[0.16,0.21)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.6 (12.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.9 (7.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[0.21,0.70]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.0 (12.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.1 (7.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eSkin reactors per 1000 cattle tested in district electoral division in year preceding breakdown start.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4.7 (11.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.5 (8.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[0.0404, 0.4705)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.4 (12.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.0 (8.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[0.4705, 2.0822)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.3 (11.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.8 (7.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[2.0822,258.0645]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.7 (12.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.8 (7.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eStandard reactors per square kilometer in district electoral division in year preceding breakdown start.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4.8 (11.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.4 (7.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[0.00805,7.29e-02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4.8 (10.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.6 (7.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[0.07289,2.54e-01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.2 (11.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.7 (7.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[0.25429,1.05e\u0026thinsp;+\u0026thinsp;02]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.8 (12.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.9 (7.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eIn-degree category (count unique source herds per annum averaged for two years preceding breakdown and year of breakdown start)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.7 (11.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.7 (6.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.6 (14.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.4 (9.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[ 2, 4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.0 (12.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.1 (7.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[ 4, 12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.2 (12.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.4 (6.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e[12,5621]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.9 (6.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.0 (1.0, 2.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3.6 (6.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 3.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eYears without bTB breakdowns prior to breakdown start\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e0 years clear\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.2 (13.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.5 (8.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e1 year clear\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.8 (13.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.2 (8.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2 years clear\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.0 (12.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.4 (8.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 6.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e3 years clear\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.0 (16.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.2 (10.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e4 years clear\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.6 (13.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e5.0 (8.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e5\u0026thinsp;+\u0026thinsp;years clear\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4.8 (10.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.0 (1.0, 4.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.4 (6.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.0 (1.0, 5.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eDairy and mixed herds were largest in size (Supplementary Fig.\u0026nbsp;1.4). Residency time distributions varied by herd type (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, Supplementary Figs.\u0026nbsp;1.5 and 1.6). In dairy and mixed herds, 42.9% and 17.1% (respectively) of residencies were of short duration (\u0026lt;\u0026thinsp;60 days). This was driven by the movement of calves off farms shortly after birth. Breeding herds (dairy, mixed, beef breeding) had sub-groups of cows with relatively longer residency times (23.8% dairy, 27.4% mixed and 22.9% beef breeding animals had residency times longer than 2 years, Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Store and fattener herds had shorter residency times compared to the breeding herds (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Traders were defined by Brock et al. to have at least 50% of their residency times shorter than 30 days, and therefore had the shortest residency times in our population (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, Supplementary Figs.\u0026nbsp;1.5 and 1.6).\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD shows that the in-degree and residency time measures were negatively correlated (Pearsons ρ = -0.53). Supplementary Figs.\u0026nbsp;1.7 and 1.8 summarise neighbourhood burden, in degree and initiating test type by herd management system. Alongside their shorter residency times (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) trader herds had the highest in-degree, followed by fattener herds, and, as would be expected, these also had the highest proportion of breakdowns initiated by slaughterhouse cases. Store and mixed herds had the next highest in-degree and dairy and beef breeding had the lowest in-degree. Dairy, mixed and fattener herds had higher neighbourhood burden measures compared to beef breeding, store and trader herds.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eA summary of 66,620,507 animal level residency times ending between 2008 and 2023, based on a movement dataset from between 2000 and 2023, stratified by herd management practice. All herds eligible for residency time estimates are included here, not only breakdown herds. \u0026ldquo;All residency times\u0026rdquo; considers all animals, whereas \u0026ldquo;Residency times\u0026thinsp;\u0026gt;\u0026thinsp;60 days\u0026rdquo; only animals with residency time\u0026thinsp;\u0026gt;\u0026thinsp;60 days. We report \u0026ldquo;\u0026gt;60 days\u0026rdquo; summaries to exclude the large numbers of very short calf residencies in dairy herds.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eAll residency times\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"3\" nameend=\"c8\" namest=\"c6\"\u003e\u003cp\u003eResidency times\u0026thinsp;\u0026gt;\u0026thinsp;60 days\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHerd type\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMedian (IQR) years\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eMean years\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e% \u0026lt; 60 days\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e% \u0026gt;2 years\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eMedian (IQR) years\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMean years\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003e% \u0026gt;2 years\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDairy\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.4 (0.07\u0026ndash;1.94)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e42.9%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e23.8%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.7 (0.84\u0026ndash;3.70)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e2.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e41.8%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMixed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.3 (0.37\u0026ndash;2.07)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.82\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e17.1%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e27.4%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.6 (0.9\u0026ndash;2.22)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e2.18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e33.0%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStore\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.9 (0.47\u0026ndash;1.41 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.06\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e9.7%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e8.0%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.0 (0.59\u0026ndash;1.46 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1.17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e8.9%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFattener\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.8 (0.42\u0026ndash;1.33 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.98\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e7.1%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e8.1%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.9 (0.52\u0026ndash;1.38 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1.05\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e8.8%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBeef breeding\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.1 (0.61\u0026ndash;1.92 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.81\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e7.0%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e22.9%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.2 (0.69\u0026ndash;1.99 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1.94\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e24.6%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUnknown\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.5 (0.34\u0026ndash;0.88 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.88\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e10.1%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e6.8%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.6 (0.4\u0026ndash;0.93 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.97\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e7.6%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTrader\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0 (0.01\u0026ndash;0.04 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e90.5%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.2%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.3 (0.22\u0026ndash;0.5 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.50\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e2.2%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAll\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.8 (0.15\u0026ndash;1.64 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.39\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e25.7%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e17.7%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.1 (0.64\u0026ndash;1.96 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1.85\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e23.8%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eDistributions of case counts\u003c/p\u003e\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eFull population A : total cases\u003c/h2\u003e\u003cp\u003eThe total case count distribution was right skewed with mean and standard deviation of 5.3 and 12.1, respectively. The median was 2 and the interquartile range (IQR) was 1\u0026ndash;4 cases. The maximum case count in the dataset was 396 cases. Of all the breakdowns, 46.7% had only a single case. In 78.7% of breakdowns, all cases were disclosed in the initial test. If only standard skin reactors and slaughterhouse cases were considered, the mean and standard deviation were 3.3 and 6.1 cases. The median was 1 case, the IQR was 1\u0026ndash;3 cases and the maximum was 228 cases.\u003c/p\u003e\u003cp\u003eThe highly skewed, fat-tailed distribution, with relatively high counts of single case breakdowns, made fitting an appropriate count distribution to the full dataset challenging. Of the three distribution fits trialled, the zeta distribution had the lowest AIC (164,466). The zero-truncated negative binomial distributions (zt-nb) and zero truncated Poisson (zt-P) distributions were associated with convergence challenges due to inflation of single case counts and had a higher AIC (167,992 for zt-nb and 475,610 for zt-P). As estimating and interpreting coefficients from a regression based on the zeta distribution can potentially present challenges (\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e), and convergence issues and non-random residual distributions indicated that the negative binomial and Poisson distributions were not appropriate, we had to develop another method of analysis. To this end we reframe the problem by discretising the case distribution and developing a series of logistic regression models for increasing sizes of breakdown (\u0026gt;\u0026thinsp;1 case, \u0026gt;4 cases and \u0026gt;\u0026thinsp;10 cases).\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eSub-population B : standard skin cases\u003c/h3\u003e\n\u003cp\u003eIn \u0026ldquo;Sub-population B : standard skin cases\u0026rdquo; (excluding breakdowns initiated by a slaughterhouse case and restricted to breakdowns with at least one lesion confirmed in an ante-mortem reactor), mean and standard deviation of standard skin reactor counts were 4.7 and 7.5 respectively. The median was 2 and the IQR was 1\u0026ndash;5. The maximum was 227 standard skin reactors.\u003c/p\u003e\u003cp\u003eIt was possible to fit all three distributions to \u0026ldquo;Sub-population B : standard skin cases\u0026rdquo; without convergence issues. The AIC of the zeta, zt-nb and zt-P fits were 77,023, 73,762 and 139,337 respectively. We selected the zt-nb based on having the lowest AIC value and no evidence of a systematic lack of fit (randomly distributed residuals).\u003c/p\u003e\u003cp\u003eLogistic regression models (Full population A : total cases)\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e summarises the odds-ratios (ORs) of the explanatory variables in the set of three logistic regressions (\u0026gt;\u0026thinsp;1 case, \u0026gt;4 cases and \u0026gt;\u0026thinsp;10 cases) in \u0026ldquo;Full population A : total cases\u0026rdquo;. Herd size and residency time were consistently associated with increasing total case counts. Herd management, initiating test type and year also helped explain the data. Slaughterhouse initiated breakdowns were strongly associated with single case breakdowns. The effect of year may have been influenced by changing interferon-γ testing policy and implementation over time. The neighbourhood metric had a small but statistically significant effect in explaining cases\u0026thinsp;\u0026gt;\u0026thinsp;1, but little effect in explaining case counts\u0026thinsp;\u0026gt;\u0026thinsp;4 or \u0026gt;\u0026thinsp;10. More detailed results from the model building process, alternative models, likelihood ratio testing and random effects are reported in Supplementary Results 2.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eZero-truncated negative binomial regression model (Sub-population B : standard skin cases)\u003c/p\u003e\u003cp\u003eIn \u0026ldquo;Sub-population B : standard skin cases\u0026rdquo; (excluding breakdowns initiated by a slaughterhouse case and restricted to breakdowns with at least one lesion confirmed in an ante-mortem reactor), the root-mean squared error (RMSE) of an intercept only model was 7.5 units. The addition of the herd-level random effect on the intercept reduced this to 6.7 units, and the full model with fixed and random effects reduced it to 6.4 units. The model with fixed effects but excluding random effects had an RSME of 7.2 units. Likelihood ratio testing indicated that herd size and mean residency time added the most explanatory power to the model (Supplementary Results 2). Herd management type and initiating skin test type also improved model fit, but to a lesser degree. Year of breakdown end added only minor explanatory power whereas there was no justification for retention of any of the neighbourhood burden metric variables in the model. Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e summarises the effect of the explanatory variables on the percentage change in mean predicted count. Detailed results for the zt-nb regression are reported in Supplementary Results 2.\u003c/p\u003e\u003cp\u003eIn our zt-nb regression, the outcome variable (mean count of standard skin reactors) was log-transformed as is routine within its generalised linear model construct. Given this log-transformation, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e suggests a non-linear relationship between herd-size and mean residency time and count of standard reactors. Supplementary Fig.\u0026nbsp;2.6 shows a marginal effects plot from extra analyses, where we replace herd size category with log of herd size, to allow for better visual characterisation of this non-linear relationship. We also present a similar marginal effects plot for log of mean residency time in Supplementary Fig.\u0026nbsp;2.7.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003eMain findings\u003c/h2\u003e\u003cp\u003eOur study is the first to integrate cattle residency times with a suite of other risk factors to explain variation in case counts. Our results demonstrate that, while herd size and residency times are important factors driving large case counts, metrics of neighbourhood risk are not. This suggests that, following initial introduction of infection, cattle-to-cattle transmission within herds plays a more important role in amplifying infections rather than further introduction of infection from the neighbourhood.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eResidency times\u003c/h3\u003e\n\u003cp\u003eDuration of contact between individuals is an essential component of modelling transmission of infection (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e). Our finding that mean residency times in herds are important predictors of case counts fits with this transmission biology. Breeding herds have subsets of cows with relatively longer residency times compared to store, fattener or trader herds. As part of national risk-based approaches to disease control, there is much to be gained from avoiding the introduction of infection into breeding herds, given the increased risk of transmission within these herds once they become infected.\u003c/p\u003e\u003cp\u003eAs well as allowing more time for cattle-to-cattle transmission to happen, longer residency times allow more time for detection of herd level infection, if present, with the diagnostic testing system. Herds with shorter residencies, fewer breakdowns and lower case counts (for example, store herds engaged in contract rearing replacement heifers for breeding herds) may still be relevant for dispersal of infection to other herds, even if the cattle are not present in the herd of origin for long enough to be detected. Residency time could also be a proxy for duration of exposure to a common extrinsic force of infection, such as wildlife in the neighbourhood. However, metrics of neighbourhood burden added little additional explanatory value to our models, making the cattle-to-cattle transmission hypothesis more plausible.\u003c/p\u003e\u003cp\u003eOur description of individual animal residencies, and variation in their distributions in the different herd management systems, highlights the potential for future studies to integrate these into \u003cem\u003eM. bovis\u003c/em\u003e transmission models. Furthermore, whilst mean residency time is the best proxy of contact times we have available for the current study deign, it is not a perfect measure of contact time for all animals in a herd. That is, it does not account for the heterogeneity of cattle management (particularly the long-term management animals in smaller epidemiological units/cohorts) within each specific herd. Better characterisation of within-herd contact patterns will enhance future modelling efforts.\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eHerd size\u003c/h2\u003e\u003cp\u003eCase count increased with herd size. Clegg et al. (2018) also reported an association between increasing herd size and large bTB breakdowns (comparing breakdowns with at least 13 reactors to those with 2\u0026ndash;4 reactors) (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). In Northern Ireland, Wright et al. (2013) similarly reported larger outbreaks with increasing herd size (\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eHerd size is also associated with the herd-level risk of having a bTB breakdown (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e). Cited reasons for this include indirect association with other risk factors for introduction of infection. For example, larger herds may buy in more cattle or be dispersed over more land parcels, have higher production intensity operations correlating with physiological or nutritional stress risk factors and aggregation (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan additionalcitationids=\"CR50\" citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e). These indirect associations could also arguably be the case for increased case counts within herds, but our study suggests that this is not the case, for the following reasons.\u003c/p\u003e\u003cp\u003eDairy herds are largest in size, raising the question whether herd size is associated with a subset of longer residencies which explain transmission. However, beef breeding herds have a similar proportion of longer cow residencies, but lower case counts. Larger herds may also occupy a greater geographical area and distinct land parcels and therefore have more contact with extrinsic forces of infection (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e). The true area occupied by cattle in a herd is challenging to quantify as land parcels associated with a herd may be used for silage or crop production (\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e). Despite this challenge, the lack of effect of any of our neighbourhood burden measures on case count in our study suggests that herd size may be associated with a separate case count driver to farm footprint.\u003c/p\u003e\u003cp\u003eThe combination of large herd size and increasing emphasis on diagnostic test sensitivity over specificity, as case count increases within a breakdown, may be associated with increased false positives in larger herds. However, the strong effect of herd size remains in our subset analysis, where we exclude the possibility of false positive breakdowns and use only standard skin reactors as our response variable. Downs et al. (2016) (\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e) highlight that a lower proportion of dairy cattle skin reactors have lesions at slaughter compared to beef breed skin reactors, which is consistent with Irish data (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). However, Madden et al. (2025) highlight increased lesioned skin reactors in dairy and mixed management types in recent years, suggesting a genuine increase in infection levels (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). This combination of evidence suggests that false positives, or increased bTB test sensitivity in dairy cattle, do not fully explain increasing case counts with herd size.\u003c/p\u003e\u003cp\u003eDairy and mixed herds have increased in size over the past decade in Ireland (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e). Production of milk solids per cow has also increased (\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e). These developments may be associated with increased cattle-to-cattle transmission. Increasing herd size may be associated with increased intensification, and extra physiological stress on cattle. For example, high yielding cattle in large and intensive dairy herds may be under increased nutritional stress, and more susceptible to infection, especially in the peri-partum period (\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e). They may need to compete for cubicles or other resources in recently expanded herds which have not yet updated their housing facilities. Optimum group size for expression of typical social behaviour may be as low as between three and five cattle (\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e), whilst behavioural studies of dairy herds suggest that peer-to-peer (\u0026ldquo;kindergarten\u0026rdquo;) bonds, developed at a young age between calves managed in the same group, shape cattle\u0026rsquo;s behaviour (\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e). Dairy herd size related increases in contract rearing of replacement heifers (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e), may increase behavioural stress due to new introductions of cattle, loss of early-life peer-to-peer bonds and group sizes far above the optimum.\u003c/p\u003e\u003cp\u003eIn our study, although case count increases, proportion of cases relative to whole herd size drops as herd size increases (Supplementary Fig.\u0026nbsp;1.10). This non-linear increase in case count with herd size in this study may be consistent with a transmission pattern somewhere between no impact of herd size on effective contact rate (\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e), termed frequency dependence, and some degree of scaling of transmission with herd size, which can be termed non-linear density dependence (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e). Both frequency and density dependence are mathematical generalisations of heterogenous biological processes influencing effective contact rates between animals (\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e). In the future, synthesis of nuanced understanding of herd management, high resolution data representing within-herd contact patterns and pathobiology related to infectiousness, with modelling, may remove much of the dichotomy between density and frequency dependent transmission of infection.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eNeighbourhood burden\u003c/h3\u003e\n\u003cp\u003eIn studies of herd-level breakdown risk, metrics of neighbourhood burden are important predictors of infection being disclosed (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e). In contrast to this, and in the different context of considering case counts once breakdowns are already initiated, we did not detect a substantial role for neighbourhood burden. In \u0026ldquo;Full population A : total cases\u0026rdquo;, neighbourhood burden had a weak positive effect on the probability of case counts greater than one, but less impact on larger case counts. In the \u0026ldquo;Sub-population B : standard skin cases,\u0026rdquo; its retention in the model was not justified by likelihood ratio testing. This may be consistent with the hypothesis that, after initial introduction of infection, cattle-to-cattle transmission, rather than extrinsic force of infection, is the dominant amplifier of infection and associated case counts.\u003c/p\u003e\u003cp\u003eContiguous herd tests are implemented at four month intervals in herds adjacent to a herd undergoing a breakdown with two or more cases (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e), and therefore may indirectly reflect neighbourhood burden. Compared to annual skin testing, contiguous testing was associated with increased probability of higher case counts, but to a lesser degree than risk-based testing. However, the effect of neighbourhood burden on case counts was similar even if initiating test type was excluded from the model (Supplementary Figs.\u0026nbsp;2.4 and 2.5), further evidencing the lack of neighbourhood effect. Application of local veterinary epidemiological assessments may explain why contiguous herd tests were positively associated with case count (compared to annual skin tests). The local government veterinary inspector decides to implement contiguous testing based on risk assessment of the index breakdown and consideration of neighbourhood land usage by cattle (\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e). Therefore, contiguous testing may be considered a special case of risk-based testing. In the case of total case count in the full population, a further explanation could be the immediate implementation of severe skin test interpretation in contiguous tests. This cannot explain the positive influence of contiguous testing on the count of standard skin reactors in the subpopulation, but it is possible that risk-based enhanced vigilance in interpreting skin readings in contiguous herd tests, may also play a role.\u003c/p\u003e\u003cp\u003eThe minimal effect of any of our metrics of neighbourhood burden on case counts in our study contrasts with that reported by Clegg et al (2018), who reported increased odds of a case count of 13 or more (compared to 2\u0026ndash;4) cases per logged unit of reactors per km\u003csup\u003e2\u003c/sup\u003e in the local electoral division (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). As well as our herd-level metric of neighbourhood burden, to allow better comparison with Clegg et al. (2018), we included animal-level metrics of neighbourhood burden as alternative explanatory variables. Despite this, none of our neighbourhood burden metrics added substantial explanatory ability to our models. The reasons for this difference with Clegg et al. (2018) are unknown, and require further investigation. We considered whether it may potentially be associated with the more extensive demographic and movement data available for the current study. We had access to the nuanced herd management categories developed by Brock et al. (2021, 2022) (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e) and incorporated mean residency time. Reactors per km\u003csup\u003e2\u003c/sup\u003e may correlate with cattle per km\u003csup\u003e2\u003c/sup\u003e, which in turn is associated with the \u0026ldquo;dairy belt\u0026rdquo; in the southern area of Ireland and the largest herd sizes (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e). However, even when we omit herd management and residency time related variables, the effect neighbourhood burden remains insignificant (results not shown).\u003c/p\u003e\u003cp\u003eClegg at al. (2018) reported that number of badgers captured over the previous 10 years within 1 km of the farm was not associated with the risk of larger breakdowns, when controlling for covariates (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). However, due to the targeted nature of badger culling around breakdown herds, such metrics are confounded by cattle bTB related badger culling effort. Milne et al. (2023) similarly reported a lack of conclusive evidence for estimates of local badger density as an explanatory variable for breakdown duration in Northern Ireland (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e), despite other risk factor analyses finding positive associations with the probability of a breakdown in both Northern Ireland (\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e) and in Ireland (\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e). Our work does not directly address force of infection from badgers in the neighbourhood. However, based on a previous study which showed that bTB cases in cattle herds are sentinels for \u003cem\u003eM. bovis\u003c/em\u003e infection in badgers (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e), we believe that measures of cattle bTB in the neighbourhood are correlated with badger bTB levels. Indeed, national scale testing of badgers from 2007\u0026ndash;2012 in Ireland found significant spatial contemporaneous and lagged associations between badger TB status and local cattle TB prevalence at scales ranging from \u0026lt;\u0026thinsp;250m to 1km around capture sites (\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e). Therefore, in the current study, lack of evidence for neighbourhood burden driving case counts within breakdowns is consistent with previous work showing lack of badger related effects on breakdown size and duration (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e). Given these results, our conceptual model is one of case counts being primarily driven by within herd dynamics and less by what is happening in the wider neighbourhood via direct or indirect routes.\u003c/p\u003e\n\u003ch3\u003eInitiating test type\u003c/h3\u003e\n\u003cp\u003eIn comparison to annual skin testing, risk-based testing had the greatest positive effect on case count, supporting the current policy of enhanced risk based surveillance (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). A similar result was reported from Northern Ireland, specifically with backward contact tracing related testing (\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e). Risk based tests in our study included forward- and backward- contact tracing check tests, tests to regain free trading status after an inconclusive slaughterhouse lesion result, miscellaneous checks or because of an association with another herd with bTB cases, for example through shared land, facilities or management. As well as increased risk of infection, these tests may be associated with increased vigilance for extra cases, due to epidemiological linkages with other bTB cases. The finding of Clegg et al. (2018) (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e) that being under the same management as a herd with a bTB breakdown is a predictor of a large breakdown is consistent with our finding. This increased vigilance may also apply to contiguous testing of herds adjacent to bTB breakdowns and post-derestriction testing at 6, 12 and 18 months after the end of a breakdown. These initiating test types had a positive effect (but smaller than risk-based testing) on case counts, relative to annual skin testing. However, they had no effect on the probability the largest breakdowns with \u0026gt;\u0026thinsp;10 cases, consistent with what Clegg et al. (2018) reported (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). This may suggest that the more intensive post-breakdown and contiguous surveillance is detecting infection before it has the chance to be amplified to produce a large breakdown.\u003c/p\u003e\u003cp\u003eSlaughterhouse initiated breakdowns were strongly associated with single case counts during the present study. This is consistent with the previous work in the Republic of Ireland (\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e) which highlighted that 19.7% and 19%, respectively, of breakdowns initiated by lesions at routine slaughter went on to have further cases, which is lower than the overall proportion of breakdowns with greater than one case (53.3% in this study). Wright et al. (2013) reported a similar finding in Northern Ireland (\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e). The underlying biology of this finding may be partially explained by the association between trader and fattener herds which have relatively larger proportions of movements to slaughter (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e), and therefore more cattle from the herd subjected to routine PM inspection for lesions (increasing the probability of early detection), and shorter residency times (decreasing the probability of within herd transmission and time window for detection). Similarly, Clegg et al. (2018) reported that herds that introduced more cattle had smaller bTB breakdowns (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e), consistent with the negative association we found between the highest three in degree quintiles and case count. This is likely to be due to the positive correlation between rates of inward and outward movements (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e) and consistent with the negative correlation we found between residency time and in degree (which in turn is also correlated with inward movements). Slaughterhouse initiated breakdowns, high rates of inward and outward movements, high in degree and low case counts are all associated with fattener and trader herds. Store herds also have high in-degree and shorter residency times, but fewer movements to slaughter. This may also explain the smaller case counts in store herds, in the absence of a high proportion of slaughterhouse cases. To summarise, multiple factors that are associated with shorter residency times are also associated with lower case counts.\u003c/p\u003e\u003cp\u003eTwo previous Irish studies, (\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e), reported that, when slaughterhouse lesions were detected in breeding herds (dairy or beef breeding), there were larger counts of skin test reactors subsequently, compared to non-breeding herds, which is consistent with the hypothesis above. Our description of case counts stratified by both initiating test type and breeding herd status (Supplementary Fig.\u0026nbsp;1.12) similarly shows higher mean case count in breeding compared to non-breeding herds.\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eHerd type\u003c/h2\u003e\u003cp\u003eOur multivariable analyses suggest that dairy herds have larger breakdowns compared to beef breeding herds, even after herd size and residency time are accounted for. Dairy management is recognised as a risk factor for breakdown probability in many studies (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e). As we discussed in the context of increased herd size, potential mechanisms for the positive effect of dairy management on case count may include physiological pressures (parturition, energy balance, nutrition), genetic susceptibility or management practices, and this is an area for future research.\u003c/p\u003e\u003cp\u003eIn \u0026ldquo;Sub-population B : standard skin cases\u0026rdquo;, when slaughterhouse initiated breakdowns were excluded, fattener herd type was associated with larger breakdowns. This was also the case for breakdowns with \u0026gt;\u0026thinsp;10 cases in \u0026ldquo;Full population A : total cases\u0026rdquo;, but not for the smaller case count categories. Controlled-finishing-units (CFUs, fattener style herds which cannot sell to other herds, are required to have increased biosecurity requirements, a subject to reduced testing requirements and often have prolonged breakdown status) were excluded from our study population, precluding CFU status as an explainer of larger case counts in fattener herds. Therefore, this subset of fattener herds with larger breakdowns warrants further investigation. In Northern Ireland, such herds were associated with small case counts (measured as the number of reactors), but tended to have higher strain diversity, indicative of multiple pathway introductions, for example via trade (\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e). It is notable that, in Northern Ireland, long-term genetic typing of infection suggested that dairy herd breakdowns were dominated by single genotype breakdowns, indicative of within-herd local spread of infection or \u0026lsquo;microepidemics\u0026rsquo;, when compared to smaller sized beef fattening enterprises, where higher strain richness occurred, linked to more buying-in of animals (\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e). We found that a subset of fattener herds have large breakdowns. Synthesising our finding of a subset of fattener herds with larger breakdowns with the Northern Irish findings of increased genetic diversity of \u003cem\u003eM. bovis\u003c/em\u003e in fattener herds (\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e), highlights the risk that fattener herds may be a source of new \u003cem\u003eM. bovis\u003c/em\u003e variants into the neighbourhood, even though they are sending cattle to slaughter rather than to other herds. It is unknown, but important to find out, whether the cattle bought in to fatten, and sent relatively quickly to slaughter, are introducing new \u003cem\u003eM. bovis\u003c/em\u003e transmission chains which become established in the neighbourhood. Therefore, investigating the proximity to such fattener herds as a risk factor for bTB breakdowns may be an important area for future research in the spatial epidemiology of \u003cem\u003eM. bovis\u003c/em\u003e.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eYear\u003c/h2\u003e\u003cp\u003eOur data descriptions, in line with the in-depth temporal analyses of (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e), show that the mean and variation of case counts in dairy and mixed herd management types have increased over recent years. However, once evolving interferon-γ testing policy and implementation is set aside (by considering only standard skin reactors in \u0026ldquo;Sub-population B : standard skin cases\u0026rdquo;), in our multivariable regression models, year does not add much explanatory ability to our multivariable regression models. Our and Madden et al. (2025) \u0026rsquo;s work (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) demonstrates that higher case counts represent a true increase in infection levels, and are not only due to increased surveillance effort, as there is an increase in lesioned skin reactors, slaughterhouse cases and standard skin test reactors. Given the minimal effect of year, after excluding interferon-γ testing, our study suggests that changes in herd size and management may account for much of the year related effect through increased cattle-to-cattle transmission of infection.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003eModelling challenges, limitations and future work\u003c/h2\u003e\u003cp\u003eOur challenges in fitting statistical distributions to our full dataset are as expected, given the multiple heterogenous processes contributing to the distribution of case counts. We addressed this issue with relatively straightforward analytic approaches. In the full population, we conducted a series of logistic regressions. Separately, based on the hypotheses about the underlying processes, we defined a sub-population which minimised bias due to heterogeny in surveillance effort. An alternative approach in future work may be to capitalise on flexible Bayesian mixture models to explicitly incorporate heterogenous processes. The analyses presented in this paper provides evidence that residency times and herd size are important predictors of case count, and that neighbourhood burden plays a lesser role. Future work will involve mechanistic modelling to better understand these processes and model interventions to reduce transmission.\u003c/p\u003e\u003cp\u003eOur study focussed on describing potential drivers of case counts in bTB breakdowns. However, if solely statistical prediction was required, methods such as principal component analyses, elastic net regression or other machine learning approaches may lend themselves to analysis of multiple options for explanatory variables, many with collinearities.\u003c/p\u003e\u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eOur analyses suggest that herd size and residency time are the most important drivers of case count in Irish bTB breakdowns. This supports cattle-to-cattle transmission as an amplifier of infection within herds. After initial introduction of infection, within herd transmission seems to be the main driver of case counts. Protecting large breeding herds from introduction of infection may be particularly beneficial in preventing large scale amplification of \u003cem\u003eM. bovis\u003c/em\u003e in such herds. Our study adds to the evidence base on potential drivers of Irish bTB burden and will inform future mechanistic model development to explicitly model the underlying processes.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eAcknowledgements\u003c/p\u003e\n\u003cp\u003eWe would like to thank Professor Eamonn Gormley of University College Dublin, and Drs Nicola Harvey, Michael Horan and Philip Breslin of the Department of Agriculture, Food and the Marine, for helpful discussions relating to this study.\u003c/p\u003e\n\u003cp\u003eFunding\u003c/p\u003e\n\u003cp\u003eThis research received no specific funding. JM, JT (and up to September 2024 MC and SM) work in the Centre for Veterinary Epidemiology and Risk Analysis in University College Dublin which is funded by the Department of Agriculture, Food and the Marine (DAFM). MC is currently funded by a University College Dublin Ad Astra Fellowship.\u003c/p\u003e\n\u003cp\u003eCompeting interests\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;None declared.\u003c/p\u003e\n\u003cp\u003eData availability\u003c/p\u003e\n\u003cp\u003eThe datasets analysed during this study are available from Department of Agriculture, Food and the Marine (DAFM), but are subject to data protection regulations and limitations:\u003c/p\u003e\n\u003cp\u003ehttps://www.gov.ie/en/department-of-agriculture-food-and-the-marine/organisation-information/data-protection/\u003c/p\u003e\n\u003cp\u003eEthics declaration\u003c/p\u003e\n\u003cp\u003eThe datasets analysed during this study are from pre-existing databases of the \u0026nbsp;Department of Agriculture, Food and the Marine (DAFM), and were analysed in compliance with the General Data Protection Regulation of the European Union. No experimental procedures, animal handling, or interventions were performed. In accordance with the policies of BMC Veterinary Research and institutional guidelines, ethical approval was not required because the study used only secondary data that were collected previously for purposes unrelated to this research.\u003c/p\u003e\n\u003cp\u003eConsent to Publish declaration\u003c/p\u003e\n\u003cp\u003eConsent to Publish declaration: not applicable. This study uses only pre-existing, anonymized data and involves no animal or human subjects.\u003c/p\u003e\n\u003cp\u003eAuthor contributions\u003c/p\u003e\n\u003cp\u003eConceptualisation: MC, AC, JM; Methodology : MC, AC, JM, JT ; Formal analysis: MC, AC; Investigation: MC, AC, JM; Data curation: MC, JM, JC, Resources: SM, DB, Writing (original draft): MC, Writing (review): All authors; Visualisation: MC; Validation: MC, JM, AC, AB, SM. AC and JM contributed equally to this study.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eMore SJ, Good M. The tuberculosis eradication programme in Ireland: A review of scientific and policy advances since 1988. Vet Microbiol. 2006;112(2-4 SPEC. ISS.):239\u0026ndash;51. \u003c/li\u003e\n\u003cli\u003eRyan E, Breslin P, O\u0026rsquo;Keeffe J, Byrne AW, Wrigley K, Barrett D. The Irish bTB eradication programme: combining stakeholder engagement and research-driven policy to tackle bovine tuberculosis. Ir Vet J [Internet]. 2023;76(1):1\u0026ndash;13. Available from: https://doi.org/10.1186/s13620-023-00255-8\u003c/li\u003e\n\u003cli\u003eMadden J, Gormley E, McGrath G, McAloon C, Casey-Bryars M. A new online tool for near real-time exploration of cattle and bovine tuberculosis trends in Ireland [Internet]. 2025. Available from: https://j-madden-m.github.io/Irish_bTB_trends/\u003c/li\u003e\n\u003cli\u003eDAFM. Bovine TB Action Plan [Internet]. 2025. Available from: https://assets.gov.ie/static/documents/d4cfc18d/7784-DAFM_TB_Action_Plan_LR.pdf\u003c/li\u003e\n\u003cli\u003eDAFM. National bovine tuberculosis statistics. 2025. \u003c/li\u003e\n\u003cli\u003eD\u0026aacute;il \u0026Eacute;ireann. D\u0026aacute;il \u0026Eacute;ireann Debate, Wednesday - 14 May 2025 [Internet]. 2025. Available from: https://www.oireachtas.ie/en/debates/question/2025-05-14/160/\u003c/li\u003e\n\u003cli\u003eMore SJ. Eradication of bovine tuberculosis in Ireland: is it a case of now or never? Ir Vet J [Internet]. 2024;77(1):4\u0026ndash;9. Available from: https://doi.org/10.1186/s13620-024-00282-z\u003c/li\u003e\n\u003cli\u003eTratalos JA, Fielding HR, Madden JM, Casey M, More SJ. Can Ingoing Contact Chains and other cattle movement network metrics help predict herd-level bovine tuberculosis in Irish cattle herds? Prev Vet Med [Internet]. 2023;211(April 2022):105816. Available from: https://doi.org/10.1016/j.prevetmed.2022.105816\u003c/li\u003e\n\u003cli\u003eMore SJ, Good M. Understanding and managing bTB risk: Perspectives from Ireland. Vet Microbiol. 2015;176(3\u0026ndash;4):209\u0026ndash;18. \u003c/li\u003e\n\u003cli\u003eAllen AR, Skuce RA, Byrne AW. Bovine Tuberculosis in Britain and Ireland \u0026ndash; A Perfect Storm? the Confluence of Potential Ecological and Epidemiological Impediments to Controlling a Chronic Infectious Disease. Front Vet Sci [Internet]. 2018 Jun 5 [cited 2021 Oct 26];5:109. Available from: https://www.frontiersin.org/article/10.3389/fvets.2018.00109/full\u003c/li\u003e\n\u003cli\u003eGriffin JM, Williams DH, Kelly GE, Clegg TA, O\u0026rsquo;Boyle I, Collins JD, et al. The impact of badger removal on the control of tuberculosis in cattle herds in Ireland. Prev Vet Med [Internet]. 2005 Mar 1 [cited 2021 Jun 24];67(4):237\u0026ndash;66. Available from: https://www.sciencedirect.com/science/article/pii/S0167587704002090?via%3Dihub\u003c/li\u003e\n\u003cli\u003eChang Y, Hartemink N, Byrne AW, Gormley E, McGrath G, Tratalos JA, et al. Inferring bovine tuberculosis transmission between cattle and badgers via the environment and risk mapping. Front Vet Sci. 2023;10. \u003c/li\u003e\n\u003cli\u003eBrock J, Lange M, Tratalos JA, Meunier N, Guelbenzu-Gonzalo M, More SJ, et al. The Irish cattle population structured by enterprise type: overview, trade \u0026amp; trends. Irish Vet J 2022 751 [Internet]. 2022;75(1):1\u0026ndash;11. Available from: https://irishvetjournal.biomedcentral.com/articles/10.1186/s13620-022-00212-x\u003c/li\u003e\n\u003cli\u003eGeneral O of the C and A. Bovine TB eradication programme [Internet]. 2025. Available from: https://www.audit.gov.ie/en/find-report/publications/2025/14-bovine-tb-eradication-programme.pdf\u003c/li\u003e\n\u003cli\u003eClegg TA, Doyle M, Ryan E, More SJ, Gormley E. Characteristics of Mycobacterium bovis infected herds tested with the interferon-gamma assay. Prev Vet Med [Internet]. 2019;168(April):52\u0026ndash;9. Available from: https://doi.org/10.1016/j.prevetmed.2019.04.004\u003c/li\u003e\n\u003cli\u003eMore SJ, Houtsma E, Doyle L, McGrath G, Clegg TA, De La Rua-Domenech R, et al. Further description of bovine tuberculosis trends in the United Kingdom and the Republic of Ireland, 2003-2015. Vet Rec. 2018;183(23):717. \u003c/li\u003e\n\u003cli\u003eClegg TA, Duignan A, More SJ. The relative effectiveness of testers during field surveillance for bovine tuberculosis in unrestricted low-risk herds in Ireland. Prev Vet Med [Internet]. 2015 Apr 1 [cited 2019 Aug 12];119(1\u0026ndash;2):85\u0026ndash;9. Available from: https://www.sciencedirect.com/science/article/pii/S0167587715000598?via%3Dihub\u003c/li\u003e\n\u003cli\u003eSkuce RA, Allen AR, McDowell SWJ. Herd-level risk factors for bovine tuberculosis: a literature review. Vet Med Int [Internet]. 2012 [cited 2021 Jun 17];2012:621210. Available from: http://www.ncbi.nlm.nih.gov/pubmed/22966479\u003c/li\u003e\n\u003cli\u003eBroughan JM, Judge J, Ely E, Delahay RJ, Wilson G, Clifton-Hadley RS, et al. A review of risk factors for bovine tuberculosis infection in cattle in the UK and Ireland. Epidemiol Infect. 2016;144(14):2899\u0026ndash;926. \u003c/li\u003e\n\u003cli\u003eClegg TA, Good M, Hayes M, Duignan A, McGrath G, More SJ. Trends and predictors of large tuberculosis episodes in cattle herds in Ireland. Front Vet Sci. 2018;5(MAY):1\u0026ndash;12. \u003c/li\u003e\n\u003cli\u003eKarolemeas K, McKinley TJ, Clifton-Hadley RS, Goodchild AV, Mitchell A, Johnston WT, et al. Predicting prolonged bovine tuberculosis breakdowns in Great Britain as an aid to control. Prev Vet Med [Internet]. 2010 Dec 1 [cited 2021 Sep 8];97(3\u0026ndash;4):183\u0026ndash;90. Available from: https://www.sciencedirect.com/science/article/pii/S0167587710002564?via%3Dihub\u003c/li\u003e\n\u003cli\u003eByrne AW, Barrett D, Breslin P, Madden JM, O\u0026rsquo;Keeffe J, Ryan E. Bovine Tuberculosis (Mycobacterium bovis) Outbreak Duration in Cattle Herds in Ireland: A Retrospective Observational Study. Pathogens [Internet]. 2020 Oct 5 [cited 2021 Sep 27];9(10):815. Available from: https://www.mdpi.com/2076-0817/9/10/815\u003c/li\u003e\n\u003cli\u003eMilne G, Allen A, Graham J, Lahuerta-Marin A, McCormick C, Presho E, et al. Bovine tuberculosis breakdown duration in cattle herds: An investigation of herd, host, pathogen and wildlife risk factors. PeerJ [Internet]. 2020 [cited 2021 Sep 27];2020(2):e8319. Available from: http://www.ncbi.nlm.nih.gov/pubmed/32117602\u003c/li\u003e\n\u003cli\u003eMadenci D, S\u0026aacute;nchez-Molano E, Madenci D, Tsairidou S, Winters M, Mitchell AP, et al. 700. Detection of genetic variability in cattle infectivity for bovine tuberculosis (bTB). J Dairy Sci [Internet]. 2025;2887\u0026ndash;90. Available from: http://dx.doi.org/10.3168/jds.2024-25697\u003c/li\u003e\n\u003cli\u003eBrooks-Pollock E, Keeling M. Herd size and bovine tuberculosis persistence in cattle farms in Great Britain. Prev Vet Med. 2009;92(4):360\u0026ndash;5. \u003c/li\u003e\n\u003cli\u003eVynnycky E, White R. An introduction to infectious disease modelling. Oxford: Oxford University Press; 2010. \u003c/li\u003e\n\u003cli\u003eConlan AJK, McKinley TJ, Karolemeas K, Pollock EB, Goodchild A V., Mitchell AP, et al. Estimating the Hidden Burden of Bovine Tuberculosis in Great Britain. Ferguson N, editor. PLoS Comput Biol [Internet]. 2012 Oct 18 [cited 2021 Apr 1];8(10):e1002730. Available from: https://dx.plos.org/10.1371/journal.pcbi.1002730\u003c/li\u003e\n\u003cli\u003eConlan AJK, Brooks Pollock E, McKinley TJ, Mitchell AP, Jones GJ, Vordermeier M, et al. Potential Benefits of Cattle Vaccination as a Supplementary Control for Bovine Tuberculosis. PLoS Comput Biol. 2015;11(2):1\u0026ndash;26. \u003c/li\u003e\n\u003cli\u003eConlan AJK, Vordermeier M, de Jong MCM, Wood JLN. The intractable challenge of evaluating cattle vaccination as a control for bovine tuberculosis. Elife. 2018;7:1\u0026ndash;39. \u003c/li\u003e\n\u003cli\u003eVanderWaal K, Enns EA, Picasso C, Alvarez J, Perez A, Fernandez F, et al. Optimal surveillance strategies for bovine tuberculosis in a low-prevalence country. Sci Rep [Internet]. 2017 Dec 23 [cited 2021 Oct 21];7(1):4140. Available from: http://www.nature.com/articles/s41598-017-04466-2\u003c/li\u003e\n\u003cli\u003eMadden JM, O\u0026rsquo;Donovan J, Casey-Bryars M, Sweeney J, Messam LL, McAloon CG, et al. The impact of changing the cut-off threshold of the interferon-gamma (IFN-\u0026gamma;) assay for diagnosing bovine tuberculosis in Ireland. Prev Vet Med. 2024;224(January). \u003c/li\u003e\n\u003cli\u003eBrock J, Lange M, Tratalos JA, More SJ, Graham DA, Guelbenzu-Gonzalo M, et al. Combining expert knowledge and machine-learning to classify herd types in livestock systems. Sci Rep [Internet]. 2021 Dec 4 [cited 2021 Feb 8];11(1):2989. Available from: http://www.nature.com/articles/s41598-021-82373-3\u003c/li\u003e\n\u003cli\u003eGriffin J, Aznar I, Breslin P, Good M, Gordon S, Gormley E, et al. What is the scope for existing (including recently developed) diagnostic methods to detect infected cattle which are not currently detected by the existing programme? Food Risk Assess Eur. 2024;1(2). \u003c/li\u003e\n\u003cli\u003ede la Rua-Domenech R, Goodchild AT, Vordermeier HM, Hewinson RG, Christiansen KH, Clifton-Hadley RS. Ante mortem diagnosis of tuberculosis in cattle: A review of the tuberculin tests, \u0026gamma;-interferon assay and other ancillary diagnostic techniques. Res Vet Sci. 2006;81(2):190\u0026ndash;210. \u003c/li\u003e\n\u003cli\u003eMurphy D, Gormley E, Collins DM, McGrath G, Sovsic E, Costello E, et al. Tuberculosis in cattle herds are sentinels for Mycobacterium bovis infection in European badgers (Meles meles): The Irish Greenfield Study. Vet Microbiol [Internet]. 2011;151(1\u0026ndash;2):120\u0026ndash;5. Available from: http://dx.doi.org/10.1016/j.vetmic.2011.02.034\u003c/li\u003e\n\u003cli\u003eGoodchild A V., Downs SH, Upton P, Wood JLN, De La Rua-Domenech R. Specificity of the comparative skin test for bovine tuberculosis in Great Britain. Vet Rec. 2015;177(10):258. \u003c/li\u003e\n\u003cli\u003eTratalos J, Madden J, McGrath G, Graham D, \u0026Aacute;ine Collins, More S. Spatial and network characteristics of Irish cattle movements. Prev Vet Med. 2020;183. \u003c/li\u003e\n\u003cli\u003eZimmermann J, Fealy RM, Lydon K, Mockler EM, O\u0026rsquo;Brien P, Packham I, et al. The Irish land-parcels identification system (LPIS) \u0026ndash; experiences in ongoing and recent environmental research and land cover mapping. Biol Environ. 2016;116B(1):53\u0026ndash;62. \u003c/li\u003e\n\u003cli\u003eHoutsma E, Clegg TA, Good M, More SJ. Further improvement in the control of bovine tuberculosis recurrence in Ireland. Vet Rec. 2018;183(20):622. \u003c/li\u003e\n\u003cli\u003eMadden J, McGrath G, Sweeney J, Murray G, Tratalos JA, More S. Spatio-temporal models of bovine tuberculosis in the Irish cattle population , 2012-2019. Spat Spatiotemporal Epidemiol. 2021;39. \u003c/li\u003e\n\u003cli\u003eHoutsma E, Clegg TA, Good M, More SJ. Further improvement in the control of bovine tuberculosis recurrence in Ireland. Vet Rec. 2018;183(20):622. \u003c/li\u003e\n\u003cli\u003eBursac Z, Gauss CH, Williams DK, Hosmer DW. Purposeful selection of variables in logistic regression. Source Code Biol Med. 2008;3:1\u0026ndash;8. \u003c/li\u003e\n\u003cli\u003eR Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing [Internet]. Vienna, Austria: R Foundation for Statistical Computing; 2024. Available from: http://www.r-project.org.\u003c/li\u003e\n\u003cli\u003eYee TW. Vector Generalized Linear and Additive Models: With an Implementation in R. Vector Generalized Linear and Additive Models: With an Implementation in R. 2015. 1\u0026ndash;589 p. \u003c/li\u003e\n\u003cli\u003eBrooks ME, Kristensen K, van Benthem KJ, Magnusson A, Berg CW, Nielsen A, et al. Package \u0026ldquo;glmmTMB\u0026rdquo; Title Generalized Linear Mixed Models using Template Model Builder. 2022;9(December):378\u0026ndash;400. Available from: https://orcid.org/0000-0001-5715-616X\u003c/li\u003e\n\u003cli\u003eDoray LG, Arsenault M. Estimators of the regression parameters of the zeta distribution. 2002;30:439\u0026ndash;50. \u003c/li\u003e\n\u003cli\u003eWright DM, Allen AR, Mallon TR, McDowell SWJ, Bishop SC, Glass EJ, et al. Field-Isolated Genotypes of Mycobacterium bovis Vary in Virulence and Influence Case Pathology but Do Not Affect Outbreak Size. PLoS One. 2013;8(9). \u003c/li\u003e\n\u003cli\u003eSkuce RA, Allen AR, McDowell SWJ. Herd-Level Risk Factors for Bovine Tuberculosis: A Literature Review. Vet Med Int [Internet]. 2012 [cited 2021 Jun 17];2012:1\u0026ndash;10. Available from: http://www.hindawi.com/journals/vmi/2012/621210/\u003c/li\u003e\n\u003cli\u003eMcGrath G, More S. Farm fragmentation in Ireland. Vet Ital. 2024;60(4). \u003c/li\u003e\n\u003cli\u003eHumblet MF, Gilbert M, Govaerts M, Fauville-Dufaux M, Walravens K, Saegerman C. New assessment of bovine tuberculosis risk factors in Belgium based on nationwide molecular epidemiology. J Clin Microbiol. 2010;48(8):2802\u0026ndash;8. \u003c/li\u003e\n\u003cli\u003eDowns SH, Durr P, Edwards J, Clifton-Hadley R. Trace micro-nutrients may affect susceptibility to bovine tuberculosis in cattle. Prev Vet Med. 2008;87(3\u0026ndash;4):311\u0026ndash;26. \u003c/li\u003e\n\u003cli\u003eMilne G, Graham J, McGrath J, Kirke R, McMaster W, Byrne AW. Investigating Farm Fragmentation as a Risk Factor for Bovine Tuberculosis in Cattle Herds: A Matched Case-Control Study from Northern Ireland. Pathogens. 2022;11(3). \u003c/li\u003e\n\u003cli\u003eDowns SH, Broughan JM, Goodchild A V., Upton PA, Durr PA. Responses to diagnostic tests for bovine tuberculosis in dairy and non-dairy cattle naturally exposed to Mycobacterium bovis in Great Britain. Vet J [Internet]. 2016;216:8\u0026ndash;17. Available from: http://dx.doi.org/10.1016/j.tvjl.2016.06.010\u003c/li\u003e\n\u003cli\u003eKelly P, Shalloo L, Wallace M, Dillon P. The Irish dairy industry \u0026ndash; Recent history and strategy, current state and future challenges. Int J Dairy Technol. 2020;73(2):309\u0026ndash;23. \u003c/li\u003e\n\u003cli\u003eTrevisi E, Amadori M, Cogrossi S, Razzuoli E, Bertoni G. Metabolic stress and inflammatory response in high-yielding, periparturient dairy cows. Res Vet Sci [Internet]. 2012;93(2):695\u0026ndash;704. Available from: http://dx.doi.org/10.1016/j.rvsc.2011.11.008\u003c/li\u003e\n\u003cli\u003eTakeda K ichi, Sato S, Sugawara K. The number of farm mates influences social and maintenance behaviours of Japanese Black cows in a communal pasture. Appl Anim Behav Sci. 2000;67(3):181\u0026ndash;92. \u003c/li\u003e\n\u003cli\u003eMarina H, Ren K, Hansson I, Fikse F, Nielsen PP, R\u0026ouml;nneg\u0026aring;rd L. New insight into social relationships in dairy cows and how time of birth, parity, and relatedness affect spatial interactions later in life. J Dairy Sci. 2024;107(2):1110\u0026ndash;23. \u003c/li\u003e\n\u003cli\u003eMarina H, Nielsen PP, Fikse WF, R\u0026ouml;nneg\u0026aring;rd L. Multiple factors shape social contacts in dairy cows. Appl Anim Behav Sci. 2024;278(July). \u003c/li\u003e\n\u003cli\u003ede Jong MCM, Diekmann O, Heesterbeek H. How does transmission of infection depend on population size. Vol. 94, Journal of the American Statistical Association. 1995. p. 84\u0026ndash;94. \u003c/li\u003e\n\u003cli\u003eFromsa A, Willgert K, Srinivasan S, Mekonnen G, Bedada W, Gebre S, et al. BCG vaccination reduces bovine tuberculosis transmission, improving prospects for elimination. 2024;3962. \u003c/li\u003e\n\u003cli\u003eBegon M, Bennett M, Bowers RG, French NP, Hazel SM, Turner J. A clarification of transmission terms in host-microparasite models: Numbers, densities and areas. Epidemiol Infect. 2002;129(1):147\u0026ndash;53. \u003c/li\u003e\n\u003cli\u003eGood M, Duignan A. Veterinary Handbook for Herd Management in the TB eradication programme [Internet]. Dublin; 2016. 23\u0026ndash;25 p. Available from: https://www.researchgate.net/publication/323402319_Veterinary_Handbook_for_herd_management_in_the_bovine_TB_Eradication_Programme\u003c/li\u003e\n\u003cli\u003eWright DM, Reid N, Montgomery WI, Allen AR, Skuce RA, Kao RR. Herd-level bovine tuberculosis risk factors: Assessing the role of low-level badger population disturbance. Sci Rep. 2015;5(November 2014):1\u0026ndash;11. \u003c/li\u003e\n\u003cli\u003eByrne AW, White PW, Mcgrath G, O\u0026rsquo;keeffe J, Martin W. Risk of tuberculosis cattle herd breakdowns in Ireland: effects of badger culling effort, density and historic large-scale interventions [Internet]. 2014 [cited 2021 Jun 25]. Available from: http://www.veterinaryresearch.org/content/45/1/109\u003c/li\u003e\n\u003cli\u003eByrne AW, Kenny K, Fogarty U, O\u0026rsquo;Keeffe JJ, More SJ, McGrath G, et al. Spatial and temporal analyses of metrics of tuberculosis infection in badgers (Meles meles) from the Republic of Ireland: Trends in apparent prevalence. Prev Vet Med [Internet]. 2015;122(3):345\u0026ndash;54. Available from: http://dx.doi.org/10.1016/j.prevetmed.2015.10.013\u003c/li\u003e\n\u003cli\u003eByrne AW, Barrett D, Breslin P, Madden JM, O\u0026rsquo;Keeffe J, Ryan E. Post-mortem surveillance of bovine tuberculosis in Ireland: herd-level variation in the probability of herds disclosed with lesions at routine slaughter to have skin test reactors at follow-up test. Vet Res Commun. 2020;44(3\u0026ndash;4):131\u0026ndash;6. \u003c/li\u003e\n\u003cli\u003eOlea-Popelka FJ, Costello E, White P, McGrath G, Collins JD, O\u0026rsquo;Keeffe J, et al. Risk factors for disclosure of additional tuberculous cattle in attested-clear herds that had one animal with a confirmed lesion of tuberculosis at slaughter during 2003 in Ireland. Prev Vet Med. 2008;85(1\u0026ndash;2):81\u0026ndash;91. \u003c/li\u003e\n\u003cli\u003eHumblet MF, Boschiroli ML, Saegerman C. Classification of worldwide bovine tuberculosis risk factors in cattle: A stratified approach. Vet Res. 2009;40(5). \u003c/li\u003e\n\u003cli\u003eMilne MG, Graham J, Allen A, Lahuerta-Marin A, McCormick C, Breadon E, et al. Herd characteristics, wildlife risk and bacterial strain genotypes in persistent breakdowns of bovine tuberculosis in Northern Irish cattle herds. In: M.L. Brennan \u0026amp; A. Lin, editor. Proceedings of the Society for Veterinary Epidemiology and Preventive Medicine Annual Meeting [Internet]. Talin, Estonia; 2018. p. 56\u0026ndash;67. Available from: https://svepm.org.uk/wp-content/uploads/2023/05/phpcK8JOc20180313220149.pdf#page=56\u003c/li\u003e\n\u003cli\u003eMilne MG, Graham J, Allen A, McCormick C, Presho E, Skuce R, et al. Variation in Mycobacterium bovis genetic richness suggests that inwards cattle movements are a more important source of infection in beef herds than in dairy herds. BMC Microbiol. 2019;19(1):1\u0026ndash;13. \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-veterinary-research","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [BMC Veterinary Research](http://bmcvetres.biomedcentral.com/)","snPcode":"12917","submissionUrl":"https://submission.nature.com/new-submission/12917/3?","title":"BMC Veterinary Research","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Bovine tuberculosis, Mycobacterium bovis, Residency time, Herd size, Statistical modelling, Within-herd transmission, Risk-factors","lastPublishedDoi":"10.21203/rs.3.rs-8181407/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8181407/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eBovine tuberculosis (bTB) poses challenges to sustainable livestock production, and to animal and human health globally. Its prevalence increased in Irish cattle herds between 2016 and 2023. Over the same period, bTB breakdowns (periods of trade restriction and enhanced surveillance within cattle herds due to bTB detection) in dairy herds saw an increase in the average and variability of bTB case numbers. To help inform Irish national control policy, we investigated the risk factors for large bTB breakdowns. Specifically, we set out to characterise the relationship between cattle residency times and case counts in Irish herds. For breakdowns ending between 2012 and 2023 (N\u0026thinsp;=\u0026thinsp;37,176 breakdowns in 24,730 herds) we describe the association between herd size, cattle residency times, breakdown history, initiating test type, year, inward movements, neighbourhood burden and bTB breakdown case counts of greater than one, four and ten cases, using logistic regressions. For a subpopulation of 15,834 skin test initiated breakdowns, we investigated the same risk factors for increasing standard skin reactor counts using zero-truncated negative-binomial regression. We show that increasing herd size and cattle residency times are the strongest predictors of increased case counts. Herd management and initiating test type are also important but play smaller roles. Dairy and fattener herd types, and risk-based skin tests are associated with the largest breakdowns. Slaughterhouse-initiated breakdowns were associated with single cases. Interferon-γ surveillance policy evolved over the study period. When cases detected by this test type were excluded, year did not account for much variation within the estimated models. Herd- or animal-level measures of neighbourhood burden added the least explanatory ability to our models. Our results demonstrate that, while herd size and residency times are important factors driving large case counts, metrics of neighbourhood risk are not. This suggests that, following initial introduction of infection, cattle-to-cattle transmission within herds plays a more important role in amplifying infections rather than further introduction of infection from the neighbourhood. Targeting larger breeding herds for prevention of introduction of infection may assist with the control of bTB. The causes and consequences of larger breakdowns in fattener herds warrant further research.\u003c/p\u003e","manuscriptTitle":"Residency time and other risk factors for large bovine tuberculosis breakdowns in Irish cattle herds","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-02 14:33:55","doi":"10.21203/rs.3.rs-8181407/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-08T04:45:58+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-05T21:22:22+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-29T09:17:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"226902587472944429783849125554013555830","date":"2026-03-16T01:43:50+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"287216153485390684974574236771983398966","date":"2026-02-24T02:29:29+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-11-28T11:34:20+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-11-28T11:15:04+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-11-27T11:59:30+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-11-27T11:57:26+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Veterinary Research","date":"2025-11-22T15:32:22+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-veterinary-research","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [BMC Veterinary Research](http://bmcvetres.biomedcentral.com/)","snPcode":"12917","submissionUrl":"https://submission.nature.com/new-submission/12917/3?","title":"BMC Veterinary Research","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2c53b4f1-3824-416d-b380-1a03d20a45f2","owner":[],"postedDate":"December 2nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-17T21:53:14+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-02 14:33:55","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8181407","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8181407","identity":"rs-8181407","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}