Article Text

Download PDFPDF

Paper
Bovine tuberculosis trends in the UK and the Republic of Ireland, 1995–2010
  1. D. A. Abernethy, BVSc, MSSc, PhD, MRCVS1,
  2. P. Upton, BSc Hons2,
  3. I. M. Higgins, Royal Statistical Society (RSS) Higher Certificate3,
  4. G. McGrath, BA(Mod)Sci, MSc3,
  5. A. V. Goodchild, BVSc, MPhil, PhD, MRCVS2,
  6. S. J. Rolfe, BSc, BVetMed, PhD, MRCVS4,
  7. J. M. Broughan, BA Hons, MSc, PhD2,
  8. S. H. Downs, BSc Hons, MSc, PhD2,
  9. R. Clifton-Hadley, MA, VetMB, MSc, PhD, MRCVS2,
  10. F. D. Menzies, BVM&S, MSc, PhD, DipECVPH, MRCVS5,
  11. R. de la Rua-Domenech, DVM, PhD, DipECVPH, MRCVS6,
  12. M. J. Blissitt, BVSc, CertSHP, PhD, MRCVS7,
  13. A. Duignan, MVB (Dublin), MRCVS8 and
  14. S. J. More, BVSc, MVB, DipPM, PhD, MACVSc, FACVSc, DipECVPH, DipECBHM3
  1. 1Department of Veterinary Tropical Diseases, University of Pretoria, Private Bag X04, Onderstepoort 0110, South Africa
  2. 2Animal Health and Veterinary Laboratory Agency, Weybridge, New Haw, Addlestone, Surrey KT15 3NB, Great Britain
  3. 3Centre for Veterinary Epidemiology and Risk Analysis, UCD School of Veterinary Medicine, University College Dublin, Belfield, Dublin 4, Ireland
  4. 4Office of the Chief Veterinary Officer, Welsh Government, Cathays Park, Cardiff CF10 3NQ, Great Britain
  5. 5Veterinary Epidemiology Unit, Department of Agriculture and Rural Development, Dundonald House, Upper Newtownards Road, Belfast BT4 3SB, Northern Ireland, UK
  6. 6Animal Health and Veterinary Laboratories Agency, Tuberculosis Programme of the Department for Environment, Food and Rural Affairs, Nobel House, London SW1P 3JR, Great Britain
  7. 7Veterinary and Science Team, Rural and Environmental Directorate, Scottish Government, Saughton House, Broomhouse Drive, Edinburgh EH11 3XD, Great Britain
  8. 8Department of Agriculture, Food and the Marine, Backweston, Celbridge, Co. Kildare, Ireland
  1. E-mail for correspondence: Darrell.Abernethy{at}up.ac.za

Abstract

Selected demographic features and trends in bovine tuberculosis (BTB) from 1995 to 2010 are described for the countries of the UK and the Republic of Ireland, using standardised definitions and measures. All countries experienced a reduction in the number of cattle and herds and in the proportion of dairy herds, while average herd size increased. In general, the trends indicate a stable situation of very low BTB prevalence in Scotland and, over most of the period, a rising prevalence in England and Wales. The prevalence in the Republic of Ireland declined while Northern Ireland experienced both a rise and fall. Differences in demography, BTB programme structure and test results were noted, particularly between the island of Ireland and Great Britain. Further investigation of these differences may provide valuable insights into risk factors for BTB and optimisation of existing BTB programmes.

View Full Text

Statistics from Altmetric.com

Introduction

Bovine tuberculosis (BTB) is an infectious disease of zoonotic potential and worldwide distribution caused by Mycobacterium bovis. Cattle are the principal host species, but the organism has been isolated from a wide range of domestic and wild animals (O'Reilly and Daborn 1995, Delahay and others 2002). In Europe, the Eurasian badger (Meles meles), wild boar (Sus scrofa scrofa) and several species of wild deer act as wildlife reservoirs, with implications for BTB eradication in domestic livestock (More and Good 2006, Delahay and others 2007, More 2009, Gortázar and others 2011).

Compulsory national eradication programmes commenced in the UK and the Republic of Ireland in the 1950s. Following entry of both countries into the European Union (EU) in 1973, EU legislation, principally 64/432 EEC and 78/52 EEC, determined the surveillance and control measures employed. Historically, BTB control policy within the UK was determined separately for Great Britain (encompassing the countries of England, Scotland and Wales) and Northern Ireland (which, with the Republic of Ireland, is located on the island of Ireland; Fig 1). This led to differences in strategy and management. Devolution of responsibility for agriculture to the Welsh and Scottish administrations in 1999 further regionalised approaches, while policy was developed independently in the Republic of Ireland. Thus, although programme measures were largely standardised through European legislation, differences have developed as a consequence of politics (UK, Republic of Ireland), geography (Great Britain, island of Ireland), epidemiological features and/or risk factors for BTB. Such differences provide an opportunity to learn from varying experiences and gain possible insight into factors affecting progress towards eradication.

FIG 1:

The UK and the Republic of Ireland illustrating political boundaries and (veterinary) administrative regions (England: N, north; W, west; E, east; Republic of Ireland: NW, northwest; SW, southwest; NE, northeast; SE, southeast)

Over time, a number of different measures have been developed to evaluate progress in the national BTB eradication programmes in the five countries, reflecting historical differences in policy, operation and epidemiological understanding. To date, however, these differences have limited the potential for intercountry comparison. In this paper, we describe the development and application of common measures of progress, calculable in all countries, to provide a composite and comparable description of BTB trends, from 1995 to 2010. Given the complexity of BTB and these country-level differences, an examination of trends provides a valuable but superficial insight into factors that affect eradication. Each country has developed a considerable knowledge base underpinning its BTB programme; an understanding of trends will help to place this country-level knowledge in a broader context.

Materials and methods

BTB eradication scheme

Under European legislation, the frequency of routine surveillance is determined by the prevalence of infected herds (Table 1). In Scotland, the only officially BTB-free (OTF) part of the UK, quadrennial herd testing has been ongoing for at least 15 years. By contrast, annual herd testing has been carried out in the Republic of Ireland and Northern Ireland for many years, and in Wales since October 2008.Test frequency in England varies from annual testing in the west (representing nearly half of all English herds) to quadrennial testing in the north and east, while those in the intermediate buffer zone are tested every two years.

TABLE 1:

Bovine tuberculosis surveillance required by EC Directive 64/432

Routine surveillance is conducted with the single intradermal comparative cervical tuberculin test using avian (2500 IU per dose) and bovine (3000 IU per dose) tuberculins manufactured by Lelystad. Cattle with an increase of greater than 4 mm at the bovine injection site after 72 hours compared with any increase at the avian site are compulsorily removed as test reactors. Clinical symptoms suggestive of BTB or evidence of pain, heat or oedema at the bovine site may also result in a positive interpretation of the test. Under certain conditions, in particular, confirmation of infection in previous reactors, the test threshold is reduced (‘severe interpretation’) to 2 mm and the ­cattle removed, otherwise they are individually restricted, isolated from the herd, and retested 60 days later. Herds from which reactors are removed lose their OTF status and may not trade in cattle until completion of one or two negative herd-level tests at minimum intervals of 42–60 days. If BTB is confirmed (presence of macroscopic lesions characteristic of BTB at slaughter, positive histopathology or positive culture results), the OTF status is withdrawn (OTW) and the herd must complete two successive negative tests. In Great Britain, should infection not be confirmed, the OTF status is suspended and only one further negative herd test is required. Northern Ireland has a similar policy, except that outbreaks with more than five unconfirmed reactors are treated as OTW, while almost all outbreaks in the Republic of Ireland are considered as OTW except for a small proportion (0.06) that are released early following an epidemiological risk assessment and laboratory analysis (Good and Duignan 2011). In all countries, a six-month post-outbreak test is undertaken following reinstatement of the OTF status.

Skin testing is undertaken by veterinary surgeons, mostly by those in private practice, although testing by trained technical officers in government employment has been underway in Great Britain since 2005. Abattoir surveillance is undertaken of all cattle, either by private veterinary surgeons (Republic of Ireland), by meat inspectors employed by the Department of Agriculture and Rural Development (DARD; Northern Ireland) or, in Great Britain, by meat inspectors employed by the Meat Hygiene Service (until April 2010, after which it merged with Food Standards Agency). The inspection protocols are standardised through European legislation, with samples of macroscopic lesions submitted for laboratory confirmation or from a pool of lymph nodes if no lesions are detected. Where BTB is confirmed in cattle slaughtered as part of routine farm production (so-called ‘abattoir cases’), the herd follows the same restriction and test regimen as one with confirmed BTB reactors.

Data sources and analysis

Cattle population data in Great Britain were extracted from the Department for Environment, Food and Rural Affairs (Defra) statistics, while testing and post-mortem information were obtained from the Animal Health and Veterinary Laboratory Agency (AHVLA) VetNet database, supplemented by the AH TB Culture System (both superseded in 2011). Data from the British Cattle Movement Service's Cattle Tracing Scheme were utilised to inform herd size and existence. In Northern Ireland, demographic data were obtained from the Agricultural Statistics ­produced by DARD (http://www.dardni.gov.uk/index/dard-statistics.htm), while slaughter data were provided by the DARD Statistical Branch. Cattle test data were extracted from the Animal and Public Health Information System. In the Republic of Ireland, BTB data were obtained from a national database (Nixdorf, prior to February 2005, and the Animal Health Computer System (AHCS) subsequently). Movement data were obtained from the animal identification and movement database (the Cattle Movement Monitoring System (CMMS) prior to 2007, and the generic Animal Identification and Movement System (AIM) subsequently).

Summary BTB statistics were compiled using agreed, standardised case definitions, as defined in the Glossary. Only herds considered to be actively holding or trading cattle (‘active’ herds) were included in summary measures, but each country used a slightly different means to define these units. These, and other differences, were explicitly considered at a series of workshops throughout the lifetime of the project, with adjustments made to ensure differences were minimised, and to consider the effect of outstanding discrepancies. It is therefore possible that summary statistics provided here will differ from official statistics published by each country. Spatial, demographic and management data used 2009 as the study endpoint, whereas, summary disease measures were calculated up to the end of 2010.

Annual herd prevalence was defined as the annual percentage of active herds with at least one reactor or abattoir case. Annual herd incidence was defined as follows: the denominator included all active herds that were BTB-free on 1 January of the year of interest, and the numerator a subset of these herds with at least one reactor or abattoir case in the following 12 months. A confirmed breakdown was one with at least one confirmed reactor, an abattoir case or at least two unconfirmed reactors. The latter criterion, which is not normally used in the UK, was incorporated to standardise differences between the UK and the Republic of Ireland in the extent to which bacteriological culture is used to determine the OTW status. Herd size is a known risk factor for BTB (Denny and Wilesmith 1999, Olea-Popelka and others 2004, Reilly and Courtenay 2007); accordingly, direct standardisation was used to adjust for varying herd size (Dohoo and others 2003). Annual stratum-specific incidence and prevalence risks were calculated for each country across six strata of herd size and then applied to the reference population, which comprised the sum of cattle populations across all countries. The standardised population was then used for herd-level disease measures, resulting in a standardised herd prevalence (SHP) and standardised herd incidence (SHI). These provide a comparable, relative measure of BTB, but they differ from the unadjusted measures reported previously by individual countries. As the frequency of testing herds in Great Britain varied by region, incidence and prevalence measures in that country are presented separately for herds on an annual testing schedule and for all herds, respectively.

The mapping procedure in this study was built on a methodology established in an earlier collaborative exercise between the Republic of Ireland and Northern Ireland (McGrath and others 2009). Great Britain has a separate National Grid projection to that of the island of Ireland, thus, a map of the latter was transformed into World Geodetic System 84 (WGS84) format and then reprojected into British National Grid format. This caused a very slight distortion in the west/northwest of Ireland and, for this reason, a north arrow and graticule reference are not included in the maps.

For aggregated spatial data, a uniform surface of 1542 hexagons, each with an effective diameter of 20 km, was created over the area of Great Britain and the island of Ireland, using a custom script in ArcGIS 9.2 (ESRI, Redlands, California, USA). All farm points were then associated with the hexagon in which they were contained using a point in polygon analysis. Thereafter, animal and herd-level data for all farms were summarised for each hexagon, enabling the creation of thematic maps.

Maps were created to enable a comparison between two time periods, 1997 and 2008. However, to include all herds in Great Britain, where test intervals varied by region from one to four years, a second series of maps was created, using data spanning two years either side of these dates (ie, 1995–1998 and 2006–2009), in herds active on January 1, 1997 or 2008, respectively. This ensured that all herd test results were represented, regardless of test interval. In summarising these multiple-year time periods, the definitions of what the data represent were slightly different to the single-year maps. Animal-level prevalence for the multiple-year maps was calculated per hexagon by dividing the sum of BTB-positive animals for the four-year period by the sum of the mean tested herd size, based on whole-herd tests, for the same period. The value is not a simple prevalence because the population (the denominator) is based on a mean for the time period, whereas the BTB-positive animals (the numerator) are summed for the period. The herd-level prevalence for the multiple-year maps was calculated as the percentage of herds in each hexagon with at least one BTB-positive animal in the four-year period. This is referred to as the ‘cumulative herd risk’. The population per hexagon for the multiple-year maps was based on mean herd size over each four-year period. The reactors per hexagon for the multiple-year maps were based on the sum of reactors for each four-year period. The BTB-positive herds per hexagon in multiple-year maps were based on a count of herds with at least one BTB-animal test over the four-year period.

Data were not available for the number of cattle tested in the Republic of Ireland between 1995 and 2004, as only the total number of tests was recorded prior to 2005. The mean number of tests per animal was therefore calculated for the years 2005–2009 and extrapolated backwards.

Results

Cattle demographics

The cattle population decreased across all countries by 15.6 per cent between 1995 and 2009 (range: −5.9 per cent in Northern Ireland to −18.7 per cent in England) and the number of active herds by 21.1 per cent (range: −11.5 per cent in Scotland to −24 per cent in the Republic of Ireland). Overall, 17.1 per cent of herds were classified as dairy in 2009 (range 11.0 per cent in Scotland to 22.0 per cent in England) but there was a 44.1 per cent reduction in the number of dairy herds during the study period (range: −32.5 per cent in Scotland to −53.1 per cent in the Republic of Ireland). The average herd size increased by 13.7 per cent (range: 3.8 per cent in England to 44.2 per cent in Scotland; Table 2). The island of Ireland was characterised by smaller herds than Great Britain, with 62 per cent having 50 or fewer cattle in 2009 ­compared with 49 per cent, respectively (Fig 2), and substantially higher ­densities of cattle and herds (Figs 3 and 4). Conversely, the number of herds with more than 500 cattle was eight-fold greater in England or Scotland compared with the island of Ireland. There have been only slight changes in cattle and herd densities from 1995 to 2009 (Figs 3 and 4).

FIG 2:

Distribution of herds by jurisdiction and herd size, 1995 and 2009 (EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; 95, 1995; 09, 2009)

FIG 3:

Herd density: 1995–1998 (left) and 2006–2009 (right)

FIG 4:

Cattle density 1995–1998 (left) and 2006–2009 (right)

TABLE 2:

  Selected demographic features

Testing

In England, herd test coverage increased from 35.2 per cent (1995) to 57.3 per cent (2010; Fig 5), associated with an increase in the annual total of tests from 1.8 million (1995) to 5.5 million (2010; Fig 6). The proportion of herds tested annually increased from 18.7 per cent to 63.9 per cent (data not shown), with the greatest regional increase in the west (48.2 per cent in 1995 to 90.7 per cent in 2010; Fig 7). The number of tests per bovine increased from 0.3 to 1.0 over the same period (Fig 8).

FIG 5:

Herd test coverage. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; shading represents duration of foot-and-mouth disease epidemic

FIG. 6:

Annual total (millions) of tuberculin tests. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; shading represents duration of foot-and-mouth disease epidemic

FIG. 7:

Percentage of unrestricted herds in England subject to field surveillance, by region, 1995–2010; shading represents duration

of foot-and-mouth disease epidemic
FIG 8:

Number of tests per head of cattle population. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; shading represents duration of foot-and-mouth disease epidemic

In Wales, herd coverage increased from 30.5 per cent (1995) to 92.5 per cent (2010), the number of tuberculin tests from 0.3 million to 1.9 million per annum (Figs 5 and 6), and the proportion of herds scheduled for annual testing from 3.1 per cent to 100 per cent (data not shown).

Testing reduced slightly in Scotland, both in herd coverage (28.4 per cent in 1995 to 22.8 per cent of herds in 2010) and in number of tests (0.25–0.24 million tests; Figs 5 and 6).

Herd test coverage averaged 89.2 per cent and 93.7 per cent over the study period in Northern Ireland and the Republic of Ireland, respectively, with a mean of 2.4 and 9.7 million tests per annum, respectively, and a mean of 1.4 tests per head of cattle population (Figs 5, 6 and 8).

The number of herds tested in Great Britain and Northern Ireland in 2001 compared with 2000 reduced by 68.2 per cent and 20.5 per cent, respectively, as a result of disruption to the BTB programme ­during the 2001 foot-and-mouth disease (FMD) epidemic (February 2001 to October 2002 in Great Britain).

Herd-level descriptive statistics

The SHP in England increased from 0.8 per cent in 1995 to 9.0 per cent in 2010 (range = 0.8 per cent (1995) to 9.1 per cent (2009); unadjusted herd prevalence = 1.1 per cent and 10.8 per cent, ­respectively), and the SHI from 0.8 per cent to 5.4 per cent (range = 0.7 per cent (2001) to 5.6 per cent (2008); unadjusted herd incidence = 1.1 per cent and 6.3 per cent, respectively; unadjusted confirmed herd incidence = 0.7 per cent and 4.8 per cent, respectively; Figs 913). The annually tested parishes in England and Wales experienced a two-fold increase in herd incidence from 2000 to 2002, which then peaked in 2003 and again in 2009 before reducing in 2010. At a regional level, the rise in BTB in England was significantly higher in the west than elsewhere (Figs 14 and 15).

FIG 9:

Standardised annual herd prevalence. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; Ann, annually tested regions; 2010 data for annually tested region in Wales omitted as all herds scheduled for annual test; shading represents duration of foot-and-mouth disease epidemic

FIG. 10:

Unadjusted annual herd prevalence. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; Ann, annually tested regions; 2010 data for annually tested region in Wales omitted as all herds scheduled for annual test; shading represents duration of foot-and-mouth disease epidemic

FIG 11:

Standardised annual herd incidence. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; Ann, annually tested regions; 2010 data for annually tested region in Wales omitted as all herds scheduled for annual test; shading represents duration of foot-and-mouth disease epidemic

FIG 12:

Unadjusted, annual herd incidence. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; Ann, annually tested regions; 2010 data for annually tested region in Wales omitted as all herds scheduled for annual test; shading represents duration of foot-and-mouth disease epidemic

FIG 13:

Unadjusted confirmed annual herd incidence. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; Ann, annually tested regions; 2010 data for annually tested region in Wales omitted as all herds scheduled for annual test; shading represents duration of foot-and-mouth disease epidemic

FIG 14:

Cumulative herd risk, 1995–1998 (left) and 2006–2009 (right)

FIG 15:

Annual herd incidence in England by region. Shading represents duration of foot-and-mouth disease epidemic

The SHP in Wales increased from 0.9 per cent in 1995 to 11.6 per cent in 2010 (range = 0.9 per cent (1995) to 13.7 per cent (2009); ­unadjusted herd prevalence = 1.0 per cent and 13.2 per cent, respectively) and the SHI from 0.8 per cent to 7.1 per cent, respectively (range = 1.0 per cent (1995) to 8.2 per cent (2009); unadjusted herd incidence = 1.0 per cent and 7.8 per cent, respectively; unadjusted confirmed herd incidence = 0.7 per cent and 5.4 per cent, respectively; Figs 913).

In Scotland, both the SHP and SHI remained level, with a mean SHP of 0.3 per cent (sd = 0.100; unadjusted mean herd prevalence = 0.5 per cent) and a mean SHI of 0.2 per cent (sd = 0.076; unadjusted mean herd incidence = 0.3 per cent; Figs 913). The mean confirmed herd incidence across the period was 0.10 per cent (range 0.05 per cent (1998) to 0.25 per cent (2002); Figs 913).

The SHP in the Republic of Ireland decreased from 9.6 per cent in 1995 to 7.4 per cent in 2010 (range over the period = 7.4 per cent (2010) to 11.7 per cent (2000); unadjusted herd prevalence = 8.4 per cent and 6.5 per cent, respectively) and the SHI from 6.3 per cent in 1995 to 5.0 per cent in 2010 (range = 5.0 per cent (2010) to 8.1 per cent (2000); unadjusted herd incidence = 5.6 per cent and 4.4 per cent, respectively; unadjusted confirmed herd incidence = 3.5 per cent and 2.4 per cent respectively; Figs 913). The frequency of infected herds was highest in the northeast and southeast, particularly in the early part of the period, but decreased more strongly in these areas towards the end, compared with the southwest and northwest (Figs 14 and 16).

FIG 16:

Annual herd incidence in the Republic of Ireland by veterinary administrative region. NW, northwest; SW, southwest; NE, northeast; SE, southeast

Northern Ireland experienced a rise in both herd prevalence and incidence up to 2002, a two-year plateau, and thereafter a decline, but infection levels were higher at the end of the study period than in 1995 (SHP in 1995 = 5.5 per cent, in 2010 = 7.9 per cent; range = 5.5 per cent (1995) and 14.3 per cent (2004); unadjusted herd prevalence = 4.9 per cent and 7.0 per cent, respectively; SHI = 5.4 per cent and 6.8 per cent, respectively; range = 5.4 per cent (1995) and 11.7 per cent (2002); unadjusted herd incidence = 4.4 per cent and 6.0 per cent, respectively; unadjusted confirmed herd incidence = 3.5 per cent (1995), 8.9 per cent (2002), 4.8 per cent (2010); Figs 913).

Animal-level descriptive statistics

The proportion of the total cattle population with BTB disclosed (apparent animal prevalence) in England increased from 0.04 per cent in 1995 to 0.46 per cent in 2010; in Wales, it increased from 0.04 per cent in 1995 to 0.83 per cent in 2008 before decreasing to 0.61 per cent by 2010. The prevalence in Scotland varied from 0.002 per cent to 0.009 per cent between 1995 and 2001 (mean = 0.005 per cent), and from 0.009 per cent to 0.004 per cent between 2002 and 2010 (mean = 0.008 per cent). The prevalence in the Republic of Ireland varied from 0.36 per cent (2004) to 0.63 per cent (1999), but decreased over the study period (0.50 per cent in 1995 to 0.37 per cent in 2010). Northern Ireland experienced an increase between 1995 and 2003 (0.24 per cent and 1.01 per cent, respectively) followed by a decrease, to 0.46 per cent, by 2010. The geographic distribution of infected animals (test reactors and abattoir cases) was similar to that observed at herd level (Figs 1719).

FIG 17:

Apparent animal annual prevalence. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; shading represents duration of foot-and-mouth disease epidemic

FIG 18:

Cumulative apparent animal prevalence, 1995–1998 (left) and 2006–2009 (right)

FIG 19:

Reactor distribution, 1997 (left) and 2008 (right)

The trends in annual test reactor incidence were similar to the apparent animal prevalence: England and Wales experienced a gradual increase prior to 2001, a marked increase in 2001, and a fluctuating trend thereafter until 2008, after which the incidence declined (England: 0.15 per cent (1995), 0.73 per cent (2010); Wales: 0.21 (1995), 0.61 per cent (2010)). The incidence reduced across the study period in the Republic of Ireland (0.43 per cent (1995) and 0.29 per cent (2010)), while Northern Ireland experienced an increase between 1995 and 2003 (0.23 per cent and 0.95 per cent, respectively) before it decreased to 0.40 per cent by 2010 (Fig 20). Between 2007 and 2009, 40 per cent of skin test reactors in Wales were removed with a net bovine rise of 4 mm or less (Fig 21); equivalent data for the other ­countries were 34 per cent (Scotland), 28 per cent (England), 30 per cent (Republic of Ireland) and 20 per cent (Northern Ireland).

FIG 20:

Annual test reactor incidence. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; Ann, annually tested regions; 2010 data for Wales omitted as all herds scheduled for annual test; shading represents duration of foot-and-mouth disease epidemic

FIG 21:

Frequency distribution of test reactors, 2007–2009, by net bovine rise. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; Comp, all of country; Ann, annually tested regions

Abattoir surveillance

The prevalence of abattoir cases, namely the percentage of cattle slaughtered with BTB-like lesions was higher on the island of Ireland than in Great Britain, particularly at the start of the study period (mean of 0.09 per cent and 0.001 per cent (1995); 0.11 per cent and 0.04 per cent (2010), respectively; Fig 22). In England, the prevalence increased from 0.001 per cent in 1995 to 0.08 per cent in 2010; Wales experienced an increase from 1995 (0.001 per cent) to 2008 (0.05 per cent), after which it reduced slightly (0.03 per cent in 2009). The prevalence in the Republic of Ireland varied around a mean of 0.12 per cent (sd = 0.012), while Northern Ireland experienced an increase from 1995 (0.07 per cent) to 2002 (0.27 per cent) before it reduced to 0.12 per cent, by 2010.

FIG 22:

Annual abattoir case prevalence. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; shading represents duration of foot-and-mouth disease epidemic

Prior to 2001, abattoir surveillance detected up to 25 per cent of new BTB breakdowns in the Republic of Ireland and Northern Ireland, but the trends diverged thereafter, with an increase in the former, peaking at 36 per cent (2006), while reducing in Northern Ireland to a low of 14 per cent in 2004 (Fig 23). The patterns in England and Wales were similar, with an increasing proportion detected at slaughter in both countries, up to 2001 (1.9 per cent and 1.3 per cent (1995); 17.1 per cent and 11.4 per cent (2001), respectively), but a divergence after 2004. By 2010, 15.2 per cent and 6.6 per cent of new breakdowns, respectively, were detected through such surveillance. Over the period, abattoir surveillance in the annually tested areas detected slightly fewer breakdowns than the country as a whole (mean difference = 1 per cent in England and 1.4 per cent in Wales). In Scotland, such surveillance detected no outbreaks in the four years prior to 2002, and an increasing percentage thereafter, to peak at 18.6 per cent in 2008. The trends at animal level were similar to herd-level measures (Fig 24).

FIG 23:

Percentage of TB-positive herds, unrestricted on January 1 each year, detected by abattoir inspection. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; Ann, annually tested regions; shading represents duration of foot-and-mouth disease epidemic

FIG 24:

Percentage of TB-positive animals detected by abattoir inspection. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; Ann, annually tested regions; shading represents duration of foot-and-mouth disease epidemic

Post-outbreak surveillance

In Great Britain, the proportion of herds positive at the six-month test increased between 1995 and 2001, from 11.0 per cent to 23.3 per cent in England, and from 10.6 per cent to 25.2 per cent in Wales. Thereafter, it varied between 18 per cent and 28 per cent. Trends in the annually tested parishes were generally similar but slightly higher than the whole-country patterns (Fig 25). In the Republic of Ireland, the trend was level, varying around a mean of 12.2 per cent (sd = 1.04). Northern Ireland experienced an increase from 1995 (6.2 per cent) to 2002 (13.4 per cent), before it declined (mean from 2003 to 2009 = 10.1 per cent). Scotland experienced an irregular pattern (range 0–13 per cent), due largely to the small number of herds involved.

FIG 25:

Percentage of positive tests at six-month post-outbreak test. EN, England; SC, Scotland; WA, Wales; NI, Northern Ireland; IE, Republic of Ireland; Ann, annually tested regions; shading represents duration of foot-and-mouth disease epidemic

Discussion

BTB was widespread in Europe at the start of the 20th century with animal prevalence exceeding 30 per cent in some regions (Francis 1947) and tuberculous-like lesions detected in 31 per cent to 40 per cent of cattle slaughtered in British and Irish public abattoirs (Grange and Yates 1994, Cotter and others 1996). Compulsory, state-­sponsored programmes successfully eradicated the disease from some EU member states (Reviriego Gordejo and Vermeersch 2006) and significantly reduced it in the UK and the Republic of Ireland, to an animal prevalence of less than one per cent by the mid-1960s (Abernethy and others 2006, More and Good 2006, Reynolds 2006), and a mean of 0.4 per cent apparent animal prevalence by 2010 (Fig 17). The zoonotic risk is extremely low, despite recent increases in bovine prevalence (de la Rua-Domenech 2006, Jalava and others 2007). By contrast, approximately 4000 human cases and 2000 deaths per annum in the UK were attributed to M bovis infection during the 1920s (Jordan 1933). Scotland has been officially tuberculosis-free since October 2009, endemic BTB in Great Britain is largely confined to Wales and west England (Fig 18), and disease levels in the Republic of Ireland, which have been stable for many years (More and Good 2006), have declined during the study period. Considerable progress has therefore been achieved, but at substantial expense, with the 2010/2011 costs to government exceeding £227 million (£152 million in Great Britain (S Rolfe, personal communication), £23 million in Northern Ireland (Anon 2011b) and ? 63 (£52) million in the Republic of Ireland (R Healy, personal communication)). Despite such investment, eradication outside of Scotland appears elusive, prompting questions regarding contributory factors, programme value and requisite actions to expedite further progress.

The purpose of this study was to provide comparative data on BTB in cattle over an extended period from countries in the same geographical region and with broadly similar BTB eradication programmes. All the countries in this study utilised the same surveillance and diagnostic procedures, implemented common European legislation, and followed similar control measures on disclosure of infection. However, there were some notable differences. Cattle and herd densities on the island of Ireland were markedly higher, herds were smaller, and there were relatively fewer dairy herds than in Great Britain. In Great Britain, cattle and herd densities were highest in the west than elsewhere, while the largest mean herd size, but lowest proportion of dairy herds, was recorded in Scotland. These factors may influence the epidemiology of infection; for example, disease transmission is likely to be influenced by both cattle and herd densities. Direct standardisation was used to adjust for the effect of herd size, which provided a more meaningful comparison of trend data than the unadjusted parameters. However, as the reference population is contrived, the standardised data provide only a relative measure of BTB.

The frequency of routine surveillance differed between countries, from quadrennial testing in Scotland, through varying intervals in England, to annual testing in the Republic of Ireland, Northern Ireland and, recently, in Wales. Several statistical methods were attempted to allow direct comparison of trends after adjusting for these, but they were not successful due to data constraints. Accordingly, trends in England and Wales are displayed at two levels: first, a composite measure included all data, but results in an underestimation of BTB in England and Wales compared with the island of Ireland, and then one describing only the annually tested areas, which overestimates the true level.

A further notable difference between countries lay in actions to mitigate the risk from badgers, a known wildlife reservoir for BTB. Badger culling has not been undertaken in Scotland or in Northern Ireland. Various strategies were followed in England and Wales from the mid-1970s to 1997, after which culling was limited to a randomised field trial carried out between 1998 and 2005 in 10 defined areas of high BTB herd incidence in the west of England (Donnelly and others 2007). Targeted (reactive) badger culling has been part of the national eradication programme in the Republic of Ireland since 2000 (Good and others 2011) but is not undertaken in Great Britain following results of the randomised field trial (Donnelly and others 2003, Jenkins and others 2010).

Compulsory premovement skin testing of cattle moved off annually and biennially tested herds was introduced in Great Britain in 2005–2006 to mitigate the risk of spread from the high BTB prevalence areas. This policy has been augmented by postmovement testing of cattle moving from those areas into Scotland (Blissitt 2006). By contrast, premovement testing has not been used on the island of Ireland and was not found to be cost-effective in the Republic of Ireland (Clegg and others 2008). Application of the ancillary interferon-gamma test has varied widely, as has the proportion of reactors removed under severe interpretation to the skin test. These differences mean that any comparisons of statistical trends and spatial distribution of BTB between these countries are necessarily crude, and caution is required in attributing causality to any specific factor.

In general, BTB trends suggest a stable situation of extremely low prevalence in Scotland, a higher but reducing prevalence in the Republic of Ireland and, up to 2009, rising trends in England, Wales and Northern Ireland. The increase (up to 280 per cent; Fig 9) in the latter three countries in 2002 was associated with curtailment of the programme in 2001, arising from measures introduced to mitigate the risk of FMD. BTB testing was markedly reduced across the UK during 2001 (Figs 58) and cattle movements restricted, both between herds and between separate premises used by the same herd. It is likely that constraints on within-herd movement resulted in higher stocking densities locally, and greater within-herd spread of BTB during that period (D Abernethy, T Goodchild, unpublished observations). The Republic of Ireland, with a single FMD outbreak in 2001 (Griffin and O'Reilly 2003), and thus less impact on the BTB programme, experienced a reduction in BTB during that period (Figs 912). The mechanisms for the increased levels of BTB post-FMD in Great Britain and Northern Ireland remain unclear, as do reasons for the decline in BTB after 2004 in Northern Ireland but a longer-term increase in England and Wales. Cox and others (2005) concluded the effect of the FMD-related ­measures should be very low after five years; this appears to have occurred in Northern Ireland, which was not included in the analysis, but not in Great Britain. A detailed comparison of changes in the programmes in Great Britain with that of Northern Ireland in the years 2002–2006 may provide valuable insights into the benefits of remedial measures introduced after the FMD epidemic, as well as the relative roles of cattle and badgers in those years. Similarly, detailed comparisons between the Republic of Ireland and Northern Ireland may also provide useful insights as these two countries are likely to be very similar in terms of other factors that might influence BTB control.

Irrespective of the putative effect of the FMD epidemic on UK BTB patterns, the current trends appear to be superimposed on an underlying, long-term increase in BTB in the UK, present since 1986 (Abernethy and others 2010, Anon 2011a). Reasons for this rise are unclear, but may be associated with increases in the size (Table 2) and complexity of cattle herds, changes in the badger population (Krebs and others 1997, Wilson and others 1997), and cattle movements associated with restocking of cattle farms post-FMD (Gopal and others 2006). Robinson (2006) described significant increases in the beef-cow herd in Northern Ireland between 1985 and 1998, associated with farm subsidies, and argued these were contributory factors to the increased BTB prevalence at that time.

Abattoir case prevalence was substantially higher on the island of Ireland than in Great Britain (Fig 22), as was the relative importance of abattoir surveillance in detecting new outbreaks (Figs 23 and 24). This may reflect differences in background risk or in the relative sensitivities of the surveillance measures employed by each country. Comparable studies on submission and confirmation rates from attested animals at slaughter will provide insights into the latter. The differences in abattoir case prevalence may also arise due to the regionalised nature of disease in Great Britain, whereby large numbers of animals slaughtered from low-incidence areas in Great Britain will reduce the prevalence. Frankena and others (2007) and Olea-Popelka and others (2012) reported substantial variation between abattoirs in the Republic of Ireland in the detection of BTB lesions, a feature also observed in Great Britain (J Broughan, unpublished observations). The sensitivity of abattoir surveillance for BTB and the epidemiological significance of such cases also merit further investigation as they account for a substantial proportion of outbreaks (up to 35 per cent in the Republic of Ireland; Fig 23) but most (>80 per cent) of the source herds have little further evidence of infection (Olea-Popelka and others 2008). By contrast, 55 per cent of all breakdowns disclosed by abattoir surveillance in Great Britain between 2001 and 2009 had at least one reactor at subsequent check tests (J Broughan, unpublished observations).

Recrudescence of infection in herds following an outbreak was substantially higher in Wales and England than on the island of Ireland (Fig 25). This might reflect a higher risk of post-outbreak exposure in Great Britain, or greater residual infection, whereby infected cattle are not all disclosed during the restriction period.

This study raises potentially important issues concerning the purpose and functioning of the current BTB programmes in the UK and Republic of Ireland. The sharp increase in BTB in the UK following the suspension of herd testing in 2001 suggests a strong role for the programme in reducing prevalence or maintaining it at low levels. Conversely, the current levels of BTB suggest the programmes are not fully effective at dealing with these underlying factors, and that further research is required to identify and resolve these drivers. The contrasting trends observed between the countries raise challenging questions as to programme efficacy and the need to learn from each others' experiences. Therefore, several aspects require further consideration. First, the epidemiological basis of the programme should be reviewed to ensure it is fit for the purpose and able to adequately address risk factors for BTB. This will likely necessitate changes to legislation given the doubts expressed about the EU Directives underpinning the programmes (Abernethy and others 2010). Accordingly, recommendations to ensure that programmes are epidemiologically sound will require formal peer-reviewed studies to support recommendations for changes to legislation. Secondly, differences observed between the countries should be formally explored to ensure that programmes are optimal in both efficiency and effectiveness. The experience of this project has been that a multinational initiative, calling on the differing experiences and perceptions within each country, will best achieve this objective.

Glossary

Abattoir case
A bovine animal, slaughtered under routine farm management, in which BTB was suspected following detection of suspicious lesions at slaughter (Republic of Ire-land) and subsequently confirmed by laboratory testing (United Kingdom). The animal was sourced from a herd that was not currently under any movement restrictions due to prior disclo-sure of BTB.
Abattoir case prevalence
The number of abattoir cases disclosed each year, expressed as a percentage of the total number of cattle slaughtered.
Active herd
A herd considered to be actively holding or trading cattle during the study period. Defined as one that presented cattle for testing or through which cattle moved within one year either side of ascheduled test (Great Britain), one that presented cattle for TB testing in the previous year or through which cattle moved in the previous two years (Northern Ireland) or one in which cattle weretested during the year or otherwise within the fifteen months prior to 1st January of that year (Re-public of Ireland).
Apparent animal prevalence
Animal prevalence calculated as the sum of test reactors and abattoir cases as per-centage of the total cattle population.
Average herd size
Mean herd size calculated by dividing total cattle population by the number of ac-tive herds
BTB
“Bovine Tuberculosis”, namely, infection of cattle with Mycobacterium bovis.
BTB-positive animal
Animal removed from a herd following a positive reaction to the skin test (SICCT) or, in abattoir cases, one in which BTB was confirmed.
Confirmed herd
A herd in which at least one confirmed reactor, multiple unconfirmed reactors or at least one abattoir case was detected
Confirmed herd incidence
Percentage of active herds, unrestricted on 1st January, with at least confirmed test reactor or abattoir case during a defined period (one year of annual herd inci-dence).
Confirmed reactor
A reactor to the skin test (or Interferon Gamma test in United Kingdom) and in which TB-like lesions were detected at post mortem examination or from which tissue samples were positive on histology or bacteriology.
Cumulative herd risk
The herd level prevalence used in the multiple year maps and calculated as the percentage of herds in each hexagon with at least one BTB-positive animal in the four-year period.
Herd incidence
Percentage of active herds, unrestricted on 1st January, with at least one test reac-tor or abattoir case during a defined period (one year for annual herd incidence). See SHI for adjustments due to differences in mean herd size between countries.
Herd prevalence
Percentage of active herds with at least one reactor or abattoir case during a de-fined period (one year for annual herd prevalence). See SHP for adjustments due to differences in mean herd size between countries.
Herd test coverage
Percentage of active herds with a herd-level test each year.
Interferon gamma test(IFNg)
Diagnostic blood test that uses an enzyme-linked immunosorbent assay (ELISA) to detect interferon, a component of cell-mediated immune reactivity to M. bovis.
Non-standard re-actor
A bovine animal slaughtered as a reactor to the SICCT test, where the reaction to bovine tuberculin was 4mm or less than the reaction to avian tuberculin.
Positive herd
Herd with at least one standard reactor to the skin test (SICCT) or an abattoir case
Reactor
Bovine removed from the herd because of a positive BTB skin or blood test result.
Reactor incidence (Annual)
The number of cattle (by definition, susceptible and free of BTB at the start of the period) positive to the test each year, expressed as a percentage of the number of cattle tested.
Restricted herd
A herd from which cattle movements are prohibited due to presence of one or more test reactors or abattoir cases.
Risk surveillance tests
Skin testing undertaken outside the normal (“routine”) test fre-quency, in response to putative exposure of the herd to TB.
Restricted tests
Tests of cattle undertaken in restricted herds. Also known as control tests.
Routine surveil-lance tests
Herd-level skin tests undertaken at the usual interval prescribed by the BTB con-trol programme, in the absence of identified risk factors
Skin (SICCT) test
Single Intradermal Comparative Cervical Tuberculin Test.
Standard reactor
A bovine animal showing a positive response to the SICCT test,where the reaction to bovine tuberculin was more than4mm greater than the reaction to avian tuber-culin.
Standardised herd incidence (SHI)
As for herd incidence but adjusted for differences between countries in tested herd size by direct standardisation, using the sum of cattle populations across countries as the reference population.
Standardised herd prevalence (SHP)
As for herd prevalence but adjusted for differences between countries in tested herd size by direct standardisation, using the sum of cattle populations across countries as the reference population.
Tested herd size
Size of the herd as measured by the mean number of cattle tested during all whole-herd tests conducted in the year of interest.

Acknowledgments

The authors wish to acknowledge the assistance of the following colleagues who provided invaluable advice and assistance: Nigel Clarke (DARD, Northern Ireland), Margaret Good (DAFM, Republic of Ireland) and Robin Sayers (AHVLA, Great Britain).

References

View Abstract

Footnotes

  • Provenance: Not commissioned; externally peer reviewed

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.