This report, provided by the APHA, summarises the key descriptive epidemiological parameters of bovine TB in cattle in Great Britain from January 1 to December 31, 2014. It summarises some of the temporal trends observed over a longer period and highlights some differences and similarities between Scotland, Wales and the three bovine TB risk areas of England. It updates the previous annual summaries for 2012 and 2013, also published in Veterinary Record (VR, June 14, 2014, vol 174, pp 600-604; March 28, 2015, vol 176, pp 326-330).
Statistics from Altmetric.com
THE aim of publishing this summary of the annual Great Britain bovine TB surveillance reports is to provide a concise description of the epidemiology of TB in cattle in GB on an annual basis and to put the surveillance data into context. The incidence of bovine TB and other related statistics are measured using surveillance data collected by the APHA through its database known as SAM. Further description and analysis of the data is available in the full reports for 2014, which are publicly available on the UK government website (APHA 2015a, b, c).
The frequency and geographical distribution of bovine TB in England and Wales has been increasing since the mid-1980s. Control measures in cattle include routine and more targeted testing of cattle herds, slaughterhouse surveillance, premovement testing, movement restriction of infected herds and rapid slaughter of infected animals. Controlling the disease in wildlife reservoirs is also a key factor towards eradication in GB; the tools to achieve this include promoting good biosecurity, supporting badger vaccination in Wales and England, and industry-led, licensed badger culling in areas of England (Defra 2014).
A new incident (or breakdown) refers to bovine TB being disclosed in an unrestricted cattle herd within the reporting period; in this case, 2014. Herds are described by their Official Tuberculosis Free (OTF) status, which can be OTF-withdrawn (OTF-W), OTF-suspended (OTF-S) or, if free from restrictions, OTF.
A new OTF-W incident refers to the detection in an OTF herd during the reporting period of at least one test reactor or inconclusive reactor with postmortem evidence of Mycobacterium bovis infection (judged by the presence of visible lesions typical of bovine TB and/or identification of M bovis in culture), or at least one suspect slaughterhouse case that yielded M bovis on culture. An OTF-S incident refers to a bovine TB incident where evidence of infection could not be confirmed in any test reactor through postmortem examination or laboratory culture of tissue samples.
Some herds in Wales that would have been classified as OTF-S (because there is no supplementary postmortem evidence of infection) are classified as OTFW-2 because they are considered to have a high probability of infection according to defined epidemiological criteria. These are an additional cohort to the traditional OTF-W incidents described above and are managed in the same way as OTF-W incidents. OTFW-2 incidents contribute to OTF-W statistics in this analysis but it should be borne in mind when comparing OTF-W statistics across countries that the definition for these is different in Wales. Historical data from 2011 onwards, when the change in policy in Wales happened, has been updated in this report.
For disease surveillance and control purposes, three bovine TB risk areas were formally established in England on January 1, 2013, replacing the previous division by parish testing interval of one, two, three or four years. Cattle herds in the high risk area (HRA) and Edge area are tested annually and those in low risk areas (LRAs) are tested every four years. The HRA includes counties where bovine TB is considered endemic in cattle herds and badgers. The Edge area includes whole and part counties of England at greatest risk of becoming new endemic bovine TB areas, but where the incidence of the disease is still lower than in the HRA. The distribution of incidents disclosed in 2014 in England has been described according to these bovine TB risk areas.
All herds in Wales have been tested annually since 2010, following the testing of all herds over a 15-month period from December 2008 (as part of the Health Check Wales initiative). Scotland adopted risk-based testing in 2012 (Anon 2011). In Scotland, herds which meet certain risk criteria (for example, a large proportion of stock slaughtered, rarely purchasing from a high risk area, and herd sizes of less than 20) are exempt from the four-yearly routine herd testing (Anon 2011).
Number of new incidents
In 2014, there were 4617 new bovine TB incidents in GB, of which 3464 (75 per cent) were classified as OTF-W and 1147 (25 per cent) as OTF-S, with an additional six incidents yet to be classified at the end of the reporting period. There was little change in the total number of new incidents detected in GB (–3.3 per cent), which was mirrored in England (–3.8 per cent) and Wales (–3.3 per cent); however, in Scotland, the number of breakdowns increased from 23 in 2013 to 41 in 2014. The number of OTF-W incidents in GB was very similar to 2013 (n=3479) and there was an 11 per cent decrease in the number of OTF-S incidents overall, compared to 2013 (n=1284), although variation was recorded at country level (Table 2.1 in APHA 2015a).
In England in 2014, there were 3738 new bovine TB incidents (3292 in the HRA, 337 in the Edge area and 109 in the LRA) (Table ES1 in APHA 2015a). The total number of new bovine TB incidents (OTF-W and OTF-S combined) in the HRA and across the country decreased in 2014 compared to 2013 (–4.9 per cent and –3.8 per cent, respectively), although there was a 5.3 per cent increase in the number of new incidents in the Edge Area, and a 3.8 per cent increase in the LRA (Table 2.1 in APHA 2015b).
A smaller decrease in the total number of new bovine TB incidents was recorded in Wales, following a large decline in 2013 (–21 per cent; –3.3 per cent in 2014) (Table 2.1 in APHA 2015a).
In Scotland, OTF-W incidents increased from eight to 15. This increase can be partly attributed to tracing investigations following an outbreak of bovine TB in a large dairy herd in Cumbria, England, which had been dispersed and resulted in six new OTF-W incidents in Scotland. The number of OTF-S incidents increased by 66.7 per cent from 15 in 2013 to 25 in 2014. The number of cattle tests in Scotland increased by 10 per cent in 2014.
As expected, the vast majority of new bovine TB incidents in 2014 (Fig 1) were located in high risk, endemic areas of England and south-west and eastern Wales and the geographical distribution of new breakdowns was consistent with previous years (Broughan and others 2015).
A series of maps by four-yearly intervals from 1992 onwards in the GB report (Figs 2.4 a-g in APHA 2015a) shows the increase in density per km2 of animals reacting positively to the skin test or interferon gamma (γIFN) test and culture-positive slaughterhouse cases in OTF-W herds. The increase is mostly located in the endemic areas of east, south-east and south-west Wales and west and south-west England. In 2014, as in previous years, there was a substantial number of areas where the density was greater than one reactor per km2.
The reason for a bovine TB test and the proportion of breakdowns that are disclosed by such testing may help to indicate the relative epidemiological and economic benefits associated with the different types of test. Although routine skin testing detected just under a third of breakdowns, tests targeted towards individual herds or particular areas of risk detected over 40 per cent of breakdowns. Routine skin tests accounted for 31.6 per cent (n=2,337,933) of the total tests carried out in unrestricted animals and detected a similar proportion (30.5 per cent) of incidents in 2014 and 2013 (Table 2.2 in APHA 2015a). Seventeen per cent (n=1,266,211) of herd tests were carried out in animals perceived to be at an increased risk of bovine TB (‘herd risk’); for example, due to a recent bovine TB incident, and these detected 27.6 per cent of all new incidents. ‘Area risk’ tests include those for herds within 3 km of an OTF-W breakdown in the LRA of England and in the Edge area sections of Cheshire and Derbyshire (enhanced ‘radial’ testing) and for herds contiguous to OTF-W breakdowns elsewhere. Those tests detected 14 per cent of all incidents in 2014 and accounted for 13.5 per cent (994,968) of herd tests.
Compulsory pre- and postmovement testing detected 366 breakdowns in 2014, representing 7.9 per cent of all breakdowns (7.5 per cent OTF-W) in GB, 8.1 per cent of breakdowns in English herds and 7.3 per cent of breakdowns in Welsh herds. Two breakdowns were detected in Scotland by postmovement testing (both OTF-S).
Postmortem meat inspection of non-reactor cattle at routine slaughter accounted for a further 26.3 per cent of all unrestricted animal ‘tests’ (using the number of cattle slaughtered for non-TB related reasons in 2014) and accounted for 14.0 per cent of total new bovine TB breakdowns.
Herd incidence and prevalence
The incidence of bovine TB in cattle herds is expressed here as the rate of newly detected incidents per 100 herd-years at risk. Herd-years at risk is the total time that herds in the population are OTF and at risk of sustaining a new bovine TB incident. For each herd, the time at risk begins with a negative herd test, or the lifting of restrictions following a breakdown, and ends with the next herd test or detection of an incident occurring during the reporting period of interest (in this case, 2014). Herds can only contribute if they have been tested or inspected and only the time when they have not been under disease restrictions is taken into account in the denominator. If a herd has more than one herd test preceded by periods of time when the herd was unrestricted the time at risk is cumulated.
Other denominators used in the past to estimate incidence include ‘100 unrestricted herds tested’ and ‘100 live herds’, but these do not explicitly take the herd testing frequency into account (Appendix Table 3.1 in APHA 2015a, b). Gathering accurate information on how incidence varies across country or risk area (Wales, Scotland and English HRA, Edge and LRA), and the effect of herd size and herd type, assists in the development of targeted surveillance strategies.
Fig 2 shows the variation in the total number of new incidents per 100 herd-years at risk for the English surveillance risk areas, Wales and Scotland between 2003 and 2014. The bovine TB incidence rate in the English HRA has been relatively stable since 2011. The incidence rate in Wales has continued to decrease, albeit less steeply than in 2013. The incidence rate in the LRA and Scotland remains consistently low, and has been less than 0.6 incidents per 100 herd-years at risk since 2006. Overall, the bovine TB incidence rate for GB was 6.8 incidents per 100 herd-years at risk.
The number of counties with an incidence of more than four new breakdowns per 100 herd-years at risk was 22 in 2014 (Fig 3) compared to 19 in 2013 and 18 in 2012. Counties with incidence rates greater than 4.0 per cent are concentrated in the HRA of England and in Wales, although some counties straddling both the HRA and Edge also exceeded 4 per cent (Cheshire, Derbyshire, Warwickshire and Oxfordshire). High risk areas in Wales and the HRA in England continued to have areas with an incidence of more than 12 per cent and three counties in England exceeded 20 per cent (Devon, Gloucestershire and Wiltshire).
Herd prevalence is described in the reports as the proportion of live herds under restriction as a result of a bovine TB incident at the middle of each month (Figs 3.3a, b in APHA 2015a). Overall, in England, an average of 4.9 per cent of herds were restricted across the 12 months of 2014 and national prevalence has been very slowly declining since early 2012 (Fig 4). An average of 9.7 per cent of herds in the HRA of England were under restrictions each month as a result of a bovine TB incident in 2014. There has been a small decrease since 2012, although with cyclical trends it is too early to tell if the decrease will continue.
Duration refers to the number of days that a herd is under movement restrictions (that is, with its OTF status suspended or withdrawn) as a result of a bovine TB incident. Trends in the duration of restrictions are presented in the GB report (Appendix Fig 4.1 in APHA 2015a). Breakdowns of longer or shorter duration result from a number of factors that affect the time taken to have a clear test, or to identify infected animals. Longer duration can be a result of changes in the breakdown management policy (such as the interpretation of short-interval tests), a high within-herd prevalence of infection at the outset of a new breakdown, intense cattle-to-cattle transmission despite removal of identified infected animals (for example, in large herds) and continued reinfection (for example, from wildlife or contiguous/ neighbouring herds).
Median incident duration in the HRA of England, Wales and GB has been consistently stable since 2005 (around 215 days in the HRA, 240 days in Wales and 175 days in GB). Median duration in Scotland rose sharply in 2013, but decreased in 2014 (from 233 days to 193 days) and has been variable since 2005. Conversely, herd breakdown durations in the Edge area and LRA of England have increased slightly.
In GB overall, the proportion of OTF-W incidents ending in 2014 that had been under restriction for more than 550 days (18 months), also known as persistent bovine TB herd incidents, decreased slightly from 11.5 per cent in 2013 to 10.3 per cent in 2014 (similar to 2012, 10.7 per cent).
Of all breakdown herds, the proportion of persistent bovine TB herds (OTF-S and OTF-W) in 2014 was 7.5 per cent in England, 9.6 per cent in Wales, 0 in Scotland and 7.9 per cent in GB (9.9 per cent in 2013). Again, there was considerable variation between the English surveillance risk areas (HRA 8.3 per cent, Edge area 1.8 per cent, LRA 1.7 per cent). The vast majority (98.0 per cent) of these were OTF-W incidents.
Routine slaughterhouse surveillance
Routine postmortem meat inspection of non-reactor cattle carcases for tuberculous lesions complements the skin testing programme on farms. A slaughterhouse case is where suspected lesions of bovine TB are found via routine postmortem meat inspection and evidence of M bovis is found. These animals are not reactors, direct contacts or inconclusive reactors. On suspicion of bovine TB in the slaughterhouse, that is, lesions are found, the OTF status of the herd of origin is suspended (OTF-S) and samples are taken for culture. If M bovis is isolated from the slaughterhouse case samples, the OTF status of the herds becomes withdrawn (OTF-W).
The proportion of individual slaughterhouse case samples submitted to the APHA from which M bovis was isolated in 2014 was similar to 2013 (72.5 per cent in 2014 compared with 71.1 per cent in 2013). A small proportion were from herds already under restrictions, not just OTF herds. In the remaining cases, the lesions were due to other organisms (Actinobacillus 23.6 per cent), were negative (23.6 per cent) or failed to yield a result on bacteriological culture (0.3 per cent).
The proportion of total incidents disclosed in the slaughterhouse was lower; 14.0 per cent in GB in 2014 and 14.9 per cent in England.
Since 2010, around one-fifth of OTF-W incidents in England and GB have been detected by inspection at the slaughterhouse. In 2014, the proportion of OTF-W incidents that were first detected by meat inspection was similar for England (19.9 per cent) and across GB (18.3 per cent) and comparable to 2013.
There was substantial variation in the proportion of OTF-W incidents disclosed in the slaughterhouse across the English surveillance risk areas: 19.8 per cent in the HRA, 17.8 per cent in the Edge area and 37.1 per cent in the LRA in 2014. There was also variation at county level (Table 5.1 in APHA 2015b).
The proportion of OTF-W incidents detected by meat inspection in the Edge area dropped from 34.8 per cent in 2012 to 17.8 per cent in 2014 per cent, more in line with that of the HRA, reflecting the increased surveillance testing of herds in the Edge area. The proportion of new OTF-W incidents detected by routine meat inspection in Wales and England was similar until 2006, whereupon the proportion in Wales began declining, reaching a low in 2009, corresponding with the placement of the whole of Wales on annual testing in 2008. Since then, there has been a more recent gradual increase in Wales, although the proportion (12.3 per cent) remains lower than England, which peaked in 2011 to 24.4 per cent and has been 20 per cent or above since 2010.
The proportion of new bovine TB incidents detected in abattoirs in Scotland (as opposed to those detected by skin testing) varies significantly from year to year and wide fluctuations are to be expected. In 2014, two out of 15 new OTF-W incidents were detected by meat inspection (Table 4.1 and Fig 4.2 in APHA 2015a). Slaughterhouse surveillance is particularly important in Scotland following its attainment of Officially TB-Free status in 2009, and also because routine herd testing using the single intradermal comparative cervical tuberculin test is no longer conducted in Scottish herds where a sufficient proportion of cattle go for slaughter and only a limited number of animals are moved on from a high risk area (Anon 2011).
Postmortem and microbiological examination
Animals slaughtered for bovine TB control reasons include reactors to the tuberculin skin test (standard or severe interpretation), reactors to the supplementary γIFN blood test, inconclusive reactors to the skin test (which may be voluntarily slaughtered before their retest 60 days later) and direct contacts (non-reactor animals in OTF-W incidents which are perceived to have been at risk of infection with M bovis). Following slaughter, animals are examined for visible lesions at postmortem examination. Where more than one reactor is detected in a herd, a representative number of those animals undergo tissue sampling for microbiological culture and (if M bovis is isolated) molecular typing at the APHA national bovine TB reference laboratory.
In 2014, the vast majority of GB counties with visibly lesioned reactors had 95 to 100 per cent of those animals confirmed by culture. In the HRA and Edge area of England many counties had 92 per cent or more; and 100 per cent of visibly lesioned reactors were confirmed by culture in most LRA counties (based on low numbers of reactors [Fig 4.5 in APHA 2015a]).
In England, the proportion of reactors with evidence of infection with M bovis, whether it be lesions at slaughter or bacterial culture confirmation, varied according to risk region (χ2=30.1, 2 degrees of freedom, P<0.001). The positive predictive value of the diagnostic test will be influenced by background prevalence of M bovis infection in the cattle population and therefore variation by risk area is expected. Confirmation was lowest in the LRA compared to the Edge area and HRA. In the HRA, 53.2 per cent of standard reactors in 2014 had evidence of bovine TB infection, compared to 42.6 per cent in the Edge area and 33.2 per cent in the LRA (Table 6.5 in APHA 2015b).
A recurrent bovine TB herd incident is defined as a herd that had any bovine TB incident in the reporting year (ie, 2014) that had previously been under movement restrictions for any bovine TB incident in the previous 36 months. Recurrence can represent a general increase in incidence (herds would have a greater probability of a previous incident, if the past incidence was high), or may be due to a change in risk factors for bovine TB, culminating in the herds being more likely to have repeated incidents. Further details on the definition of recurrence used in the surveillance reports can be found in the respective sections in the GB, England and Welsh reports (APHA 2015a, b, c). The causes of recurrence are difficult to ascertain with surveillance data, since it is often not possible to disentangle undisclosed, residual cattle infection resulting from a previous bovine TB incident from reinfection derived from either exposure to infected cattle or wildlife, particularly in endemic areas where the same M bovis genotype may be isolated in successive bovine TB incidents.
Recurrent OTF-W incidents tend to be concentrated in the high-risk areas of GB (Fig 5). The majority of herds with recurrent OTF-W incidents had one or two incidents in the previous 36 months, with a small number of herds sustaining three, four or five previous incidents, mostly in endemic areas of south-west England and parts of Wales. There were two beef herds with a history of four OTF-W incidents in the previous 36 months, and these were located in Cambridgeshire and Dorset. The recurrent breakdown herd in Cambridgeshire was an Approved Finishing Unit for fattening only (animals purchased mainly from markets). It was described as wildlife-proof and inspected regularly with a continuous flow of animals going straight to slaughter. One beef herd in Devon had a history of five OTF-W breakdowns in the previous 36 months.
In 2014, 11 per cent of OTF herds in GB had experienced a bovine TB incident in the previous 36 months. Since 2006, England mirrors the same increasing proportion as GB but is approximately 1 per cent higher. The proportion of OTF herds with a history of bovine TB was highest in the HRA (26.7 per cent in 2014) and lower in the Edge area (6.7 per cent). In Wales, the proportion was 17.1 per cent. As expected, the proportion was much lower in the LRA and in Scotland (around 1 per cent since 2006).
Since 2011, approximately half of all bovine TB breakdowns in GB and England have occurred in herds with a history of a breakdown in the previous 36 months (England 52.5 per cent in 2013, 54.0 per cent in 2014; GB 51.0 per cent in 2013, 52.0 per cent in 2014). In England in 2014, this varied by surveillance risk area: HRA 58.4 per cent; Edge area 25.6 per cent; LRA 14.0 per cent. In Wales, the proportion has gradually declined from 50.0 per cent in 2011 to 45.9 per cent in 2014. In Scotland, low numbers of breakdowns mean the proportion has varied extensively, ranging between 2.3 per cent and 18.5 per cent over the past nine years, and in 2014 was 4.9 per cent (APHA 2015a).
Bovine TB genotyping
Genotyping is a molecular typing technique used to differentiate the strains of M bovis responsible for the bovine TB incidents in cattle and other species. DNA is taken from a sample of bacteria and genetic sequences unique to a particular strain are identified using two PCR-based techniques, known as spoligotyping and variable number tandem repeat (VNTR) typing, which target different regions or markers in the M bovis genome. This approach works well on bacteria grown in the laboratory, but is insensitive when using pathological samples taken directly from infected animals at postmortem examination. Therefore, M bovis bacteria from field samples need to be cultured in the laboratory before molecular typing of the isolates can take place, a process that can take up to six weeks and occasionally longer.
In 2014, 3304 cattle isolates from new breakdowns (3234 in 2013) and 170 isolates from other, non-bovine hosts were spoligotyped. For each cattle incident, an average of 1.1 isolates was genotyped (1.1 in 2013). The full genotype (spoligotype plus a six-locus VNTR) was obtained for 90 per cent of the cattle isolates, representing 2958 incidents (96 per cent in 2013). Maps displaying the major spoligotypes, plus home ranges for 27 genotypes (based on data from 2010 to 2014) can be found in Section 6 of the GB surveillance report (APHA 2015a).
Of the 170 mycobacterial isolates from non-bovine species genotyped in 2014 (2013=153), 161 were M bovis and nine were Mycobacterium microti. Eight M bovis isolates had no host species assigned. The 153 M bovis isolates with a host species assigned (Table 1) represent 85 separate incidents, where an incident may include multiple isolates from the same type of non-bovine host at the same location. Of the nine M microti isolates, eight were detected in cats. The total numbers of isolates from cats was comparable to previous years (22 in 2013); however, the proportion that were M microti was lower than in 2013 (n=19). Between 2009 and 2013, the number of incidents in sheep and pigs increased; however, in 2014, only one incident in sheep was recorded. Incidents in pigs also declined, with 15 incidents in 2014 compared to 22 in 2013.
In general, the genotypes of M bovis found in non-bovine species subject to scanning bovine TB surveillance reflect the frequency and geographical distribution of those found in the local cattle herds. Of the 68 non-bovine incidents for which home range map analysis was possible, 52 incidents were within the relevant cattle home range and 16 were out of home range. A total of 76.5 per cent of non-bovine incidents were either in the home range or within 10 km of the equivalent cattle home range.
The overall bovine TB incidence rate for GB in 2014 was 6.8 incidents per 100 herd-years at risk. This means that for every 100 herds that undergo a full year of surveillance, about seven breakdowns were detected in 2014. There were differences between countries of GB, risk areas and counties of England and Wales (APHA 2015a, c).
The disease situation is complex and heterogeneous. In Wales, the sharp decrease in the number of new incidents observed between 2012 and 2013 has stabilised between 2013 and 2014, and disease prevalence remains relatively high (4.7 per cent in 2014). In Scotland, there was an increase in the number of new incidents in 2014; however, this is largely explained by the multiple purchase of infected stock from the sale of a dairy herd from Cumbria, which resulted in the dispersal of a number of infected cattle to several farms in Scotland. Overall, incidence rates and the number of incidents detected in England have been relatively stable since 2008 (7.7 breakdowns in 2014 per 100 herd-years at risk) and prevalence has fallen in the HRA (9.7 per cent in 2014). In the LRA of England, bovine TB incidents continue to be few and sporadic and the herd incidence and prevalence are comparable to Scotland. In the Edge area of England, the incidence rate has been relatively stable (3.9 breakdowns per 100 herd-years at risk), but prevalence has risen steeply (2.8 per cent in 2014), coinciding with a 40.3 per cent increase in herd tests in this area.
The bovine TB disease situation in GB is difficult to evaluate, due to frequent changes in surveillance and control policy; for example, regarding testing frequency and local policies such as in the Intensive Action Area in Wales and the Edge area of England, and also because of the extremely variable disease levels across the country as a whole. Changes in testing policy affect incidence, prevalence, duration and the number of new incidents detected, resulting in frequent fluctuations within the data that suggest extreme caution should be exercised in examining changes between years. To be sure that there is a reduction in the frequency of bovine TB in England and Wales, there would need to be a marked and sustained downward trend in the incidence and prevalence of infected herds for more than five consecutive years (to avoid historical cyclical trends associated with testing frequency), as well as a reduction in the geographical range of new incidents in cattle and non-bovine animals.
The authors would like to thank James Tiller and Alison Prosser, who assisted with the compilation of the annual surveillance reports, and James Dale, Monika Klita and Karen Gover, for their technical expertise in genotyping the strains. They would also like to thank Glyn Hewinson and their policy customers and veterinary advisers, Ricardo de la Rua-Domenech, Malla Hovi, Simon Rolfe and Martyn Blissitt, for their advice and guidance in the production of the surveillance reports and for reviewing this summary.
The compilation of the annual bovine TB surveillance reports is funded by Defra and the Scottish and Welsh Governments (project SB4500). Bovine TB genotyping was funded by Defra project SB4020.
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.