Poor delivery of a poor test method

testing with calipers

Under the current UK ‘test and cull’ policy, cattle are tested at intervals determined by risk of infection. Those that fail the test are slaughtered and severe restrictions on cattle movement are placed on the farm.

In the test primarily used in the UK, a small amount of tuberculin (a sterile extract obtained from a culture of M. bovis) is injected into the animal and a skin measurement is taken with calipers (left).

On revisiting the injection site 72 hours later, measurement is undertaken again A swelling will occur if the animal has previously been challenged by, and the immune system has reacted to, tuberculosis bacteria.

However, reactions of the animal’s immune system to other types of mycobacteria can also cause the swelling. To reduce the number of false diagnoses this would lead to, a preparation of M. avian, the avian form of tuberculosis, is injected nearby. M. avian is widely present in the environment. It is not harmful to cattle, although their immune system reacts to it.

The swellings are compared after 72 hours. Statistical studies, rather than an understanding of the underlying biological mechanism, have indicated that if the M. bovis swelling is significantly larger than the M. avian swelling, a reaction to M. bovis has occurred. The animal is then deemed to be a ‘reactor’ and killed. This test is known as the ‘skin test’ (or more properly as the single intradermal comparative cervical tuberculin test).

A blood test, the gamma interferon test, is also used on some occasions as an ancillary test. It evaluates the same immune response but is conducted in the laboratory. There are documented cases of wildly differing results from using the two tests on the same herd.
If an animal fails either the skin or blood tests it is slaughtered.
Serious concerns (some of which we explain below).

  • Compromised
  • Functionally flawed
  • Worse than the disease!
Healthy cattle are often falsely condemned

After sixty painful and expensive years of testing and slaughtering cattle, and an intervening period of relatively few incidents of Bovine TB, we are again several decades from any chance of ‘official TB free’ status.

Bovine TB has been a difficult and demanding problem for many years. There are reasons for believing that it can be controlled and finally eradicated but this will require a long-term commitment by all stakeholders and take at least 20 years.

Can we afford the cost and will farmers tolerate another 20 years of movement restrictions, disruptive and inaccurate testing and compulsory cattle culling?
Over the period 1998 - 2009 the number of new herd incidents13 in Great Britain increased by 276%, and the number of cattle consequently culled, by a staggering 477%. Defra states that; “The causes of the long-term increase in bTB in GB are not well understood as there are likely to be many factors involved”.
In other words Defra does not know why there has been a dramatic long term increase, or why Bovine TB policy has failed.

A functional test should detect cattle that have, or will have, Bovine TB.
The skin test does not do this, it identifies animals that have come into contact with M. bovis and mounted an immune reaction - exactly what a healthy animal should do.
The latent infection that remains may in some of these animals re-emerge as Bovine TB, but not in all. All are slaughtered.

Only about one third of reactors show evidence of infection at postmortem and can be listed as ‘confirmed reactors’. Much of the compensation paid to farmers is for healthy cattle that were unlikely to develop Bovine TB, or would have been slaughtered in the normal course of farm production long before any symptoms developed.

The present testing regime:
* Condemns thousands of cattle in error.
* Fails to detect a significant proportion of those cattle that are infected.
* Looks for the wrong thing.

Culled cowBesides the shortcomings in the ‘imperfect’ testing regime, Bovine TB policy is having severe effects on farming. Healthy cattle are being slaughtered and farmers are consequently suffering unnecessarily. If Bovine TB itself was affecting farm productivity, evidence would have emerged by now. It is not easy for cattle to catch Bovine TB and clinical symptoms are rarely seen on farms. Testing requires unfamiliar and stressful handling of cattle, compromising both animal welfare and human safety.

Compensation does not always cover the value of the animals and certainly not the consequences of movement restrictions and loss of critical breeding stock.
Defra, in its ‘Bovine Tuberculosis Evidence Plan 2011/12’ 6 states, “... we also recognise that the current TB surveillance and control regime in cattle (based on test and slaughter of reactors) is not preventing the spread of TB to clean areas, and that the incidence of disease in endemic areas appears to be increasing.” In other words Defra, to its credit, recognises failure of the policy. Sadly Defra has not proposed cessation or a viable alternative.

The ‘skin test’ is compromised by major shortcomings.

False positives
According to Defra 11 the test falsely condemns only 1 in 1,000 cattle tested. (The ‘specificity’ of the test). This makes the test sound accurate until what it really means is realised.

Using Defra’s testing figures 15 (for 2009):
In England 4,899,144 tests were performed,
1 in 1,000 tests, 4,899 in those 4,899,144, will be false reactors or ‘false positives’, 24,924 cattle were actually condemned as reactors,
4,899 or 1 in every 5 of those cattle will have been incorrectly condemned.

In Wales 1,812,666 tests were performed,
1 in 1,000 tests, 1,812 in those 1,812,666, will be false reactors or ‘false positives’,
10,117 cattle were actually condemned as reactors, 1,812 cattle, or 1 in every 6 of those cattle will have been incorrectly condemned.
In Scotland 229,800 tests were performed,
1 in 1,000 tests, 229 in those 229,800, will be false reactors or ‘false positives’, 323 cattle were actually condemned as reactors, 229 or a staggering 2 out of every 3 of those cattle will have been incorrectly condemned.

The principle of this calculation has been confirmed by Defra in a letter ref. RFI 3725 & RFI 3749 of 19 January 2011. Not only are thousands of cattle being falsely condemned, but as herd size increases, the chance of a false positive in the herd increases. Thus the larger the herd, the greater the chance that such a false positive will be the sole cause of movement restrictions and repeat testing on the entire farm, with all the accompanying disruption, costs and anxiety to the owner.

The gamma interferon blood test, used on some occasions as an ancillary test, has a higher sensitivity, thus showing less false negatives (see below) but, having a massively inferior specificity. It condemns an even higher proportion of cattle as false positives.

False negatives
According to Defra the skin test misses 1 in 5 cattle that it should identify as reactors. (This is the ‘sensitivity’ of the test). For every four ‘reactors’ slaughtered in the belief that they are or will become infectious or infected, one more remains undetected and potentially infectious in the herd or worse still, moved to infect another herd or area. If one or more reactors have been found in the herd, a further test is done 60 days later and it may then detect the missed reactors, or maybe not.

In many countries this shortcoming is recognized and the skin test is used as a herd test. All animals in the herd are tested individually as in Britain, but if a single reactor is found, the entire herd is slaughtered and restocking is delayed.