Annex
RESPONSE TO
"TUBERCULOSIS IN
CATTLE AND
BADGERS: A REPORT
BY THE
CHIEF SCIENTIFIC
ADVISER"
FJ Bourne, CA Donnelly, DR Cox, G Gettinby, JP
McInerney, WI Morrison and R Woodroffe
Former members, Independent Scientific Group
on Cattle TB
SUMMARY
ISG1 The Independent Scientific Group on
Cattle TB (ISG) was an independent body charged with developing
science-based policy options for the control of tuberculosis (TB)
in cattle. As the ISG was dissolved in June 2007 following publication
of its final report (Bourne et al, 2007), this response
to the Chief Scientific Adviser's recent report (King et al,
2007) presents the views of the seven former members of the ISG.
ISG2 A key conclusion of our work was that
"badger culling can make no meaningful contribution to cattle
TB control in Britain" (Bourne et al, 2007). This
conclusion is based on the analysis and interpretation of data
derived from nearly 10 years' scientific research by ourselves
and our colleagues. Our major scientific findings, and their interpretation,
have been published in top-quality peer-reviewed journals, and
our conclusions are widely accepted within the scientific community.
We are surprised, therefore, that King et al (2007) reached
a different conclusion, namely that "the removal of badgers
could make a significant contribution to the control of cattle
TB in those areas of England where there is a high and persistent
incidence of TB in cattle".
ISG3 We believe that a key reason for these
differing conclusions is that King et al (2007) were constrained
within their terms of reference, which prevented them from fully
evaluating policy options. While we aimed "to present Ministers
with a range of scientifically-based policy options which will
be technically, environmentally, socially and economically acceptable"
(Bourne et al, 1998), King et al (2007) were "...asked
to make comment on scientific issues..."; their "...brief
did not extend to economic or other practical issues" (Environment,
Food and Rural Affairs Committee, 2007). Unfortunately, the complex
relationship between badger abundance and cattle TB risks, as
revealed by our work, means that "economic [and] practical
issues"—which determine how, where, when, and on what
scale badger culling might be conducted—are absolutely critical
in determining whether culling would reduce or increase the incidence
of cattle TB. By excluding consideration of such issues from Sir
David King's remit, Ministers severely hampered his ability to
inform policy development.
ISG4 In addition to this broad concern about
King et al remit, we have identified a large number of
scientific problems with their report, which have led them to
draw conclusions from our work which are not consistent with the
data available. In particular, King et al (2007) dismiss
as "unsound" our finding that badger culling increases
TB risks for cattle on neighbouring unculled land, yet their conclusion
is undermined by (i) incorrect interpretation of statistical confidence
intervals; (ii) exclusion of data accrued between the first and
second proactive culls, even though this cannot be justified either
by statistical bias or by the time taken for changes in the badger
population to cause detectable effects in TB risks for cattle;
and (iii) incomplete consideration of ecological data consistent
with detrimental effects observed among cattle. In addition, misinterpretation
of our mathematical modelling work is likely to have led King
et al (2007) to under-estimate the likely benefits of improved
cattle-based controls.
ISG5 Given these concerns, we are not persuaded
by the arguments in King et al (2007) report and stand
by our published recommendations concerning the control of cattle
TB in Britain.
INTRODUCTION
ISG6 The Independent Scientific Group on
Cattle TB (ISG) worked from its outset to build and interpret
a science base to inform the control of tuberculosis (TB) in cattle.
As members of the ISG, we analysed and interpreted data from the
Randomised Badger Culling Trial (RBCT) and related studies. The
ISG was dissolved in June 2007 following publication of its final
report (Bourne et al, 2007). This response to the Chief
Scientific Adviser's recent report (King et al, 2007) thus
represents the views of the seven former members of the ISG.
ISG7 Our primary findings, including their
interpretation, were peer reviewed, both before publication in
scientific journals, and by the Defra-appointed statistical auditor;
comments were also sought from colleagues at Defra and its associated
agencies, some of whom were co-authors on our papers. We have
throughout our work encouraged informed debate and discussion,
and continue to welcome further constructive comment and dialogue
on any aspect of our work.
ISG8 Like King et al (2007), we recognised
from the start of our work that "the overriding aim is to
control TB in cattle". Indeed, the opening sentences of our
final report noted that "Bovine TB is a serious infectious
disease of cattle. It has public health implications, has major
economic consequences for Government and the farming industry,
and causes distress to farmers and their families" (Bourne
et al, 2007). Hence we firmly agree that "...strong
action needs to be taken now to reverse the upward trend of this
important disease" (King et al, 2007). Our recommendations,
summarised in our final report (Bourne et al, 2007) on
the basis of data published in a broad array of peer-reviewed
papers, therefore represent our views of the best way to achieve
control of cattle TB using methods currently available.
ISG9 A key conclusion of our work was that
"...while badgers are clearly a source of cattle TB, careful
evaluation of our own and others' data indicates that badger culling
can make no meaningful contribution to cattle TB control in Britain.
Indeed, some policies under consideration are likely to make matters
worse rather than better" (Bourne et al, 2007). We
note that our broad conclusions regarding the role of badger culling
are consistent with views expressed previously by experts on the
ecology of TB in badgers such as Dr Chris Cheeseman,[7]
Professor David Macdonald[8]
and Professor Tim Roper,[9]
as well as with recent statements by leading scientists charged
with past independent reviews of the issue such as Lord Krebs.[10]
However, our conclusion contrasts with King et al (2007)
recommendation that "the removal of badgers could make a
significant contribution to the control of cattle TB in those
areas of England where there is a high and persistent incidence
of TB in cattle" (King et al, 2007).
THE IMPORTANCE
OF TERMS
OF REFERENCE
ISG10 At the start of our work, our stated
aim was "to present Ministers with a range of scientifically-based
policy options which will be technically, environmentally, socially
and economically acceptable" and the RBCT was therefore designed
as a trial of potential policy options (Bourne et al, 1998).
Our approach contrasts with the remit of King et al. (2007),
who were "...asked to make comment on scientific issues...";
their... brief did not extend to economic or other practical issues"
(Environment, Food and Rural Affairs Committee, 2007). We consider
this distinction critical, as we believe it partially explains
the difference between our conclusions and those of King et
al (2007).
ISG11 We recognise three major reasons why
the "economic [and] other practical issues" excluded
from King et al (2007) terms of reference should be given
detailed consideration in any scientific discussion of future
TB control policy. (i) Although our findings suggest that, in
principle, modest reductions in the overall incidence of cattle
TB would result from simultaneous, coordinated and repeated culls
of badgers over extremely large areas of the countryside, using
skilled staff and ideally within geographical barriers to badger
movement, trying and failing to achieve this is likely to make
matters worse, increasing the incidence of disease in cattle and
spreading infection to new areas. As discussed in our final report
(Bourne et al, 2007), it is highly unlikely that coordinated
culls could be conducted simultaneously and repeatedly across
hundreds of square kilometres—especially since Defra's Wildlife
Unit was dissolved—whereas our work shows that culling which
is asynchronous, patchy, small scale or discontinuous is likely
to increase rather than reduce the incidence and spatial spread
of disease (Donnelly et al, 2007; Donnelly et al,
2003; Woodroffe et al, 2006b; Jenkins et al, 2007).
(ii) There are too few natural geographical barriers to badger
movement in TB-affected areas of Britain to contribute to national
control strategies, while building and maintaining artificial
barriers would be extremely costly and highly impractical on the
scale at which TB control is needed (Bourne et al, 2007;
Poole et al, 2002). (iii) Even if benefits were achieved
by culling, our results indicate that these would be modest in
comparison with the substantial financial costs of conducting
the widespread culls that would be required. These three issues
illustrate the critical importance of the "economic [and]
other practical issues" excluded from Sir David King's remit.
Our own consideration of these issues was based on systematic
evaluation of ecological and epidemiological data derived from
the RBCT and other studies, and our conclusions were reinforced
by economic data. We feel that it would have been very difficult
for Sir David and his team to reach meaningful policy recommendations
without similarly detailed consideration of such issues, and it
is therefore unfortunate that their terms of reference were so
narrow.
INTERPRETATION OF
SCIENTIFIC DATA,
STATISTICAL ANALYSES,
AND MODELLING
RESULTS
ISG12 In addition to this broad concern
about the Chief Scientific Adviser's terms of reference, we wish
to express six major concerns about the scientific basis of King
et al (2007) report. These concern (i) incorrect use and
interpretation of statistical confidence intervals in subgroup
analysis; (ii) inappropriate exclusion of data accrued between
the first and second proactive culls; (iii) failure to consider
or cite the ecological data and analyses relevant to our conclusions;
(iv) misunderstanding of our mathematical modelling work and its
implications; (v) incorrect interpretation of our conclusions
regarding temporal trends in the effects of culling; and (vi)
over-reliance on assumptions concerning the effects of culling
on transmission among badgers. These major issues are addressed
below; a point-by-point response to King et al (2007) report
is provided in an Appendix to this document. This report is also
accompanied by copies of three scientific papers describing findings
from the RBCT, which have been published in the peer-reviewed
literature since publication of our final report (Jenkins et
al, 2007; Pope et al, 2007b; Woodroffe et al,
2007). Key findings reported in these three papers were cited
in our final report (Bourne et al, 2007), referring to
the papers as "in press" (in the case of Woodroffe et
al), "in review" (in the case of Jenkins et al)
or citing a report to Defra (in the case of Pope et al).
(i) Interpretation of statistical confidence
intervals in subgroup analysis
ISG13 King et al (2007) place great
weight on interpretation of results presented in Figure 5.2B of
our final report (Bourne et al, 2007). This figure shows
the results of a subgroup analysis, stratifying overall RBCT results
into beneficial and detrimental effects recorded at different
distances from the boundaries of proactive trial areas. As is
often the case in scientific studies, each effect was presented
as a point estimate, associated with 95% confidence limits. The
latter give a measure of the uncertainty associated with the point
estimate; there will be substantial uncertainty when small sample
size prevents a precise estimate from being obtained from a subset
of the data.
ISG14 In paragraph 43 of King et al
(2007), the detrimental effect of badger culling on cattle TB
incidence outside culling area boundaries is dismissed since "Three
out of four [confidence intervals] go through zero (ie one cannot
be confident that the overall effect is detrimental)". This
interpretation is incorrect, as it omits one very crucial proviso
in the interpretation of our subgroup analysis. The limits attached
to any subgroup of data concern what can be learned from that
subgroup of data on its own. While it is useful to know this,
it is rarely the primary focus of analysis, and was not so in
this case. Indeed if the data are broken into a large number of
small subgroups, each on its own will have substantial uncertainty
and be indecisive on any issue of concern, even if the overall
picture is entirely clear on the point under study, as was the
case here. The overall picture must be studied, not fragments
of it. This is what we have done in all our analyses, subject
to tests of the uniformity of the effect under study.
ISG15 In the interests of consistency, we
note that the same analysis showed that all five of the subgroup
estimates of beneficial effects of culling inside proactive areas
had confidence intervals which included zero. Thus, were the same
(erroneous) interpretation to be placed on those findings, evidence
of the beneficial effects of culling should likewise have been
dismissed.
(ii) Exclusion of the first year of data post-culling
ISG16 In our published accounts of the effects
of proactive culling on cattle TB incidence (Donnelly et al,
2007; Donnelly et al, 2006), we provided estimates of culling
effects following completion of the first proactive cull, and
also following the second cull; the latter analyses excluded roughly
a year of data accrued in each triplet between the first and second
culls. King et al (2007) chose, where possible, to present
results from the second cull only, noting in paragraph 29 that
our use of the full data set to assess detrimental impacts outside
culling areas might "overestimate the effect". In fact,
there is no such bias toward overestimation, nor do King et
al (2007) put forward any case for such a bias. Since cattle
in RBCT areas were tested annually but not simultaneously, some
breakdowns detected in the first year would have originated from
infections which occurred prior to culling, and some infections
which occurred in the first year would not be detected as breakdowns
until the second year. Contrary to King et al (2007) statement,
this would make culling and no-culling areas appear more similar
in the first year: the estimated effects on the incidence of detected
breakdowns—whether beneficial or detrimental—would
be smaller than the underlying effects on the incidence of new
infections in this time period. This is important because the
exclusion of this initial time period, which cannot be justified
on the basis of any bias toward overestimation, appears to have
contributed to King et al (2007) tendency to downplay the
importance of detrimental effects.
ISG17 Paragraphs A16 and A41-A44 in the
Appendix give more in-depth interpretation of results from different
time periods.
(iii) Failure to consider ecological data
ISG18 King et al (2007) dismiss detrimental
effects of badger culling on cattle TB as "hard to interpret"
and "unsound", noting that they were "not fully
persuaded by" our explanation for these effects. In this
context, it is unfortunate that King et al (2007) appear
to have considered only a part of the wealth of ecological data
that we have published on the impacts of culling on badger abundance,
distribution, ranging behaviour and infection status, citing none
of our primary papers on these issues. This work shows that badger
culling prompted immigration into culled areas (Woodroffe et
al, 2007) as well as disruption of badger territories and
expanded ranging (Pope et al, 2007b; Woodroffe et al,
2006a). These ecological changes were associated with reduced
clustering of infection in both badgers and cattle (Jenkins et
al, 2007), and also with elevated prevalence of M bovis infection
in badgers in both proactive and reactive culling areas (Bourne
et al, 2007; Woodroffe et al, 2006b). These ecological
findings are consistent with, and hence support, our observation
of elevated cattle TB incidence on land neighbouring proactive
culling areas, and in reactive areas.
ISG19 King et al (2007) failure to
give full consideration to the ecological data is important as
it may have given them an incomplete or biased picture of badger
ecology and M bovis epidemiology, leading to inappropriate conclusions
and recommendations. For example, the recommendation to "reduc[e]
the migration of badgers into the removal area by... soft boundaries
(such as arable land with no cattle)" is unsupported by any
ecological data and highly unlikely to be effective.
(iv) Conclusions of our mathematical modelling
work
ISG20 King et al (2007) discuss the
conclusions of our mathematical modelling work (Cox et al,
2005), on which we based our conclusion that "the rising
incidence of disease can be reversed, and geographical spread
contained, by the rigid application of cattle-based control measures
alone" (Bourne et al, 2007). We used this model essentially
to estimate how close the epidemic is to criticality and hence
to infer the likely consequences of policy action, including controls
aimed at cattle. Of course all such modelling is based on highly
idealised assumptions but it, combined with careful study of all
the relevant data that have been collected, is the only rational
basis for assessing the likely consequences of policy actions
not yet undertaken.
ISG21 Unfortunately, King et al (2007)
have misinterpreted this work. Cox et al (2005) present
an explicitly two-species model and the R0 estimates obtained
do not, as stated in paragraph 11 of King et al (2007),
refer to cattle-to-cattle transmission alone. Rather, these estimates
are for cattle within the badger-cattle disease system, and thus
represent contributions from both cattle-to-cattle and badger-to-cattle
transmission. Indeed, equations 13 and 14 in Cox et al
(2005) present analytical solutions based on two different assumptions
regarding disease levels in the badgers, the first that the badger
disease levels stay constant and the second that they follow the
increasing pattern observed in cattle. These equations give indistinguishable
fits to the data, and so the true situation cannot be inferred.
The important issue for the conclusions presented in Cox et
al (2005) is that the estimates of R0 are essentially identical.
We agree with King et al (2007) that the levels of badger-to-cattle
and cattle-to-cattle transmission are likely to vary considerably
throughout the UK. However, if the incidence of cattle TB in low
incidence areas is driven largely by the epidemic in the South-West
of England then the growth rate, and thus the R0 estimates, for
both areas will be similar.
ISG22 We are concerned that the failure
by King et al (2007) to fully appreciate the structure
of this important modelling work, and the estimates obtained from
it, will have led them to under-estimate the potential for reducing
the incidence of cattle TB in Britain using cattle-based measures
alone.
(v) Conclusions regarding temporal trends
in the effects of culling
ISG23 In paragraph 37 of their report, King
et al (2007) raise the possibility that detrimental effects
of badger culling on badger ranging and disease transmission might
be "transient", indicating that we did not consider
this possibility. This is incorrect; several of our papers present
and discuss evidence for temporal trends in the effects of culling.
This issue is discussed in detail in paragraphs A28-A30 below.
ISG24 King et al (2007) restrict
their discussion to the transience (or otherwise) of disruption
to badger spatial organisation; as detailed in paragraph A30 below,
there is evidence to suggest that such ecological effects were
sustained in proactive culling areas. However, as discussed in
Donnelly et al (2007), the possibility remains that detrimental
effects on cattle TB could be transient—or at least of smaller
magnitude following later culls when sustained reductions in badger
densities have been achieved—despite continued disruption
of badger ecology. We believe that the likely outcome of future
TB control options involving badger management can be reliably
predicted only through separate, but parallel, consideration of
effects on cattle and badgers.
(vi) Capacity to reduce badger-to-badger transmission
by culling
ISG25 King et al (2007) note that
"the likelihood of uninfected badgers being exposed to infectious
badgers will... be reduced [by culling]" (paragraph 38).
While such an effect would be expected in the randomly-mixing
populations assumed in simple epidemiological models, more complex
effects can occur in socially structured populations (Keeling
and Eames, 2005). In fact, published data suggest that the substantial
reductions in badger density which were achieved by proactive
culling (Woodroffe et al, 2007) increased badger-to-badger
transmission of infection rather than reducing it (Bourne et
al, 2007; Woodroffe et al, 2006b). The likely reasons
for this pattern are detailed in paragraph A31 below.
ISG26 Although suppression of badger densities
to extremely low levels (substantially lower than those achieved
in the RBCT) would be expected to reduce badger-to-badger transmission
of infection, no data are available on the densities at which
this might be achieved in TB-affected areas of Britain. However,
evidence for widespread cattle-to-badger transmission (Jenkins
et al, 2007; Woodroffe et al, 2006b) suggests that
even extremely low-density badger populations would rarely remain
TB-free. Hence, data do not support King et al (2007) characterisation
of culling as an "intervention... [to] reduce the prevalence
of disease in... wildlife" (paragraph 5).
CONCLUSIONS
ISG27 In summary, King et al (2007)
appear to have given incomplete consideration both to the scientific
data and to the wider issues critical for determining the likely
outcomes of particular approaches to the management of bovine
TB. In addition, we have major concerns about their interpretation
of our scientific data. Thus, we are not persuaded by the arguments
in their report and stand by our conclusion that "badger
culling can make no meaningful contribution to cattle TB control
in Britain" (Bourne et al, 2007).
7 Dr Chris Cheeseman, Oral evidence to Environment,
Food & Rural Affairs Committee 2006 "I would venture
to suggest now that I do not believe that any culling policy is
sustainable in the long term." http://www.publications.parliament.uk/pa/cm200506/cmselect/cmenvfru/905/6020704.htm Back
8
Prof David Macdonald, Letter to The Guardian 2006 "The
evidence is that a badger cull on a scale or level of efficiency
that seems feasible will not solve cattle farmers' problem"
http://www.guardian.co.uk/environment/2006/mar/10/guardianletters.conservationandendangeredspecies Back
9
Professor Tim Roper, Press release on behalf of the Mammal Society
2006 "While we understand the farming community's concerns,
we believe the available evidence does not justify a policy of
badger culling" http://www.abdn.ac.uk/mammal/badgercull-press.shtml Back
10
Lord Krebs, Discussions in House of Lords 2007 "We now
know from reading the report of the Independent Scientific Group
that culling is not a viable policy option. There is no wriggle
room." http://www.publications.parliament.uk/pa/ld200607/ldhansrd/text/70726-0001.htm Back
|
