Examination of Witnesses (Questions 40-59)
10 DECEMBER 2003
MR BEN
BRADSHAW, MS
SUE EADES,
DR NICK
COULSON, PROFESSOR
JOHN BOURNE
AND PROFESSOR
CHRISTL DONNELLY
Q40 David Taylor: That there is no
empirical proof that culling induced perturbation has epidemiological
consequences which show themselves as an increase in cattle TB.
Professor Bourne: I would agree
exactly with that, we do not know.
Q41 David Taylor: You said there
is evidence but no proof.
Professor Bourne: We do not know
if the effect that has been observed in reactive culling is down
to perturbation. It is a likely possibility but we simply do not
know.
Q42 Mr Wiggin: I have been listening
very carefully to what you said. First of all, I would like to
know what extra benefit could have been derived from continuing
the reactive cull to the end of the year?
Professor Bourne: It would have
given us breathing space to look again at the data with the more
complete data set. The data set we had was only until the end
of August. We would have had a complete data set to the end of
January so there would have been that advantage. It would have
given us breathing space and it was simply that which we were
asking for but, as you saw in our report, we did present the view
to Ministers that at that particular time the situation was unlikely
to change very much.
Q43 Mr Wiggin: Was it possible that
you could have got a decrease in the number of TB breakdowns if
you had continued, is it possible?
Professor Bourne: Perhaps my colleague
could answer this. Certainly the range of results would have decreased
but the impact would remain the same. Confidence intervals would
have narrowed but the result would have been the same.
Q44 Mr Wiggin: Did you want to say
anything?
Professor Donnelly: Yes. We have
looked at the possibility of just projecting forward in discussions
about what would happen if we had continued. Even if there had
been no breakdowns in reactive areas from that point onwards and
other treatments had continued as they were, that would only have
brought the estimate that there was no difference between survey
only and reactives. Even if we had continued with the data, we
would not have got to a point where that estimate shifted and
said there was a benefit of reactive.
Q45 Chairman: I think the Minister
wants to make a comment on your earlier question.
Mr Bradshaw: Yes, just in case
you do not ask a supplementary along the lines of "Well,
if Professor Bourne said he wanted to carry on until the end of
this coming season why did you not take that advice." I think
the answer is pretty clear and it is one that Professor Bourne
and I discussed and he agreed with me that we had to put these
figures in the public domain. I think a member of your own Committee
in an adjournment debate had asked for an assurance, and I had
given it, that if there were any dramatic or startling interim
results from the ISG work that we should make that known and as
soon as Professor Bourne found these dramatic results he came
and told me about them. I wanted to put them in the public domain
and it would have been inconceivable I suggest to this Committee
that having discovered very dramatic and, as I have been assured
by both John and Christl, safe categorically and peer reviewed
science of a 27% increase in the reactive culling areas, if I
had then said "Oh, well, just carry on until the end of the
culling season, that is okay" knowing that policy would cause
extra TB outbreaks. I think that would have been a completely
politically unsustainable position to adopt.
Q46 Chairman: Notwithstanding the
scientific point that Professor Bourne has just made? Or, again
for clarification sake, you would have liked to have carried on
but you did not think the result would have been significantly
different if you had carried on?
Professor Bourne: That is true.
We would have liked to have had more time to think about the results,
get more data and do other analyses which may in fact have gone
some way towards explaining why this effect occurred. We were
unable to do that because spatial analysis, for instance, we did
not have time to do.
Mr Wiggin: Many critics believe that
what you are doing scientifically is rather as you outlined, it
is about getting empirical evidence for policy formation, it is
not about deciding how TB travels from badger to cattle or cattle
to badger. Therefore, I do not understand how the Minister can
have reacted in the way he did because having been asked by the
ISG if they could continue the tests, aware, obviously, that there
was some penalty, of course, and I understand why the Minister
made the decision but I do not think the Minister made the right
decision because if he had we would not have needed to have culled
badgers in the first place, we could have done a great deal of
other things if we were trying to save particular lives. Where
he has gone wrong in my opinion is that he has reacted without
following the scientific guidelines that you were trying to get
done. Therefore, I am very unhappy as to the answer he just gave
because it does seem as though he has parted company with the
science and just gone for what he thought was right, which I do
not blame him for but we all know about badgers and cattle already.
Paddy Tipping: You are blaming him.
Mr Wiggin: All right, I am blaming him.
I am very unhappy with the fact we do not have the proper scientific
evidence now to build policy going forward.
Q47 Chairman: Could we have that
on the record.
Professor Bourne: From the outset
as a group we considered the ethics and morality and the scientific
validity of carrying out this work. We recognised that we were
trialling two policy options, this was not an experiment in the
way that you might think of it. It was trialling a policy option
but we put it in the field in a way that it would provide us with
a range of epidemiological data which had scientific validity
but we had to do it also in a way that if it was adopted as a
policy option it could be translated seamlessly into policy if
it showed to have an effect which was regarded by ministers as
making a useful contribution to disease control, useful to features
of economics and all the rest of it. There is no doubt in my mind
about the importance of doing this work, about the importance
of continuing a proactive culling and I have no doubt also I support
the Minister in bringing the reactive cull to an end because clearly
it was not going to be a policy option. To have continued it for
the benefit of scientists just to get more scientific data would
have been fine from my perspective but I can see why the Minister
did not support that, and I agree with it. The proactive cull
is a very different matter. It is a proactive cull which is giving
us the main epidemiological data. The reactive cull was only ever
included to trial a potential policy option with a serious limitation
on the epidemiological data that we could get from it. When the
decision was made to release the reactive data I advised the Minister
that I thought that decision brought an end to reactive culling
and I was concerned that the proactive cull was not impugned in
any way and I could envisage that if one continued with the reactive
cull on the basis that we are killing badgers unnecessarily, creating
a problem for dairy farmers or certainly not helping the problem,
the rise in emotion which would inevitably result from that could
seriously impugn the proactives. Now I feel very strongly that
the proactives must continue unhindered to give us the data we
need for future policy. I accept that culling badgers may or may
not be a part of future policy but really we have to know what
impact the culling that we are allowed to do, given the sustainable
approach we have had to adopt, will have on cattle TB breakdowns.
Q48 Mr Drew: I think to start with
it would help me if you could explain in pure science terms what
we mean by the stopping rule? This was the idea that if there
was a statistical outcome which proved that one of your premises
was either proved or disproved then you would not continue with
it. Can you just put it into words that we understand.
Professor Bourne: I will ask my
colleague to comment on that. You say our initial premises, we
did not have any initial premises. We were carrying out a piece
of work to determine whether culling badgers on a reactive or
proactive basis could form part of future policy.
Q49 Mr Drew: Could Professor Donnelly
say then what you take to be the meaning of the stopping rule
and why that resulted in where we are at now?
Professor Donnelly: We considered
in this the estimate comparing the TB incidence in proactive and
reactive culling to the TB incidence in survey only areas. We
considered with the advice given to us as well by the statistical
auditor, Professor Denis Mollison—the statistical auditor
is independent of the ISG and in fact was put into place on a
recommendation of this Committee—the stopping rule looking
at getting a sufficiently precise estimate. Specifically at that
point we were looking for a reduction in TB incidence, so how
much of a reduction we would see in TB incidence associated with
either the reactive or proactive. In fact, when we did an interim
analysis in May of this year, which was only seen by myself and
Sir David Cox, we found evidence that was indicating that there
appeared to be an increase in TB incidence associated with reactives.
Then we discussed that and the uncertainty associated with that
estimate with Denis Mollison who agreed with us that there was
not sufficient precision in that estimate, we should wait for
another six months of data.
Q50 Chairman: Professor Donnelly,
I am sorry to interrupt. When you mention the name of somebody
who may not be entirely familiar to every member of the Committee,
or indeed to those here, would you be kind enough to tell us for
what organisation they work or what position they hold, it would
be very helpful.
Professor Donnelly: Sure. Denis
Mollison is a professor at Heriot-Watt University. He is a professor
of statistics and he was appointed as the statistical auditor
for the ISG.
Professor Bourne: On the advice
of previous select committees.
Professor Donnelly: Both Sir David
Cox and myself and also the statistical auditor, Denis Mollison,
recommended that we continue and re-examine the data after an
additional six months of this later interim analysis. It was the
feeling of all three of us that it had reached a sufficient level
of precision, given particularly that it was an increase in TB
incidence, that we should take this to the ISG and subsequently
to the Minister. It would be important to realise that there is,
I would argue, a big difference between looking at a 27% increase
or a 27% decrease and having in the lower end of our confidence
interval a very small negative effect. That would be very different
from having a 27% decrease in TB incidence and it could be virtually
no effect because that would not have been then sufficiently precise
to argue to stop the reactive cull because it was working. We
felt it was sufficient in this case to carry on discussions about
the reactive cull on the basis that it was at best showing no
effect.
Q51 Mr Drew: If I can follow through
the logic of that because that is very clear and it helps me.
What account did you take in terms of the reactive area of the
percentage of animals that were culled on each occasion? Presumably
that was factored in as a variable to what the actual outcome
in terms of enforcing the stopping rule was?
Professor Donnelly: No, that was
not included explicitly. One of the difficulties in analysing
any of the culling data, whether it be reactive or proactive,
is that we have the number of badgers which were culled but we
cannot know directly the number of badgers which were there which
were targeted and were not culled. While in the ideal world it
would have been very helpful, I am sure, if we could have included
the proportion of targeted badgers which had been culled we could
not include that. As had been previously outlined before doing
this analysis, our primary analysis was a direct comparison of
incidence rates. It was as the strategies were implemented, so
a sort of warts and all real world approach which I think is sensible
when you are doing a trial of policies.
Q52 Mr Drew: What you are saying
is even though you would have done some tracking of how many badgers
there were in reactive areas, you would not know what percentage
of those badgers were caught and were culled and it would not
have made any difference anyway?
Professor Donnelly: It may well
have made a difference, I would argue that we could not precisely
say that particularly dealing with it on a very small scale which
some of the reactive culls are. There has been work funded by
Defra—which the Defra people may wish to speak to in more
detail—to look at the estimation of badger numbers on the
basis of signs that they make, for example looking not only at
the number of setts including main setts but also various features,
latrines and so on. That gives a rough estimate of how many badgers
are there because it gives things which correlate with more badgers.
They do not provide for any specific farm, an estimate of the
number of badgers which are there. We cannot for any particular
culling operation say precisely how effective it has been. Now
there are post cull surveys which are undertaken, Chris Cheeseman
in particular does a number of them. He is really looking at very
broad scale differences. Those are carried out on proactives:
is there evidence that there are fewer badgers there, look to
see if setts are active or inactive but, unfortunately, I do not
think it is possible under any approach to say precisely how many
badgers are taken.
Q53 Mr Drew: If I could look at the
third part of this and I am sorry I am going into the results
but I think this is absolutely crucial. To what extent have you
factored in also illegal activity both in terms of farmers culling
badgers in the trial areas and also activists interfering with
the trials? Is there any statistical study which shows this is
of any significance whatsoever or is this something you have effectively
factored out?
Professor Donnelly: We have not
included any measure of illegal culling in this analysis. We did
look to see whether or not the level of compliance—that
is the extent to which farmers had agreed with survey and cull
in different areas—was associated with more or less effectiveness
of the treatments and we found no evidence of that. We did not
include a measure of illegal culling in that. I do not believe
we have any rigorous measure of that.
Q54 Mr Drew: Can I just look at this.
This is hypothesis. Let us say there is a substantial amount of
illegal culling that has not been statistically analysed, that
is bound in itself to add to the perturbation impact on badgers.
What validation is there in terms of the work that you have been
doing given that that is a perfectly reasonable hypothesis? At
the same time there may be activist interference, letting badgers
out of traps and so on, and that would impact, also, seriously
on the trials. Where are we in terms of those hypotheses?
Professor Bourne: I think there
are a number of points one can make here. You must accept that
what we are doing is comparing the number of cattle herd breakdowns
in reactive areas with survey only areas where there is no legal
badger intervention. One is able to compare the number of breakdowns
in survey only areas at the start of reactive culls and also continue
to record these during the reactive culling periods. The analyse,
in fact, did look at the incidence of breakdowns in reactive areas
assuming that they have behaved and were treated as survey only
areas throughout the period under which they were being tracked.
Maybe you wish to comment on that analysis. It simply found that
if one compared the number of breakdowns, if we had continued
as survey only and compared it with the number of breakdowns as
a result of reactive treatment in every case the situation was
worse.
Q55 Mr Drew: I understand, what you
are saying is that you would compare with areas outside and if
there is not the same incidence of TB one presumes then there
is some other variable. I am just putting it to you that this
to me would have been something worthy of study in its own right.
Clearly it is something you have to keep incredibly confidential
because it could invalidate your results but I think it is important
that we understand—
Professor Bourne: It could change
the results.
Q56 Mr Drew: —what impact activity
on either side could have on the validity of the results.
Professor Bourne: The only handle
we have on illegal culling is through the surveying that we carry
out. This surveying has been conducted throughout the trial area,
including the survey only areas, and the indication from that
is there is no high level of illegal culling. This is an imprecise
tool we accept that. I think you would also accept that if perturbation
is an issue, then the fact perturbation is the result of illegal
culling going on within the survey only areas, which would increase
the incidence of cattle TB, that would dampen down the impact
that we are measuring in the reactive areas as a result of the
illegal culling of badgers and the perturbation that is being
created. You are right, we cannot be totally definite about the
impact of illegal culling, and we recognised that from the outset.
We have tried to accommodate this by continually surveying those
areas using the crude tools we have relating badger activity to
an estimate of badger numbers and looking at whether setts are
dead or whether they are not dead which is exactly the same approach
we have used to give us an estimate of the level of trapping efficiency
in proactive areas immediately after we have carried out a proactive
cull. The data show in almost all cases there is a significant
reduction in badger activity, I say almost all, following culling.
It is true in some proactive areas which were culled in January
that we saw a limited reduction in badger numbers which was reflected
in the number of badgers we caught so the trapping efficiency
is influenced by the time of year but we have every reason to
believe that the rigour used by the wildlife unit in trapping
proactive areas is the same as they use in reactive areas.
Q57 Mr Drew: Minister?
Mr Bradshaw: I simply want—and
John has really answered it—to reassure the Committee that
I asked exactly these questions and specifically the one how do
we know that the survey only areas are safe and there is not a
lot of illegal culling going on? John reassured me, as he has
said, that they are studied and they are surveyed, they are part
of the trial. They do not just sit there and as far as they could
discover there was only very limited illegal culling going on
which was insignificant and certainly, as he has stressed, would
not have had any impact on the difference between the TB breakdown
levels in the reactive areas and in the survey only areas.
Q58 David Taylor: Can we develop
a little bit more an examination of the experimental rigour, Chairman.
The Krebs report did recognise the potential perturbation effect
of partial removal and said that the reactive strategy should
aim to remove all badgers from all social groups, did it not?
The ISG, and your first report, suggested that by a combination
of snaring and cage trapping the capture rate of 80% with cage
trapping alone could be driven up to between 90 and 100%. You
recognised that but you chose not to implement a more rigorous
control strategy involving snares?
Professor Bourne: That is true.
Q59 David Taylor: Why was that?
Professor Bourne: These reasons
are laid out in some length in our first report and immediately
after Krebs there was a public consultation. It was on the basis
of public consultation and the scientific evidence available to
us on the use of snares that we advised that we restrict ourselves
to cage trapping recognising that we would not eliminate every
badger. This was also following the line that was laid down to
us by the then Minister of Agriculture, Dr Cunningham, that elimination
of the badgers from large tracks of the countryside was not a
policy option. It follows, also, of course, the Bern Convention
with respect to the elimination of badgers. I remind you we are
testing a policy option of sustainability and this was why we
restricted ourselves to cage trapping with all the reservations
we have and did express about not removing 100% of badgers. Krebs
in his initial proposal was envisaging 100% removal on an experimental
basis. Clearly we could not do that because we were trialling
a sustainable policy option and that is the course we have taken.
| 
