Previous page | Contents | Next page
17  |  Civilian casualties
commissioned, had been produced by Dr Bill Kirkup, one of the Department of Health’s
Regional Directors of Public Health and its lead on health in Iraq.
176.  Dr Kirkup’s assessment was more detailed and more critical of the Lancet
study than the assessments undertaken earlier by Professor Anderson and Mr Butler.
He stated:
“Less than a thousand [households] … is a small number on which to base death
rates … The confidence intervals are correspondingly very wide … A confidence
interval this large makes the meaning of the estimate very difficult to interpret …
“Cluster sampling may not be appropriate when there is a large element of
discontinuity in the population experience. Clearly, some parts of Iraq have seen
much more violence than others …”
177.  Dr Kirkup stated that, according to his calculations, the study’s conclusion that
“violence accounted for most of the excess deaths” was only true if the “bizarre” Fallujah
cluster was included (the study stated that that cluster was not included in its central
estimate of 98,000 excess deaths). Dr Kirkup calculated that if the Fallujah cluster
was not included, just over 23,000 of the 98,000 estimated excess deaths were due
to violence.
178.  Dr Kirkup stated that it was not possible, from the data provided in the study, to
confirm the study’s conclusion that “air strikes from coalition forces accounted for most
violent deaths”.
179.  Dr Kirkup explained his characterisation of the Fallujah projection as “bizarre”. The
study estimated that there had been 200,000 excess deaths in Fallujah (using the same
techniques as for other areas). That would represent a loss of nearly 28 percent of the
population of Fallujah in just 14 months. Dr Kirkup commented: “Something has plainly
gone so badly wrong with the estimates in Fallujah that it must cast doubt on the validity
of the rest of the findings.”
180.  Dr Kirkup concluded:
“… the paper suffers from wide confidence intervals, dubious methodology, the
likelihood of significant respondent bias and results that are disastrously skewed by
the Fallujah outlier. The authors have been tempted into extrapolations based on
shaky data that lack face validity, and in two cases are not even borne out by their
own results.”
201
Previous page | Contents | Next page