COMARE Statement on Childhood Cancers in North Wales
The Welsh Assembly Government (WAG) asked COMARE to consider a
paper produced by Green Audit [1] and reports from the Welsh Cancer
Intelligence and Surveillance Unit (WCISU) [2] and the National
Public Health Service (NPHS) for Wales [3]. WAG requested advice
from the Committee on these studies and in particular asked COMARE
to consider five questions. This statement contains COMARE's response.
The Committee noted that the Green Audit analyses had not undergone
peer review as recommended by the Committee in the past. Green
Audit analyses certainly contained some methodological shortcomings
as detailed in the WICSU report (for example, the use of different
time period for calculating numbers of observed and expected cases).
The Committee also had considerable reservations about the reliability
of self-reported data as collected by Green Audit and the mechanism
for case finding. Further, it appeared that Green Audit had conducted
a large number of analyses which leads to problems of multiple
significance testing. On a specific point, Green Audit assumes
that retinoblastoma cases were diagnosed over a ten year period.
Their assertion that this is conservative is incorrect.
The Committee also felt that the temporal pattern of the claimed
excesses of childhood cancer made a connection with Sellafield
discharges less plausible. The largest Sellafield discharges were
in the 1960s and 1970s. The original suggestion that the Sellafield
‘cluster’ of childhood cancer may be explained by
the preconception irradiation of fathers, has received little
subsequent support in the literature. However, even if there was
a putative effect of preconceptual exposure of fathers, increasing
the risk of their children developing childhood cancer, then any
excess would have to be seen shortly after exposure, ie, in the
1960s and 1970s for the association to be biologically plausible.
Nevertheless, some of the WCISU analyses result in relative risks
above one, even if they do not reach statistical significance.
This is particularly the case in the most recent period. The Committee
notes WCISU's proposals for further research on the question of
childhood cancer rates around the North Wales coastline and would
support such research if it were well conducted and directed at
a specific hypothesis.
The Committee also noted the very considerable discrepancy in
the numbers of cases of brain and spinal tumours reported by WCISU
and by HTV. The data collected by the cancer registries had undergone
extensive validation, including cross checks against national
records held by the Childhood Cancer Research Group, and are more
likely to be reliable. However, the Committee felt that, subject
to ethical and practical considerations, this discrepancy should
be resolved. It recommended that the list of cases accumulated
by HTV should, if at all possible, be compared with that set up
by WCISU and any differences investigated.
However, both these lines of investigation are likely to take
some time. In the interim, the Committee's views were as follows.
Question 1: Was the analysis and methodology adopted by
WCISU appropriate?
Yes, though the Committee noted the inescapable problems of epidemiological
studies involving small numbers of cases.
Question 2: Was the interpretation by NPHS Wales appropriate?
Yes, broadly speaking, though the Committee felt that further
investigation of levels of childhood cancer in the area would
be worth considering.
Question 3: Do the studies demonstrate or suggest a raised
incidence of childhood leukaemia, retinoblastoma or brain tumours
near the Menai Straits as alleged in the Green Audit report?
Top
