Occupation and cancer in Britain.

Occupation and cancer in Britain
L Rushton*,1, S Bagga2, R Bevan2, TP Brown3, JW Cherrie4, P Holmes2, L Fortunato1, R Slack2,
M Van Tongeren3, C Young3 and SJ Hutchings1
1Department of Epidemiology and Biostatistics, Faculty of Medicine, Imperial College London, St Mary’s Campus, Norfolk Place, London W2 1PG, UK;
2Institute of Environment and Health, Cranf ield Health, Cranf ield University, Cranf ield, Bedfordshire MK43 0AL, UK; 3Health and Safety Laboratory,
Mathematical Sciences Unit, Harpur Hill, Buxton, Derbyshire SK17 9JN, UK; 4Institute of Occupational Medicine, Research Avenue North, Riccarton,
Edinburgh EH14 4AP, UK
BACKGROUND: Prioritising control measures for occupationally related cancers should be evidence based. We estimated the
current burden of cancer in Britain attributable to past occupational exposures for International Agency for Research on Cancer
(IARC) group 1 (established) and 2A (probable) carcinogens.
METHODS: We calculated attributable fractions and numbers for cancer mortality and incidence using risk estimates from the literature
and national data sources to estimate proportions exposed.
RESULTS: 5.3% (8019) cancer deaths were attributable to occupation in 2005 (men, 8.2% (6362); women, 2.3% (1657)). Attributable
incidence estimates are 13 679 (4.0%) cancer registrations (men, 10 063 (5.7%); women, 3616 (2.2%)). Occupational attributable
fractions are over 2% for mesothelioma, sinonasal, lung, nasopharynx, breast, non-melanoma skin cancer, bladder, oesophagus, soft
tissue sarcoma, larynx and stomach cancers. Asbestos, shift work, mineral oils, solar radiation, silica, diesel engine exhaust, coal tars
and pitches, occupation as a painter or welder, dioxins, environmental tobacco smoke, radon, tetrachloroethylene, arsenic and strong
inorganic mists each contribute 100 or more registrations. Industries and occupations with high cancer registrations include
construction, metal working, personal and household services, mining, land transport, printing/publishing, retail/hotels/restaurants,
public administration/defence, farming and several manufacturing sectors. 56% of cancer registrations in men are attributable to work
in the construction industry (mainly mesotheliomas, lung, stomach, bladder and non-melanoma skin cancers) and 54% of cancer
registrations in women are attributable to shift work (breast cancer).
CONCLUSION: This project is the first to quantify in detail the burden of cancer and mortality due to occupation specifically for Britain.
It highlights the impact of occupational exposures, together with the occupational circumstances and industrial areas where
exposures to carcinogenic agents occurred in the past, on population cancer morbidity and mortality; this can be compared with the
impact of other causes of cancer. Risk reduction strategies should focus on those workplaces where such exposures are still occurring.
British Journal of Cancer (2010) 102, 1428 – 1437. doi:10.1038/sj.bjc.6605637 www.bjcancer.com
& 2010 Cancer Research UK
Keywords: occupation; cancer burden; attributable fraction; industry sector; carcinogen
������������������������������������������������������������������
Reduction of occupationally related cancers requires sound
evidence on which to base priorities. Recognition and classifica-
tion of a substance as a carcinogen is made through scrutiny and
assessment of a wide range of evidence including in vivo, in vitro
and human studies. For substances that are already established as
human carcinogens, for example using the International Agency
for Research on Cancer (IARC) classification, estimation of
attributable burden provides useful indicators of the contribution
of different risk factors and has become widely used in public
health research (Lopez et al, 2006). Risk reduction can thus
incorporate consideration of risk level with the size of potentially
exposed populations. A comprehensive review of the proportions
of cancer from different causes in the United States estimated, the
contribution of occupational factors as 4% with an uncertainty range
of 2–8% (Doll and Peto, 1981). These estimates have been used to
formulate occupational health policies in many countries, including
Britain. More recently Doll and Peto (2005) have also produced an
estimate for Britain of 2% with a range of 1–5%; they suggested that
less than 1% is avoidable by practicable ways (Doll and Peto, 2005).
Budget rationing based on underestimates of the cancer burden
attributed to occupation will almost certainly overlook important
issues. Conversely, overestimation of the cancer burden may result
in tighter regulation, which may impede industry. Therefore, it is
important to develop rigorous methodology to estimate cancer
burden, which is adaptable to a given country.
The attributable fraction (AF) is widely used, for example, in the
estimation of global burden of disease (Driscoll et al, 2005);
quantification and ranking of the burden by diseases and causes
of disease facilitate decision making for risk reduction measures.
Our paper presents an evaluation of the burden of cancer in Britain
for all carcinogenic agents and occupations classified by IARC as a
Received 11 November 2009; revised 2 March 2010; accepted 9 March
2010
*Correspondence: Dr L Rushton; E-mail: l.rushton@imperial.ac.uk
British Journal of Cancer (2010) 102, 1428 – 1437
& 2010 Cancer Research UK All rights reserved 0007 – 0920/10 $32.00
www.bjcancer.com
E
p
id
em
io
lo
gy

group 1 (established) or 2A (probable) carcinogen that, for
occupational exposures, had either ‘strong’ or ‘suggestive’ evidence
of carcinogenicity in humans for the specific cancer site, as defined
by Siemiatycki et al (2004) and subsequent IARC publications
(Rousseau et al, 2005; Straif et al, 2005, 2007). We identify
carcinogens, occupations and industry sectors that make an
important contribution to the total burden.
MATERIALS AND METHODS
Preliminary results for six cancer sites, showing the developing
methodology have been previously published (Rushton et al,
2008). For this paper, estimations were carried out using 2005 data
for mortality and 2004 for cancer incidence. Mortality data were
obtained from the Office for National Statistics (ONS), and the
General Register Office for Scotland. Cancer incidence data were
obtained from ONS, Cancer Statistics, Registrations, Series MB1
for England, the Scottish Cancer Registry, and the Welsh Cancer
Intelligence and Surveillance Unit.
We estimated the AF, that is, the proportion of cases that would
not have occurred in the absence of an occupational exposure; this
was then used to estimate the attributable numbers (ANs). There
are several methods for estimating the AF but all depend on
knowledge of the disease risk due to the exposure of interest and
the proportion of the target population exposed (Steenland and
Armstrong, 2006).
Risk estimates were obtained from key studies, meta-analyses or
pooled studies, taking into account quality (including relevance
to Britain, sample size, extent of control for confounders, adequacy
of exposure assessment and clarity of case definition). Where
possible we selected risk estimates adjusted for important
confounders or non-occupational risk factors, for example,
smoking for lung cancer, smoking and alcohol use for laryngeal
cancer. Where only a narrative review was available giving a range
of risk estimates, we calculated a combined estimate of the relative
risks (RRs) based on a random-effects (for heterogeneous RRs) or
fixed-effects (for homogeneous RRs) model. Formal systematic
reviews and meta-analyses were carried out to estimate risk
estimates for laryngeal and stomach cancers related to asbestos
exposure.
Dose–response risk estimates were generally not available in the
epidemiological literature nor were proportions of those exposed
at different levels of exposure over time available for the working
population in Britain. However, where possible risk estimates were
obtained for an overall ‘lower’ level and an overall ‘higher’ level of
exposure to the agents of concern. The risk estimates for
occupational exposure to ionising radiation were derived using
generalised linear dose–response models of excess RR per unit
of cumulative radiation dose from the United Nations
Scientific Committee on the Effects of Atomic Radiation
(UNSCEAR, 2006). Cumulative lifetime dose was estimated using
data from the Central Index of Dose Information (CIDI) (HSE,
1998). For aircrew, who are not covered by CIDI, the mean total
lifetime radiation dose per pilot was obtained from a large cohort
study of European airline pilots (Langner et al, 2004) and
combined with numbers employed obtained from the British
Airways Stewards and Stewardesses Union (BASSA, personal
communication, 2008).
A substantial proportion of the excess is likely to occur in the
large number of workers with low exposures for whom our
estimates of average risks are inevitably unreliable. Where no risk
estimate could be identified for very low or background levels of
exposure, we estimated an RR for the ‘lower exposed’ group by
(1) taking the harmonic mean of all the available ratios of ‘higher’
to ‘lower’ RR estimates for cancer-exposure pairs for which data
were available and (2) applying this average ratio to the ‘higher’
level estimates to obtain ‘lower’ level RR estimates; if this was less
than one, it was set to one. (In the developmental phase of the
study, an RR of one was arbitrarily assigned, giving zero AF and so
possibly underestimating the burden; a large number of people
exposed at low levels associated with a low risk of disease may
contribute more to the burden than a small number exposed at
high levels associated with a high risk.)
We defined the period during which the relevant exposure for
the cancer in the target year 2005 as the risk exposure period
(REP). For solid tumours, a latency of 10–50 years was assumed
giving an REP of 1956–1995; for haematopoietic neoplasms, 0–20
year’s latency was assumed giving an REP of 1986–2005. The
proportion of the population ever exposed to each carcinogenic
agent or occupation in the REP was obtained from the ratio of the
numbers ever exposed to the carcinogens of interest in each
relevant industry or occupation within Britain over the total
number of people ever employed (Equation (4) in the Statistical
Appendix).
If the study from which the risk estimates were obtained
was population based, an estimate of the proportion of the
population exposed was derived directly from the study data,
although such studies were rarely available for Britain. If the risk
estimate was obtained from an industry-based study, national data
sources such as the CARcinogen EXposure (CAREX) database
(Pannett et al, 1998), the UK Labour Force Survey (LFS) (LFS,
2009) and the Census of Employment (ONS, 2009) were
used. CAREX was used for estimating the numbers of the British
population ever exposed to a carcinogen by industry sector. As
highlighted above, data are not available on the levels of exposure
in all industry sectors for all the carcinogens considered, nor the
numbers exposed at these levels. Industry sectors were therefore
allocated to ‘higher’ or ‘lower’ exposure categories assuming
distributions of exposure and risk that corresponded broadly
to those of the studies from which the risk estimates were
selected. The initial allocations were based on the judgement of
an experienced human exposure scientist; each assessment
was then independently peer reviewed and if necessary, a
consensus assessment agreed. Data from CAREX are not
differentiated by sex; 1991 Census data by industry and occupation
were used to estimate the relative proportions of men and women
exposed. The LFS and Census of Employment data were used to
estimate numbers ever employed in specific occupations, for
example, welder, painter and so on, and for specific industries for
carcinogens not included in CAREX.
CAREX data for Britain relate only to the period 1990–1993. For
the LFS and CoE, an available year was chosen to represent
numbers employed about 35 years before the target year of 2005,
as this was thought to represent a ‘peak’ latency for the solid
tumours, and is also close to the mid-point of the REP for
estimating numbers ever exposed across the period (for which a
linear change in employment levels was implicitly assumed).
Where the Census of Employment was used, the data are for 1971.
Where the LFS was used, the first year available and therefore used
was 1979 for solid tumours, and 1991 for short latency cancers.
When CAREX data were used, adjustment factors were applied to
take account of the change in numbers employed in the primary
and manufacturing industry and service sectors in Britain over the
REP. Adjustment for employment turnover over the period for
grouped main industry sectors was also carried out (see Equation
(3) in the Statistical Appendix). Ideally this requires full national
starter and leaver data across the REP for all industry sectors.
In the absence of this quality of data, estimating turnover directly
using new starters in years within the REPs gives the best
approximation for the purpose of estimating those ever exposed.
This method estimates starters in the past year as a proportion
of the average number employed (Gregg and Wadsworth, 2002).
As exposure in occupational epidemiological studies is usually
defined as for at least 1 year, we have adapted this to exclude short-
term labour turnover by taking new starters in the past year who
Occupational cancer burden in GB
L Rushton et al
1429
British Journal of Cancer (2010) 102(9), 1428 – 1437& 2010 Cancer Research UK
E
p
id
em
io
lo
gy

are expected to remain employed for at least 1 year as a proportion
of all those expected to be employed for at least 1 year. This is
estimated as the number recorded as employed for between 1 and 2
years divided by the total employed for at least 1 year using
LFS data averaged over the REP.
Statistical analysis
To estimate the AFs for each cancer and occupational carcinogen,
we used Levin’s method if risk estimates came from an industry-
based study, a review or meta-analysis, together with estimates
of the proportion of the population exposed from independent
national sources of data (Levin, 1953). Miettinen’s method was
used if risk estimates and proportions exposed came from a
population-based study (Miettinen, 1974) (Equations (1) and (2),
respectively, in the Statistical Appendix). For each AF, a random
error confidence interval was calculated using Monte Carlo
simulations (Steenland and Armstrong, 2006). The AFs were
applied to total numbers of cancer-specific deaths (2005) and
cancer registrations (2004) for ages that could have been exposed
during the REP to give ANs. Where risk estimates were only
available from mortality studies, AFs derived from these were used
for estimation of attributable registrations and vice versa. Similarly
if separate AFs for women could not be estimated, those for men or
for men and women combined were used.
The AF for mesothelioma was derived directly from several UK
mesothelioma studies that suggest 96–98% of male mesothelioma
cases are due to occupational or paraoccupational (e.g., exposure
from living near an asbestos factory or handling clothes
contaminated due to occupational exposure) exposure (Howel
et al, 1997; Yates et al, 1997; Rake et al, 2009). Combining the
results from Rake et al (2009) with those from two studies in which
results were reported separately for females (Spirtas et al, 1994;
Goldberg et al, 2006) gave estimates of 75–90% for females. The
ratio of asbestos-related lung cancer to mesothelioma deaths has
been suggested to be between two-thirds and one (Darnton et al,
2006). Rather than our standard method for the estimation of
numbers of lung cancers attributable to asbestos, we therefore used
a ratio of 1 : 1 mesothelioma to lung cancer deaths; this takes into
account the impact that past levels of asbestos exposure are having
on current incidence by the direct link to mesothelioma deaths
that are still climbing whereas lung cancer in general is declining
due to the reduction in smoking. This assumes, however, that
lung cancer has a similar pattern of latency as mesothelioma.
For lung cancer associated with radon exposure from natural
sources, estimates of rates of lung cancer due to exposure to radon
in domestic buildings (NRPB, 2000) were applied to estimates
of the time employees spend in workplaces where radon exposure
occurs.
AFs for all the relevant carcinogenic agents and occupational
circumstances were combined into a single estimate of AF for each
separate cancer. To take account of potential multiple exposures,
we used strategies including partitioning exposed numbers
between overlapping exposures or estimating only for the
‘dominant’ carcinogen with the highest risk. The IARC Monograph
process has been taking place over many years and has resulted in
overlap between substances evaluated. For lung cancer, for
example, 32 occupations or carcinogenic agents are evaluated.
We estimated AFs for 21 of these; for example, substances such as
coal tars and pitches and processes such as coal gasification and
coke production were included within our evaluation of polycyclic
aromatic hydrocarbons (PAHs). Where exposure to multiple
carcinogens remained, it was assumed that the exposures were
independent of one another and that their joint carcinogenic
effects were multiplicative. The AFs were then combined to give an
overall AF for that cancer using a product sum (Equation (5) in the
Statistical Appendix). An overall AF for all cancers was estimated
by summing the ANs for each and dividing by the total number of
cancers in Britain.
RESULTS
The overall burden by cancer site (AFs, ANs and 95% confidence
intervals) is given in Table 1. In all, 8.2% (n¼ 6362) of cancer
deaths in 2005 in men and 2.3% (n¼ 1657) in women in Britain
have been estimated to be due to occupation giving an overall AF
of 5.3% (n¼ 8019). The combined AFs for registrations are 5.7%
(n¼ 10 063) for men in 2004 and 2.2% (n¼ 3616) for women
giving an overall AF based on registrations of 4.0% (n¼ 13 679).
These are lower than for deaths because of the large numbers
of non-melanoma skin cancer (NMSC). The results for four of the
cancers, bladder, leukaemia, NMSC and sinonasal, are lower
than previously estimated (Rushton et al, 2008) mainly due to
reallocation of some of the industry sectors from ‘higher’ to ‘lower’
exposure categories after more in-depth review of the exposures in
Britain. If only agents and occupations classified by IARC as group
1 and having strong evidence of carcinogenicity in humans are
considered, the overall burden reduces to 4.0% (5123 total deaths,
8277 total registrations) (Supplementary Table A1). Only nine
cancer sites are involved (bladder, larynx, leukaemia, liver, lung,
mesothelioma, NMSC, sinonasal and thyroid). The dominance of
asbestos exposure and mesothelioma, asbestos and the many other
group 1 carcinogens relevant to lung cancer, and solar radiation
(SR) and NMSC means that the reduction in the AFs and ANs
for men (6.6%, 5123 deaths, 8277 registrations) is far less than for
women (1.2%, 862 deaths, 1313 registrations) for whom shift work
is most dominant.
The AFs by cancer site range from less than 0.01–95% overall,
the most important cancer sites for occupational attribution being,
for men, mesothelioma (97%), sinonasal (46%), lung (21.1%),
bladder (7.1%) and NMSC (7.1%), and for women, mesothelioma
(83%), sinonasal (20.1%), lung (5.3%), breast (4.6%) and
nasopharynx (2.5%). Occupation also contributes 2% or more
overall to cancers of the larynx, oesophagus, soft tissue sarcoma
(STS) and stomach, with in addition for men melanoma of the eye
(due to welding) and non-Hodgkin’s lymphoma (NHL).
Table 2 gives the number of registrations in 2004 by cancer
site for each cancer attributable to each of the 41 agents and
occupations for which separate estimation was carried out,
considered in rank order. Of these, 15 contributed over 100 total
cancer registrations, the largest being asbestos exposure
(mesothelioma (1937), lung (2223), larynx (8) and stomach
cancers (47)), followed in order by shift work, including flight
personnel (breast (1969), mineral oils (bladder (296), lung (470),
NMSC (902), sinonasal (63)), SR (NMSC (1541)), silica (lung
(907)), diesel engine exhaust (DEE) (lung (695), bladder (106)),
PAHs from coal tar and pitches (NMSC (545)), occupation as a
painter (bladder (71), lung (282), stomach (5)), dioxins (lung
(215), NHL (74), STS (27)), environmental tobacco smoke (ETS) at
work in non-smokers (lung (284)), radon exposure from natural
exposure in workplaces (lung (209)), occupation as a welder (lung
(175), and melanoma of the eye due to UV radiation (6)),
tetrachloroethylene (cervix (18), NHL (17), oesophagus (130)),
arsenic (lung (129)) and strong inorganic acid mists (larynx (46),
lung (76)). The results in this table highlight the fact that many
carcinogenic exposures in the workplace can affect multiple
cancer sites.
Table 3 gives numbers of registrations within industry sectors
or occupations for which there were at least 50 total attributable
registrations. The exposures concerned in each industry sector are
listed, with those contributing most (at least a total of 10 cancers)
being shown in bold. Painters and welders are assumed to be
exposed to many different carcinogens. Also given are the relevant
cancer sites for each industry sector with the most important
Occupational cancer burden in GB
L Rushton et al
1430
British Journal of Cancer (2010) 102(9), 1428 – 1437 & 2010 Cancer Research UK
E
p
id
em
io
lo
gy

T
ab
le
1
Es
tim
at
ed
at
tr
ib
ut
ab
le
fr
ac
tio
ns
,d
ea
th
s
an
d
re
gi
st
ra
tio
ns
by
ca
nc
er
sit
e
in
20
05
(d
ea
th
s)
an
d
20
04
(r
eg
ist
ra
tio
ns
)
A
tt
ri
b
u
ta
b
le
fr
ac
ti
o
n
(%
)
(9
5%
co
n
fi
d
en
ce
in
te
rv
al
)
A
tt
ri
b
u
ta
b
le
n
u
m
b
er
s
(9
5%
co
n
fi
d
en
ce
in
te
rv
al
)
D
ea
th
s
(2
00
5)
R
eg
is
tr
at
io
n
s
(2
00
4)
C
an
ce
r
si
te
IC
D
-1
0
co
d
e
M
al
e
F
em
al
e
T
o
ta
l
(b
as
ed
o
n
d
ea
th
s)
M
al
e
F
em
al
e
T
o
ta
l
M
al
e
F
em
al
e
T
o
ta
l
Bl
ad
de
r
C
67
7.
1
(4
.6
,9
.7
)
1.
9
(1
.3
,3
.9
)
5.
3
(3
.4
,7
.7
)
21
5
(1
39
,2
96
)
30
(2
1,
62
)
24
5
(1
59
,3
58
)
49
6
(3
21
,6
84
)
54
(3
7,
11
0)
55
0
(3
57
,7
95
)
Bo
ne
C
40
–
C
41
0.
04
0.
01
0.
02
0
0
0
0
0
0
Br
ai
n
C
70
–
C
72
0.
5
(0
.1
,1
.1
)
0.
1
(0
,0
.2
)
0.
4
(0
.0
,0
.7
)
10
(1
,2
0)
1
(0
,3
)
11
(1
,2
3)
12
(1
,2
5)
2
(0
,4
)
14
(1
,2
8)
Br
ea
st
C
50
4.
6
(3
.3
,6
.0
)
4.
6
(3
.3
,6
.0
)
55
5
(3
97
,7
27
)
55
5
(3
97
,7
27
)
19
69
(1
40
7,
25
79
)
19
69
(1
40
7,
25
79
)
C
er
vi
x
C
53
0.
7
(0
.0
,2
.1
)
0.
7
(0
.0
,2
.1
)
7
(0
,2
2)
7
(0
,2
2)
18
(1
,5
6)
18
(1
,5
6)
Ki
dn
ey
C
64
–
C
66
,C
68
0.
04
(0
,0
.1
6)
0.
04
(0
,0
.1
4)
0.
04
(0
,0
.1
5)
1
(0
,3
)
1
(0
,2
)
1
(0
,5
)
2
(0
,7
)
1
(0
,4
)
3
(0
,1
0)
La
ry
nx
C
32
2.
9
(1
.4
,5
.7
)
1.
6
(0
.6
,3
.5
)
2.
6
(1
.2
,5
.2
)
17
(8
,3
4)
3
(1
,6
)
20
(9
,4
0)
50
(2
4,
99
)
6
(2
,1
3)
56
(2
6,
11
2)
Le
uk
ae
m
ia
a
C
91
–
C
95
0.
9
(0
.2
,3
.5
)
0.
5
(0
.1
,4
.5
)
0.
8
(0
.2
,3
.9
)
18
(4
,7
1)
6
(1
,4
9)
24
(5
,1
20
)
31
(7
,1
18
)
9
(1
,8
0)
40
(8
,1
99
)
Li
ve
r
C
22
0.
2
(0
.1
,0
.3
)
0.
1
(0
.1
,0
.2
)
0.
2
(0
.1
,0
.3
)
4
(2
,6
)
2
(1
,2
)
5
(3
,8
)
4
(2
,6
)
1
(1
,2
)
5
(3
,8
)
Lu
ng
C
33
–
C
34
21
.1
(1
9.
2,
24
.7
)
5.
3
(4
.3
,6
.9
)
14
.5
(1
3.
0,
17
.2
)
40
24
(3
65
9,
46
96
)
72
6
(5
92
,9
46
)
47
49
(4
25
1,
56
43
)
46
32
(4
21
2,
54
06
)
81
6
(6
66
,1
06
3)
54
47
(4
87
7,
64
69
)
Ly
m
ph
o-
ha
em
at
op
oi
et
ic
C
81
–
C
96
0.
00
4
(0
,0
.0
14
)
0.
00
2
(0
,0
.0
07
)
0.
00
3
(0
,0
.0
11
)
0
(0
,1
)
0
(0
,0
)
0
(0
,1
)
0
(0
,1
)
0
(0
,0
)
1
(0
,2
)
M
el
an
om
a
(e
ye
)
C
69
2.
9
(0
.6
,6
.6
)
0.
4
(0
.1
,1
.0
)
1.
6
(0
.3
,3
.6
)
1
(0
,3
)
0
(0
,0
)
1
(0
,3
)
6
(1
,1
3)
1
(0
,2
)
6
(1
,1
6)
M
es
ot
he
lio
m
a
C
45
97
.0
(9
6.
0,
98
.0
)b
82
.5
(7
5.
0,
90
.0
)b
94
.9
(9
3.
0,
96
.9
)b
16
99
(1
68
1,
17
17
)
23
8
(2
16
,2
60
)
19
37
(1
89
8,
19
76
)
16
99
(1
68
1,
17
17
)c
23
8
(2
16
,2
60
)c
19
37
(1
89
8,
19
76
)c
M
ul
tip
le
m
ye
lo
m
a
C
90
0.
4
(0
,1
.0
)
0.
1
(0
,0
.3
)
0.
3
(0
,0
.7
)
5
(0
,1
0)
1
(0
,2
)
6
(0
,1
2)
8
(0
,1
8)
2
(0
,3
)
10
(0
,2
1)
N
as
op
ha
ry
nx
C
11
11
.0
(2
.3
,4
7.
9)
2.
5
(0
.6
,6
.8
)
8.
2
(1
.8
,3
4.
3)
7
(2
,3
2)
1
(0
,2
)
8
(2
,3
3)
14
(3
,6
1)
2
(0
,4
)
16
(3
,6
5)
N
H
L
C
82
–
C
85
2.
1
(0
,6
.9
)
1.
1
(0
.1
,2
.9
)
1.
7
(0
,5
.4
)
43
(0
,1
38
)
14
(1
,3
7)
57
(1
,1
76
)
10
2
(0
,3
28
)
39
(3
,1
01
)
14
0
(3
,4
30
)
N
M
SC
d
C
44
7.
1
(1
.3
,1
5.
1)
1.
1
(0
.0
,2
.9
)
4.
6
(0
.8
,1
0.
0)
21
(4
,4
4)
2
(0
,6
)
23
(4
,5
0)
25
76
(4
81
,5
47
5)
35
2
(0
,9
00
)
29
28
(4
81
,6
37
5)
O
es
op
ha
gu
s
C
15
3.
3
(1
.5
,7
.5
)
1.
1
(0
.3
,2
.8
)
2.
5
(1
.1
,5
.9
)
15
6
(7
0,
35
8)
28
(8
,7
0)
18
4
(7
8,
42
9)
15
9
(7
1,
36
5)
29
(9
,7
4)
18
8
(8
0,
43
9)
O
va
ry
C
56
0.
5
(0
,1
.2
)
0.
5
(0
,1
.2
)
23
(0
,5
2)
23
(0
,5
2)
33
(0
,7
6)
33
(0
,7
6)
Pa
nc
re
as
C
25
0.
02
(0
,0
.0
7)
0.
01
(0
,0
.0
4)
0.
01
(0
,0
.0
5)
1
(0
,2
)
0
(0
,1
)
1
(0
,4
)
1
(0
,2
)
0
(0
,1
)
1
(0
,4
)
Si
no
na
sa
l
C
30
–
C
31
46
.0
(2
7.
3,
74
.0
)
20
.1
(1
4.
4,
31
.6
)
34
.4
(2
1.
5,
54
.8
)
29
(1
7,
47
)
10
(8
,1
6)
39
(2
5,
63
)
10
1
(6
0,
16
2)
32
(2
3,
50
)
13
3
(8
3,
21
2)
ST
S
C
49
3.
4
(0
,1
1.
4)
1.
1
(0
,3
.8
)
2.
4
(0
,8
.1
)
11
(0
,3
6)
3
(0
,9
)
13
(0
,4
5)
22
(0
,7
5)
4
(0
,1
5)
27
(0
,9
0)
St
om
ac
h
C
16
3.
0
(1
.5
,5
.1
)
0.
3
(0
.1
,0
.5
)
2.
0
(1
.0
,3
.4
)
10
2
(5
2,
17
6)
6
(3
,1
1)
10
8
(5
5,
18
7)
14
9
(7
7,
25
8)
9
(4
,1
5)
15
8
(8
1,
27
4)
T
hy
ro
id
C
73
0.
12
0.
02
0.
05
0
0
0
1
0
1
T
ot
al
ba
se
d
on
de
at
hs
8.
2
(7
.2
,9
.9
)
2.
3
(1
.7
,3
.2
)
5.
3
(4
.6
,6
.6
)
63
62
(5
64
6,
76
90
)
16
57
(1
24
9,
22
87
)
80
19
(6
88
8,
99
77
)
T
ot
al
ba
se
d
on
re
gi
st
ra
tio
ns
5.
7
(4
.0
,8
.5
)
2.
2
(1
.4
,3
.2
)
4.
0
(2
.7
,5
.9
)
10
06
3
(6
94
1
14
82
2)
36
16
(2
37
0,
54
13
)
13
67
9
(9
31
0,
20
23
5)
T
ot
al
ca
nc
er
s
in
G
B
in
ag
es
15
+
77
91
2
72
21
2
15
0
12
4
17
5
39
9
16
8
18
4
34
3
58
3
A
bb
re
vi
at
io
ns
:G
B
¼
G
re
at
Br
ita
in
;N
H
L
¼
no
n-
H
od
gk
in
’s
ly
m
ph
om
a;
N
M
SC
¼
no
n-
m
el
an
om
a
sk
in
ca
nc
er
;S
T
S
¼
so
ft
tis
su
e
sa
rc
om
a.
T
ot
al
s
do
no
ta
lw
ay
s
su
m
ac
ro
ss
ro
w
s
du
e
to
ro
un
di
ng
er
ro
r.
C
on
fid
en
ce
in
te
rv
al
s
no
te
st
im
at
ed
fo
r
ca
nc
er
s
at
tr
ib
ut
ed
to
io
ni
sin
g
ra
di
at
io
n,
as
th
ey
ar
e
no
t
ye
t
av
ai
la
bl
e
fo
r
th
e
ex
ce
ss
re
la
tiv
e
ris
k
m
od
el
s
us
ed
(U
N
SC
EA
R
,
20
06
).
a A
F
ap
pl
ic
ab
le
to
al
l
le
uk
ae
m
ia
s.
b
In
cl
ud
es
ca
se
s
de
sc
rib
ed
as
du
e
to
pa
ra
oc
cu
pa
tio
na
l
or
en
vi
ro
nm
en
ta
le
xp
os
ur
e
to
as
be
st
os
.c
T
ak
en
as
eq
ua
lt
o
at
tr
ib
ut
ab
le
de
at
hs
fo
r
th
is
sh
or
t
su
rv
iv
al
ca
nc
er
.d
Ba
se
d
on
re
gi
st
ra
tio
ns
.
Occupational cancer burden in GB
L Rushton et al
1431
British Journal of Cancer (2010) 102(9), 1428 – 1437& 2010 Cancer Research UK
E
p
id
em
io
lo
gy

T
ab
le
2
C
an
ce
r
re
gi
st
ra
tio
ns
in
20
04
at
tr
ib
ut
ab
le
to
oc
cu
pa
tio
n
by
ex
po
su
re
an
d
ca
nc
er
sit
es
w
ith
at
le
as
t
15
to
ta
la
tt
rib
ut
ab
le
re
gi
st
ra
tio
ns
C
an
ce
r
si
te
a
C
ar
ci
n
o
ge
n
o
r
o
cc
u
p
at
io
n
B
la
d
d
er
B
ra
in
B
re
as
t
C
er
vi
x
L
ar
yn
x
L
eu
-
ka
em
ia
L
u
n
g
M
es
o
th
e-
lio
m
a
N
as
o
-
p
h
ar
yn
x
N
M
S
C
N
H
L
O
es
o
-
p
h
ag
u
s
O
va
ry
S
in
o
n
as
al
S
T
S
S
to
m
ac
h
O
th
er
si
te
s
T
o
ta
l
re
gi
st
ra
ti
o
n
s
A
sb
es
to
s
8
22
23
19
37
47
42
16
Sh
ift
w
or
k
(in
cl
ud
in
g
fli
gh
t
pe
rs
on
ne
l)
19
69
19
69
M
in
er
al
oi
ls
29
6
47
0
90
2
63
17
30
So
la
r
ra
di
at
io
n
15
41
15
41
Si
lic
a
90
7
90
7
D
ie
se
le
ng
in
e
ex
ha
us
t
10
6
69
5
80
1
PA
H
s
fr
om
co
al
ta
rs
an
d
pi
tc
he
s
54
5
54
5
Pa
in
te
rs
71
28
2
5
35
9
T
C
D
D
(d
io
xi
ns
)
21
5
74
27
31
6
En
vi
ro
nm
en
ta
lt
ob
ac
co
sm
ok
e
(n
on
-s
m
ok
er
s)
28
4
28
4
R
ad
on
20
9
20
9
W
el
de
rs
17
5
17
5
T
et
ra
ch
lo
ro
et
hy
le
ne
18
17
13
0
16
4
A
rs
en
ic
12
9
12
9
St
ro
ng
in
or
ga
ni
c-
ac
id
m
ist
s
co
nt
ai
ni
ng
su
lp
hu
ric
ac
id
46
76
12
2
C
hr
om
iu
m
V
I
67
22
89
N
on
-a
rs
en
ic
al
in
se
ct
ic
id
es
11
19
33
M
M
(1
0)
73
C
ob
al
t
73
73
In
or
ga
ni
c
le
ad
2
42
23
67
A
ro
m
at
ic
am
in
es
66
66
H
ai
rd
re
ss
er
s
an
d
ba
rb
er
s
15
14
33
63
So
ot
s
60
60
W
oo
d
du
st
14
39
54
Le
at
he
r
du
st
31
31
St
ee
lf
ou
nd
ry
w
or
ke
rs
29
29
Fo
rm
al
de
hy
de
12
1
1
14
PA
H
s
7
4
11
N
ic
ke
l
10
0
10
C
ad
m
iu
m
9
9
Be
ry
lliu
m
7
7
T
ric
hl
or
oe
th
yl
en
e
3
Ki
dn
ey
(3
)
Li
ve
r
(2
)
7
Be
nz
en
e
7
7
U
V
ra
di
at
io
n
(w
el
de
rs
on
ly
)
M
el
an
om
a
(e
ye
)
(6
)
6
R
ub
be
r
in
du
st
ry
3
1
4
Io
ni
sin
g
ra
di
at
io
n
1
2
Bo
ne
(0
)
Li
ve
r
(0
)
T
hy
ro
id
(1
)
4
V
in
yl
ch
lo
rid
e
Li
ve
r
(3
)
3
T
in
m
in
er
s
2
2
1,
3-
Bu
ta
di
en
e
0
LH
(1
)
1
A
cr
yl
am
id
e
Pa
nc
re
as
(1
)
1
Et
hy
le
ne
ox
id
e
1
1
Pe
tr
ol
eu
m
re
fin
in
g
0
0
T
ot
al
re
gi
st
ra
tio
ns
at
tr
ib
ut
ab
le
to
oc
cu
pa
tio
n
55
0
14
19
69
18
56
40
54
47
19
37
16
29
28
14
0
18
8
33
13
3
27
15
8
26
13
67
9
T
ot
al
re
gi
st
ra
tio
ns
in
G
B
(2
00
4)
b
98
78
39
33
43
20
2
26
12
21
12
52
02
37
37
8
20
37
18
9
67
22
0
82
36
74
98
61
97
37
8
10
63
79
70
22
03
4c
33
9
15
6d
A
bb
re
vi
at
io
ns
:G
B
¼
G
re
at
Br
ita
in
;L
H
¼
ly
m
ph
oh
ae
m
at
op
oi
et
ic
ca
nc
er
s;
M
M
¼
m
ul
tip
le
m
ye
lo
m
a;
N
H
L
¼
no
n-
H
od
gk
in
’s
ly
m
ph
om
a;
N
M
SC
¼
no
n-
m
el
an
om
a
sk
in
ca
nc
er
;P
A
H
¼
po
ly
cy
cl
ic
ar
om
at
ic
hy
dr
oc
ar
bo
n;
ST
S
¼
so
ft
tis
su
e
sa
rc
om
a;
T
C
D
D
¼
2,
3,
7,
8-
T
et
ra
ch
lo
ro
di
be
nz
o
di
ox
in
;U
V
¼
ul
tr
av
io
le
t.
a B
la
nk
ce
lls
in
di
ca
te
th
at
at
tr
ib
ut
ab
le
ca
nc
er
re
gi
st
ra
tio
ns
w
er
e
no
t
es
tim
at
ed
fo
r
th
is
oc
cu
pa
tio
na
l
ex
po
su
re
.Z
er
o
re
pr
es
en
ts
an
es
tim
at
e
of
le
ss
th
an
0.
5.
b
R
eg
ist
ra
tio
ns
ag
ed
25
+
ye
ar
s
fo
r
so
lid
tu
m
ou
rs
,a
ge
d
15
–
84
ye
ar
s
fo
r
ha
em
at
op
oi
et
ic
ne
op
la
sm
s
fo
r
m
en
an
d
15
–
79
ye
ar
s
fo
r
ha
em
at
op
oi
et
ic
ne
op
la
sm
s
fo
r
w
om
en
;f
ig
ur
es
fo
r
m
es
ot
he
lio
m
a
ba
se
d
on
de
at
hs
.c
In
cl
ud
es
bo
ne
,
ki
dn
ey
,l
iv
er
,m
el
an
om
a
(e
ye
),
m
ul
tip
le
m
ye
lo
m
a,
pa
nc
re
as
,t
hy
ro
id
.d
A
ll
m
al
ig
na
nt
ne
op
la
sm
s.
Occupational cancer burden in GB
L Rushton et al
1432
British Journal of Cancer (2010) 102(9), 1428 – 1437 & 2010 Cancer Research UK
E
p
id
em
io
lo
gy