Protocol for a further study of cancer among the current and former employees of National Semiconductor (UK) Ltd, Greenock

EXECUTIVE SUMMARY
1.1 BACKGROUND AND AIM
This document is a study protocol for the further investigations that HSE wishes to undertake
into the risks of cancer among workers and ex-workers of the National Semiconductor UK
(NSUK) plant at Greenock in Scotland. It sets out what follow-up investigations we plan to
do and how we intend to do them, including how we intend to report the findings.
The new investigations follow on from an earlier HSE study of the occurrence of cancer in a
cohort study group of 4388 workers and ex-workers at the NSUK plant. That study found
some evidence that, in women, cancers of the lung and stomach were occurring more
frequently than would be expected in the general population, after taking account of the social
and economic factors associated with postcode of residence. There was also some suggestion
of excess breast cancer in women and brain cancer in men.
The present revised protocol follows on also from an extensive consultation carried out by
HSE on an earlier draft. That consultation included the independent Study Steering
Committee (SSC), National Semiconductor and its scientific advisors, the NSUK workforce
and Phase Two (Appendix 1).
In the light of that consultation, and of concurrent developments in other studies, the present
revised protocol describes linked studies of a simpler design than originally proposed.
However, the main purpose (aim) of the further investigations remains unchanged; it is to
help clarify whether the rates of any of these cancers in the NSUK workforce might be related
to work done there. As well as studying work done at NSUK, we will need to take into
account work done elsewhere, and non-work factors, to the extent that these are important risk
factors for the cancers of main interest in this further study. 1.2 UPDATING THE OVERALL COHORT STUDY GROUP
Results from the existing NSUK cohort of 4388 workers and ex-workers will be updated, to
give additional information about who has developed cancer in recent years. This will enable
us to make a further assessment of whether more cancers have occurred in the cohort than
would have been expected, and if so how many. Only a few years will have elapsed since the
earlier HSE study and so we expect that the new results will not be very different from those
already reported, except that now we will also study occurrence of cancer in relation to the
length of time that people were employed at NSUK and in addition brain cancer, brain and
CNS cancer combined and ovarian cancer will be included in the list of cancers examined.
Nevertheless, the update may provide some additional cases for inclusion in the case-only and
case-control studies which we now describe.
1.3 OVERVIEW OF FURTHER STUDIES OF WORK AND OCCURRENCE OF
SELECTED CANCERS
In the remainder of the further investigations, we will try to find out if work activities at
NSUK may have contributed to the risks of developing lung, breast or stomach cancer among
women, or brain cancer among men, in the workforce. Even though we will use wellestablished methods, these studies will be much more complex than the cohort update because
of the relatively small numbers of cases, the many possible risk factors that need to be taken
into account, and the difficulties of obtaining reliable information about them. This makes it
especially important that we get good co-operation from NSUK management, the workforce,
and the local community, including relatives and friends of former workers who have died, so
that the further studies can be as good as possible.
1
(Version 2.0.4a for HSE website – October 2007)
To help understand why the cancers occurred, we will carry out a set of separate
investigations. All of these will be based on (sometimes described as ‘nested within’) the
existing NSUK cohort of 4388 workers and ex-workers.
· Following detailed consultation, and careful assessment of whether further study is
really likely to be useful, we have decided to focus effort principally, but not
exclusively, on studying lung cancer and breast cancer in women. For each of these
two cancers, we will carry out a traditional, standard, case-control study. To do this,
we will need to collect further information about all the people who developed the
cancers of interest (the cases), and a sample of the rest of the cohort (the controls).
· For stomach cancer in women, and brain cancer in men, we will carry out more
limited investigations, of the cases only – there are simply too few of these cancers to
justify a full case-control study. These case-only studies will also involve collecting
new information.
· Because the indications of a cancer excess arose mostly in women, we will also
examine the similarities and differences in the work carried out by women and by
men, over the years at NSUK.
1.4 NESTED CASE-CONTROL STUDIES OF LUNG CANCER AND BREAST
CANCER, IN WOMEN
These two studies will examine whether work activities and other potential risk factors were
similar or different in people who developed these cancers (the ‘cases’) compared with a
sample of those who didn’t (the ‘controls’). For example, histories of exposure to certain
factors of interest (say work with a chemical or on a process) will be compared between the
cases and the controls. Both studies will be restricted to women – breast cancer, because this
is an issue for women in particular; lung cancer, because there was no evidence of an excess
in men. The risk of these cancers varies with age; we will design the study so all case-control
comparisons will be between people at the same age, so that comparisons are fair (i.e. not
biased).
To help in exposure assessment, we will carry out a separate but linked investigation – the
Historical Hygiene Assessment (HHA) – to describe the work activities and exposures
associated with various jobs at NSUK, in different time-periods. This will focus on jobs and
time-periods, not on individuals, and will include interviews with current or ex-workers who
have a long-standing familiarity with the work processes and exposures at NSUK over the
years. Information from the HHA will help us develop work history questionnaires, based on
job and work area, to be used by those attending at surveys when we gather information about
cases and controls. The information collected during the HHA will be kept confidential by
the study team until the final report is produced, and no information regarded as commercially
confidential will be published. (We describe this and other data gathering more fully below.)
Those work histories will later also be linked with detailed information from the HHA, to
derive estimates of exposures for individuals. That linkage will be carried out without
knowledge of the case-control status of individuals. Important non-work factors will also be
compared between cases and controls, to help ensure that effects of non-work factors are not
mistakenly attributed to activities and exposures at NSUK.
The results will be used to assess whether work at NSUK might have played a part in the
development of the disease, taking account of other factors. For instance, if a greater
proportion of cases than controls had the exposure or risk factor, then this would weigh in
favour of it being related to the occurrence of the disease. However, when many comparisons
are being made, it increases the likelihood of finding some differences that may have occurred
by chance. These possibilities will be considered carefully, using standard scientific
approaches.
2
(Version 2.0.4a for HSE website – October 2007)
1.5 A LIMITED EXAMINATION OF WORK HISTORIES OF CASES ONLY (NOT
CONTROLS) OF STOMACH CANCER IN WOMEN AND BRAIN CANCER IN
MEN
In two separate small studies, the work histories of cases of stomach cancer in women and
brain cancer in men will be examined, to see if there are common patterns in the of work
activities or exposures at NSUK of either set of cases. These small studies are case-only
rather than case-control studies because there are too few cases to be able to show reliably any
associations between exposures and occurrence of the relevant cancers. If, in a particular case
series, there are no exposures or activities in common that could even plausibly explain the
development of cancer, we will interpret this as evidence against the work-relatedness of the
small excesses found. If, conversely, there is some evidence of similarity, then we will
interpret this as something to investigate in larger, more powerful studies elsewhere than at
NSUK. There is no way, on the small numbers available, that such associations can be
regarded as proof of a work-related excess – the likelihood of such patterns occurring by
chance cannot be ruled out when small numbers of cases are involved.
1.6 A STUDY OF GENDER DIFFERENCES IN WORK ACTIVITIES AT NSUK
We will also study the similarities and differences between women and men in terms of their
work patterns at NSUK, to see if there are differences which might help explain why the
excess cancers occurred in one group or other, but not both. This will be done as part of the
Historical Hygiene Assessment (HHA). That means that it will be a study of general patterns
of gender differences in work activities and exposures at NSUK; it will not refer specifically
to cases and controls, almost all of whom will be women.
1.7 SUBJECTS FOR CASE-CONTROL AND CASE-SERIES STUDIES
The case-control and case-series studies will include all confirmed cases of the cancers of
interest among the identified NSUK cohort of 4388 workers and ex-workers, and any other
cases identified later who are confirmed (at least one month before the end of study data
collection at interviews) as falling within the cohort definition. To help make better
comparisons in the case-control studies of lung and breast cancer, we will choose controls of
similar ages to the cases. This means that the study will be focused more towards the older
people in the cohort, because the cases are older on average than the cohort as a whole.
Almost all the subjects will be women, the only men being those who have developed brain
cancer.
1.8 GATHERING NEW INFORMATION AT INTERVIEWS
A key piece of information will be the work histories of the cases and controls over the years
they worked at NSUK. Collecting these work histories will involve interviewing individual
cases and controls where practicable. Where somebody has died, or is otherwise not capable
of taking part in interviews, we will seek to gather corresponding information about their
work activities at NSUK by interviews with former work colleagues. The interviews will be
carried out by a member of the research team who is trained in survey methods generally and
in gathering work histories in particular. He (or she) will know about the workplace and the
jobs at NSUK but will not know about detailed exposures in the various occupations at
NSUK, or about which women developed cancer and who did not. This is to help ensure that
information is gathered in a fair and reliable way.
Information from the interviews will be combined with expert assessments of the exposures
that might have occurred in the course of work at NSUK – what we have earlier called the
Historical Hygiene Assessment (HHA). This expert assessment will be based on several
sources:
3
(Version 2.0.4a for HSE website – October 2007)
· whatever relevant records there are at NSUK, including standard risk assessments and
records of any exposure incidents, and statutorily compulsory ‘suitable and sufficient’
COSHH assessments;
· associated discussions with appropriate NSUK health and safety staff;
· special interviews and detailed discussions with long serving workers and ex-workers
who are most familiar with the processes at the plant over the years – for example,
supervisory staff.
The interviews with study subjects, or their relatives, during the main studies will also include
details of employment in other workplaces, including prior to employment at NSUK. This is
to allow an assessment of the exposure of the cases and controls to any known and important
occupational causes of the cancers of interest (especially asbestos, which can cause lung
cancer) where these had occurred in jobs other than at NSUK in Greenock.
It will also be important to know about the exposure of cases and controls to possible nonwork causes of the cancers of interest, so that the overall findings can be interpreted properly.
The most important risk factor is smoking tobacco, in relation to lung cancer. We propose
also to collect some information on issues such as residence, socio-economic status, some
aspects of medical history (e.g. family history of breast cancer), and limited information on
eating habits and alcohol consumption. This information will also be collected by interview,
with the person themselves where possible, or else with a close relative or, possibly, work
colleague. Interviews will be carried out by an experienced research nurse or other
occupational health professional. We will as far as practicable ensure that the interviewer is
not aware of who developed cancer and who did not.
All information collected about individuals will be confidential to the research team and none
of it will be published in such a way that individuals can be identified. This may well
influence the extent to which detailed information about the findings of the case-only studies
of stomach and brain cancer can be reported.
We will find a suitable location in Greenock or nearby for carrying out the interviews with
study members, relatives and friends and travel will be arranged where necessary. There will
be a small attendance payment. Under some circumstances we may be able to visit and
interview people at home. Telephone interviews will be necessary for those that live far away
or who cannot travel to be interviewed. Information will be recorded to paper forms, but will
be kept in a secure way, for later checking, computerisation and analysis.
1.9 APPROVALS, CONSULTATIONS AND THE NEED FOR ACTIVE SUPPORT
FOR THE STUDY
We will seek ethical approval for the study from South Glasgow and Clyde Research Ethics
Committee and the Scottish Privacy Advisory Committee, and the study will not proceed
without their agreement. In addition, as for the previous study, an independent Study Steering
Committee (SSC) of respected scientists and professionals will monitor the work done, from
the viewpoints of scientific reliability and good professional practice on ethical and other
issues.
Also, the investigation will proceed only if we obtain the consent of the NSUK management.
Their consent and active co-operation is essential so that the research team can visit the
factory to examine records of workplace exposures and assess the exposures that might have
occurred during various processes. As noted earlier, that assessment will then enable us to
develop, in much more detail than we can at present, the study procedures for recording work
histories and estimating exposures.
4
(Version 2.0.4a for HSE website – October 2007)
5
We will also discuss our plans with the workforce, and thereafter with various stakeholders in
the community, including the Phase Two group and its advisors. Ex-employees who are not
directly involved will be informed as best we can by general publicity about the study. The
consent and active co-operation of the present and past workforce, and of the local
community as a whole, is essential if we are to get the participation we need in collecting the
new and detailed information that is necessary for the studies to be successful.
In addition to general community support for the study, all participating cases and controls, or
those providing information about them and all interviewees for the HHA, will individually
be asked for their informed consent to being involved in the investigation. We hope to
complete the study within 12-15 months of having all the agreements we need in place.
1.10 COMMUNICATING THE FINDINGS
When the study is completed, the work and the results will be presented to NSUK
management, to meetings of the workforce, and to other stakeholders (e.g. Phase Two,
relevant Government Departments and Parliamentary representatives). A short leaflet
summarising the study and its findings will also be prepared and made available widely. We
will make this as non-technical as we can, while ensuring that it is fair and accurate. A full
research report will be prepared and published on the HSE website – paper copies will be
available to main stakeholders. One or more scientific papers will be prepared and submitted
for publication to appropriate peer-reviewed scientific journals.