Supplementary letter from Professor Peter
Burney, Professor of Respiratory Epidemiology and Public Health,
Imperial College, London
I promised a further brief note on the current
problems of undertaking epidemiological research.
The principal difficulty is in accessing good
representative samples of the population at reasonable cost.
Previously this was possible by accessing the
names and addresses of people registered with general practices.
A computerised list of these was kept together with sex and date
of birth by the health authorities and age- sex-specific samples
could be drawn by post code. No clinical information was kept
on these files. Other sources such as electoral lists were more
limited and less complete and are even more so now.
Although use of these data was acceptable to
the public, the data are now regarded as not only confidential
but also "sensitive" and access is denied.
The nearest alternative now is to approach individual
General Practices and invite their collaboration. The first difficulties
with this approach are:
1. General practices that refuse (for many
different reasons) are not the same as those who do not and so
samples are often unrepresentative. Particularly in research into
health care this can be a fatal problem.
2. Even if it were possible, it is far too
expensive in time and resources to approach all GPs and get small
samples of patients from each. Studies therefore need to be of
a "clustered" design and this leads to loss of power,
or the need for larger more costly studies.
However, even if general practices are in principle
willing to help they are very often unable to. Under the current
arrangements they have to write to their patients to ask whether
they are willing to participate and this involves 1) drawing an
appropriate sample (which they are generally not qualified to
do), 2) printing out the letters and posting them (a process that
requires not only time but space), 3) marking off those who have
replied and 4) sending reminders to those who have not.
Under the current rules, because we cannot have
access to the names and addresses until the patients have replied
to say that they are willing to participate, we are unable to
help with any of this process. For a busy general practice this
is all but an impossible task and it is amazing that we have any
volunteers. In addition we have none of the valuable information
about the age, gender and postcode of the non-respondents.
In all of this I see a large scale repeat of
the legal nonsense that held up anonymous testing for HIV and
any chance of understanding the spread of AIDS in the UK for some
years. What we need is permission to use the names and addresses
of patients registered with GPs together with their dates of birth
and gender providing:
1. The programme of work, including the letter
of invitation to participate and questionnaire, has ethical committee
approval.
2. The staff are adequately trained and have
honorary contracts with a health authority or trust.
PIAG was asked to allow us to take on some of
this work for a particular project either outside the surgeries
or sending our staff into the surgeries to help. This would have
been only a partial solution at best but was turned down on the
grounds that the study was "too small" to warrant an
exception. That is a judgement that in my view PIAG was not qualified
to make. The relevant perception of the size of the problem is
that of the general practitioners and they certainly saw the problem
as one that made their participation impossible.
Further absurdities are also now creeping into
the research governance. Recent advice to us about a questionnaire
survey included an injunction to send a separate consent form
with the questionnaire and a separate letter and "patient
information sheet". These may seem simple requirements, but
there is a world of difference between sending a two page questionnaire
with a letter giving a brief explanation and asking the recipient
to fill it in return it, and sending a legal document with an
extensive "information sheet" with what many of the
public would see as "small print" and asking them to
"sign" that they have agreed to the "small print".
The upshot of all this is that we are organising
a European wide survey of allergic conditions this summer and
may well not be able to comply with the survey design ourselves.
The irony is that a properly designed and well vetted study to
improve our knowledge of public health is forbidden, but any company
can ring me up at home and conduct surveys or try and sell me
whatever they please. If your committee were able to find a solution,
this would be greatly and widely welcomed.
10 January 2007
| 
