Recruiting patients to medical research: double blind randomised trial of “opt-in” versus “opt-out” strategies
BMJ 2005; 331 doi: https://doi.org/10.1136/bmj.38583.625613.AE (Published 20 October 2005) Cite this as: BMJ 2005;331:940
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
We recently published a method to determine if consenters to
population surveys are representative of the target study population [1].
Recently introduced data protection legislation has changed research
practice in the UK, although uncertainties about the interpretation of the
legislation remain [2]. Methods such as searching general practice records
to investigate whether non-responders are different to responders, are no
longer acceptable as they use information about identifiable individuals
without their consent.An alternative approach to estimating the effect of
non-response is to reformulate the problem as the extent to which
responders are representative of the total target population. Since
anonymised data may be obtained and used for the target population as a
whole, this represents an indirect way of assessing the impact of non-
response on representativeness. We compared consultation rates for some
common conditions among responders to a headache survey [3] who had
consented to the use of their practice records for research with
anonymised rates derived from the practice populations from which they
were sampled.
We found that the adjusted consultation rates were similar but
generally higher in the consenters group compared with the population
group. This suggests that consenters have more morbidity than those who do
not participate, not only with respect to the topic of the survey
(headache) but for other conditions also. However, the differences were
not large and suggest that non-response bias is unlikely to be substantial
with respect to the extent and pattern of ill health.
The alternative method which we described offers one potential
approach to determine whether respondents to a study are representative of
the population from which they were sampled with respect to a particular
set of characteristics, namely their morbidity as measured by consultation
with a general practitioner, and as such can be used as one measure of the
representativeness of the responding sample.
1.Strobl J, Cave E, Walley T. Data protection legislation:
interpretation and barriers to research.
BMJ 2000;321:890-892.
2.Helen F. Boardman, Elaine Thomas, Helen Ogden, Peter R. Croft, and
David S. Millson
A method to determine if consenters to population surveys are
representative of the target study population
J. Public Health Med., Jun 2005; 27: 212 - 214.
3.Boardman HF, Thomas E, Croft PR, Millson DS. Epidemiology of
headache in an English district.
Cephalalgia 2003;23:129-137.
Competing interests:
None declared
Competing interests: No competing interests
Dear editor,
Patients, who are with cardiovascular functional impairment, may have
more difficulties to deal with their daily life. They may consider
writing a reply letter or ring up someone who is not familiar as a
burden/great ”personal sacrifices”, and they would be less willing to make
such an effort to become participants. However to answer a phone and “say
yes” would be much more easier for these people, and therefore more
people, who were in poorer health, were recruited by the “opt-out” method
in this study. [1]
Given this explanation is correct, the “opt-out” strategy may not be
better than the “opt-in” strategy to recruit patients with higher risk, if
the risk does not associate with functional impairments which are”
barriers” for active response. For example: to recruit women for breast
cancer prevention research, “opt-out” method may not be able to obtain a
sample with a bigger proportion of women who are at higher risk, than the
“opt-in” method.
Reference
1. Junghans Cornelia, Feder Gene, Hemingway Harry, Timmis Adam ,
Jones Melvyn, Recruiting patients to medical research: double blind
randomised trial of “opt-in” versus “opt-out” strategies. BMJ,
doi:10.1136/bmj.38583.625613.AE (published 12 September 2005)
Competing interests:
None declared
Competing interests: No competing interests
The kind we look up Vs The kind we make up.
Dear Editor,
I am talking of statistics. What are we comparing in a controlled
study? It is only a few phenotypic features. Those phentotypic features,
by themselves, can not predict the future time evolution. The latter
depends on the knowledge of the total initial state of the organism
(patient or control)for any controlled study. Randomisation does not help
either. That being the case what good are the controlled studies, in the
first place?
Be that as it may, the new "avatar" of clinical drug trial research
in the third world countries involves a new breed of CROs (clinical
research organisations) that work like the middlemen in any business! Less
said about their authenticity the better. Where are we headed in medical
research?
Yours ever,
bmhegde
Competing interests:
None declared
Competing interests: No competing interests