BMJ Open
Public health policy, context and ethics: the example of unlinked anonymous testing of blood
Journal:
BMJ Open
rp Fo Manuscript ID: Article Type:
Date Submitted by the Author:
Complete List of Authors:
Kessel, Anthony; London School of Hygiene and Tropical Medicine, Social and environmental health research Datta, Jessica; London School of Hygiene and Tropical Medicine, Wellings, Kaye; London School of Hygiene and Tropical Medicine, Perman, Sarah; Health Protection Agency, Public Health Strategy Public health
Ethics, Infectious diseases
ev
Keywords:
28-May-2012
rr
Secondary Subject Heading:
Research
ee
Primary Subject Heading:
bmjopen-2012-001427
PUBLIC HEALTH, MEDICAL ETHICS, Health policy < HEALTH SERVICES ADMINISTRATION & MANAGEMENT
w
ie ly
on For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 1 of 26
Original paper
Public health policy, context and ethics: a qualitative study of unlinked anonymous testing of blood Unlinked anonymous testing of blood (running title) Anthony S Kessel (corresponding author and guarantor of study), Honorary Professor, London
rp Fo
School of Hygiene & Tropical Medicine, Faculty of Public Health and Policy, 15-17 Tavistock Place, London WC1H 9SH (Tel: 020 7759 2715; email:
[email protected])
Jessica Datta, Research Fellow, London School of Hygiene & Tropical Medicine
ee
Kaye Wellings, Professor, London School of Hygiene & Tropical Medicine
rr
Sarah Perman, Health Protection Agency
Key words
w
ie
ev Unlinked, anonymous, blood, ethics, public health, policy, HIV
Word count
ly
6437 words
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
1
BMJ Open
Abstract In this study we explore the ethical issues around unlinked anonymous testing (UAT) of blood, a method of sero-prevalence surveillance for infectious diseases. Our study focused on UAT for HIV, although UAT can be used for other infectious diseases. Objectives The objectives of the research were to gain a better understanding of the views of key informants in
rp Fo
countries adopting different UAT testing strategies, and to use the findings of the research to inform health policy Design
Qualitative study using in-depth interviews and ethical analysis.
ee
Setting
rr
Four countries using different strategies around unlinked anonymous testing of blood for HIV (United Kingdom, United States, Netherlands and Norway). Participants
Twenty-three key informants in the four countries.
w
Results
ie
ev
Participants from the four countries have different views on UAT of blood, and that the approaches and policies on UAT adopted by different countries have been historically and culturally determined.
on
We use our findings to explore the relationship between public health policy and ethics, framing our discussion in relation to two important contemporary debates: informed consent for participation in
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 26
medical and public health research; and the balance between individual rights and the public good. Conclusions Qualitative research and ethical analysis of unlinked anonymous testing of blood in different countries has yielded important findings for consideration by policy makers. The policy of unlinked anonymous testing of blood for HIV and other diseases in the UK needs re-consideration in light of these findings.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
2
Page 3 of 26
Public health policy, context and ethics: the example of unlinked anonymous testing of blood
In this paper we explore unlinked anonymous testing (UAT) of blood for HIV and other diseases, and examine the ethical issues associated with this method of sero-prevalence surveillance for infectious diseases. We report original results from a study funded by the UK Economic and Social Research Council (ESRC) – qualitative interviews with stakeholders in
rp Fo
four countries adopting different UAT policies – and use the subject matter, and our findings, to explore the relationship between public health policy, context and ethics.
BACKGROUND
ee
Unlike voluntary testing for clinical purposes, unlinked anonymous testing (UAT) of blood is
rr
carried out in order to provide information about the prevalence of particular infectious diseases in a population. The use of UAT began in the late 1980s for HIV, and remains most
ev
commonly associated with this infection, but can be applied to other infectious diseases such as hepatitis C.1 In this paper we use the term UAT synonymously with sero-prevalence surveys, or sero-prevalence surveillance.
w
ie
The principles behind UAT can be explained with reference to the approach used in the UK for HIV. Leftover blood from samples taken for diagnostic or routine screening purposes is
on
made anonymous, irreversibly unlinked from the source, and then tested for HIV.2 3 This means that individuals whose blood tests positive for HIV cannot be contacted or informed.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
The data derived from UAT are then used for monitoring purposes and also to inform the planning and evaluation of health services and health promotion activities.
Historically, UAT has developed in different ways in different settings, yet exploration of the moral dimensions of such developments has been relatively limited.4 5 6 Further, recent attention given to individual rights and autonomy – in the health and healthcare context – make it an appropriate time to reassess moral considerations around UAT. A brief summary
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
3
BMJ Open
of developments in UAT in the four countries used in our study is provided as a necessary introduction to the research project.
History of UAT in different countries A national sero-prevalence surveillance programme using UAT to estimate the prevalence of HIV has been in place in the UK since 1990.1 The programme was first developed in
rp Fo
response to concerns about the spread of HIV during the 1980s7 and, despite early opposition, the methodology eventually gained political support.8 The programme was funded initially as a pilot project by the Medical Research Council, and has since been funded by the Department of Health. [1] [2] The programme currently uses data collected from three ongoing surveys (newborn babies, [maternal antibodies], users of genitourinary medicine
ee
(GUM) clinics, and intravenous drug users [saliva]) to measure the proportion of individuals within the general population, and within subgroups of the population, infected with HIV.
rr
Since the programme’s inception, around ten million blood samples have been tested in this way.9
ev
A system of ‘prior notification’ operates in locations where testing goes on, involving the
ie
display of leaflets or posters explaining that blood may be tested anonymously. Consent to participation is implicit rather than explicit; provision is made to opt out of testing, although this naturally depends on awareness.9 10
w
on
In the USA, UAT was extensively used in the late 1980s and early 1990s using discarded blood samples. The largest survey was national and used newborn heel prick tests. Other
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 26
UAT surveys focused on population groups, for instance a survey of young people volunteering for job corps, although these were regional or local. In contrast to the UK the prior notification system, with the opportunity to opt out, was not used. UAT essentially stopped in the USA in 1996, as evidence emerged in relation to reducing vertical transmission of the HIV virus from a pregnant mother to the foetus through use of anti-viral medicine.11 The interplay at the time of clinical, social and political factors was complex.12 The main method for HIV surveillance in the Netherlands is named voluntary testing rather than UAT (these are ‘opt-out’ programmes, such as the antenatal testing programme and the
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
4
Page 5 of 26
testing of users of STI clinics, as well as the extensive ‘opt-in’ testing carried out by general practitioners). The UAT that does occur in the Netherlands is carried out using blood samples from intravenous drug users (IDUs) and saliva samples from other groups. Unlike the UK or USA, informed consent is required from those surveyed by UAT, either verbal or, more recently, written consent. From 2002 to 2006, other high risk groups haven been included in UAT research surveys in the Netherlands (e.g. migrants from the Cape Verde islands, Surinam and the Netherlands Antilles, as well as commercial sex workers), because of concerns of a higher proportion of undiagnosed HIV in such groups, who were also felt to be
rp Fo
accessing health services less than others. In Norway, public health efforts have focused on diagnosing all those infected, as it was thought (certainly in earlier stages) that those who were diagnosed would be less likely than those undiagnosed to transmit infection to others. Mandatory named AIDS reporting was
ee
introduced in Norway in 1983 and mandatory anonymous HIV case reporting in 1986, both including cases diagnosed before those dates.13 The Norwegian approach has therefore
rr
centred on individual testing and clinician reporting, and UAT has not been deemed epidemiologically necessary and has never been undertaken.
ie
ev AIMS AND OBJECTIVES OF THE RESEARCH STUDY
w
The ESRC-funded project had three components: a questionnaire survey of 471 patients
on
attending two inner-urban genito-urinary medicine (GUM) clinics in two English cities, exploring their views of, and attitudes towards, UAT;14 qualitative in-depth interviews with a sub-sample (20) of these patients; and in-depth interviews with key informants in four
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
countries chosen to represent different approaches towards sero-surveillance for HIV. In this paper, we present findings from this third component, the qualitative interviews with key informants.
The over-arching aim of this component of the research study was to use the subject of UAT, and our findings, to explore the relationship between public health policy, context and ethics. There were two more specific objectives of the research. The first was to gain a better understanding of how UAT has developed in different settings over time, with particular
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
5
BMJ Open
attention to the ethical considerations. The second was to use UAT as an instrument to gain empirical understanding of the ethical tensions between surveillance and patients’ rights, and to explore how these tensions may have developed and been handled in different contexts.
METHODS Four countries were selected purposively to participate in this study component (the key
rp Fo
informant interviews), based on their representing a range of approaches, or positions taken, with regard to UAT: the UK (specifically England), USA, the Netherlands and Norway. Purposive sampling was used to select the key informants from these countries, informed by the following criteria: range of stakeholder perspectives (e.g. epidemiology, clinical, public health policy, charity or pressure group, and academia (public health or medical ethics));
ee
involvement in UAT, historically and/or current; existing links with the research team or advisory group members; and practical issues (e.g. availability and likelihood of participation).
ev
rr
A topic guide was developed that allowed for a semi-structured interview. Elements within the topic guide included: introduction and background to the purpose and nature of the
ie
interview (prior information had been provided); knowledge and views on unlinked anonymous testing of blood (and national programme as appropriate), including advantages
w
and disadvantages; ethical considerations including consent, prior notification, confidentiality, anonymity, individual rights/liberties, social responsibilities, and any changes
on
in ethical position over time; the value to public health of different surveillance methods; blood as a tissue; legal considerations; and any other relevant issues the respondent wished to
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 26
discuss. The topic guide was initially tailored for the different countries based on our own understanding, but then applied flexibly during the interview.
A researcher, with a background and skills in qualitative interviewing, was employed for the duration of the project, and undertook all the interviews. Interviews were carried out either face-to-face or by telephone between January and September 2008. Interviews were recorded (except for two: the first due to unsuitable location and the second to mechanical failure) with the participant’s permission, and transcribed verbatim. Interviews lasted between 20 minutes and one hour. In total 21 interviews were completed with 23 respondents (two interviews
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
6
Page 7 of 26
were with two respondents together). Of these, 12 were with UK respondents (three clinicians in genito-urinary medicine/HIV; one person working in government health policy; two HIV epidemiologists; two ethics academics; two charity representatives [MedFash, Terence Higgins Trust]; one person working at the British Medical Association; one editor of a medical journal), four with respondents in the USA (one health policy academic; one academic in health law; one academic in public health; and one epidemiologist), three with respondents from the Netherlands (two epidemiologists; and one academic in medical ethics), and four with respondents in Norway (one epidemiologist; three academics in medical ethics
rp Fo
– one of whom is also a hospital clinician). There were three interview refusals, all in the UK. [3]
Interview data were coded by the researcher, and emergent themes assessed by the same researcher (in the context of the aims and objectives of the research); a second researcher
ee
assisted with coding and analysis. Inevitably some of the themes inter-link with each other (e.g. ‘value of UAT’ and ‘ethical issues’). Content was analysed for description of UAT
rr
developments in the country concerned, the context in which UAT may have changed, ethical tensions in UAT between surveillance and human rights, and broader connections between UAT and public health policy.
ly
on
UK
w
Value of UAT
ie
RESULTS
ev
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
The programme is largely seen by respondents as indispensable in providing an unbiased measure of HIV prevalence and risk factors in the UK population. The most important outcome from the surveys has been information on undiagnosed HIV infection which, respondents suggested, has led to changes in policy on named testing; for example, HIV testing in pregnant women has changed from opt in to opt out, and sexual health clinics have moved from targeted to universal voluntary testing. The neonatal dried blood spot survey is used as a check on the performance of local antenatal screening. The public health rationale for the programme has not changed since 1990, and respondents cited the continuing high infection rate, particularly among men who have sex with men, as justification:
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
7
BMJ Open
"There was far more feeling of a public health emergency (then) than there is today, although you could argue … that it's actually much worse today because of where we see the disease." (UK4) Several respondents suggested that a review of UAT would be timely, including consideration of extending the surveys to other health care settings and other diseases. UK6 discussed how advances in HIV testing, including self-testing, might alter the questions asked
rp Fo
by the programme. Some suggested that if named testing was normalised, there would be less need for anonymous testing.
A particular theme was whether the British public has a duty to participate in public health monitoring, possibly as part-and-parcel of receiving the benefits of a national health service
ee
(NHS):
rr
"There is a very good case for really exploring with the public, and then pushing very hard to get public agreement, that in general terms that sort of tissue, extra bits of
ev
tissue or blood, should be available for use." (UK4)
ie
USA
According to one respondent, the surveillance programme was valuable in terms of planning,
w
resource allocation, advocacy and interventions to prevent transmission. The heel prick survey provided accurate national data on prevalence in childbearing women and was seen as
on
extremely valuable, with nothing since that has given as accurate a picture of national prevalence (USA 2).
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 26
However, given the questionable ethics of UAT (see later), as well as the resultant lack of good quality prevalence data, USA3 argued that enough became known about prevalence to target the most vulnerable groups for health promotion. “And how much more detail do you want from that, so what would you do differently if you knew that in the next level of detail.” (USA3) Netherlands
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
8
Page 9 of 26
Unlinked anonymous surveys have been used in the Netherlands to measure HIV prevalence among high risk populations, but with informed consent (verbal of written). The populations surveyed by UAT have changed over time. Injecting Drug Users (IDUs) have been surveyed bi-annually since 1994 in different cities in the Netherlands, and are recruited via methadone care; sex workers and migrant groups have also been subject to UAT surveys.
There are two reasons, indicated by respondents, why UAT in Netherlands has been of limited value. First, the uptake of voluntary testing is generally high in the Netherlands,
rp Fo
meaning the quality of that data are generally good. And second, UAT has only been deemed ethically permissible in the Netherlands with informed consent, making large-scale UAT surveys impracticable, so restricting UAT to specific population groups and/or the research context.
ee
Norway
The unknown, undiagnosed segment of the population is known as the ‘dark figure(s)’ in
rr
Norwegian, although neither the ‘dark figure’ nor the number of cases of HIV has been perceived as significant issues in Norway. Uptake of voluntary testing for HIV was described
ev
as very high in intravenous drug users and antenatal clinic users, meaning that UAT has never been perceived as to be of value epidemiologically. The system of surveillance for HIV in
ie
Norway instead covers all diagnosed cases (i.e. not undiagnosed cases as with UAT), and was reported as being simple, inexpensive and capable of providing a representative and timely
w
overview that guides prevention, while respecting confidentiality and adhering to ethical principles.
ly
Quality and nature of data
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
UK The epidemiological data yielded from unlinked anonymous testing are considered by respondents to be of unrivalled quality internationally. It was suggested, however, that the apparent high number of undiagnosed HIV infections from UAT surveys might be inflated by double-counting:
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
9
BMJ Open
"You can quite understand that if you've got gonorrhoea and you've been told by your doctor not to have unsafe sex, you might well decide to go to a different clinic to have your gonorrhoea treated." (UK1) Others defended the statistics and argued that they were explained by a resurgence of high risk behaviour, particularly among gay men. UK1 argued that people who refuse a named test and who are subsequently found to be HIV positive through UAT are likely to be people who know that they are positive.
USA
rp Fo
There was little comment from USA interviewees about the quality of UAT data, since UAT surveys stopped in 1996. However, a theme that emerged in relation to UAT data was around the difference between research and surveillance.
ee
The fact that there is no ‘right’ to health care in the US was felt to have an effect on people’s
rr
views on surveillance – in other words, the ‘disconnect’ between health care provision (not available to all) and surveillance (done to all) (USA 1). The differences between surveillance,
ev
research and practice, with perhaps different ethical standards required for each, emerged as a theme from the interviews (USA 4).
ie
One point of view was that surveillance (because it is named and part of routine care) is part
w
of clinical care rather than purely research (USA 2). However, while some of the surveillance undertakings at state level in the US were considered to be practice, at the level
on
of CDC they were treated as research (USA 1). The same respondent commented that cancer sufferers advocated in favour of surveillance for the purposes of research on the disease,
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 26
since “they were willing to trade privacy and consent for a good which was surveillance that they saw as serving their interests”. Norway and the Netherlands There were no comments from Norway respondents about the quality of UAT data since such surveys are not carried out in Norway, in part because of the good quality of data on HIV and AIDS from other sources.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
10
Page 11 of 26
Two respondents from the Netherlands judged that large-scale UAT programmes were not necessary because named HIV testing is widespread. The quality of information obtained from named testing was felt to be robust and could be used to model estimates of HIV prevalence in the population. Coverage of pregnant women is very high and this group is considered a valid proxy for the heterosexual population. Named HIV testing among users of STI clinics was also presented as being a valid method for estimating prevalence. It was argued that, as long as the uptake of named testing by pregnant women and STI clinic users remains high (above 90%) then a national programme of UAT was not thought to be
rp Fo
necessary. However, notwithstanding the above point it was acknowledged that current methods of surveillance do not allow undiagnosed HIV infection to be measured accurately.
Ethical issues UK
rr
ee
Respondents felt that, on balance, no harm was caused by the programme. A recurrent argument for this was that, in settings where anonymous testing takes place, people are offered a named HIV test:
ie
ev
"I think the ethics are not in fact, are not a problem because everybody is now very strongly advised to have a test and therefore the residual rump who don't have a test,
w
it's important to know what their rate of HIV positivity is." (UK1)
on
Some respondents, however, said there would be ethical concerns if people in these settings were not routinely offered a named test. Several respondents considered the ethics of UAT to
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
be relatively unimportant, for instance in the context of more significant health-related problems in Africa:
"There are these enormous ethical, real ethical issues, and to me, this one … I just can't understand why one should put so much emphasis on it." (UK9) UK11 suggested that the question of ownership of the sample was relevant, but that this could be argued both ways:
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
11
BMJ Open
"Is it yours or is it somebody else's? ... If it's waste that's going to be chucked away then why shouldn't people use it if they find it useful? On the other hand, 'I gave permission for this test for my own benefit and I didn't give you permission to do things to help others.'" (UK11) There was universal agreement that changing to a system of informed consent would damage the programme. There was a fear that introducing information about UAT into a clinical consultation could confuse patients. Patients might think they were getting a named test:
rp Fo
"You would lose someone's confidence and trust because they might not understand the words or the reason behind why we're doing it, and they would think that you could get back to them and you did know." (UK12)
ee
Other respondents argued that it would reduce time for discussing important clinical matters:
rr
"When the midwife turns up, does she pull out a leaflet and say by the way, I'm putting this beside you here? … it would be a perversion of the consultation to divert it into discussing this ..." (UK5)
ie
ev
Further concerns were that a system of informed consent would increase refusals and lead to bias in the surveys, and also the time and cost of gaining consent from patients:
w
"You know we've got ten minute appointments. It could take you five minutes, half of
on
your time just to explain what this is all about. So doing it in a consultation is not realistic." (UK3) The respondent suggested that a system of clinic self-registration,
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 12 of 26
using computers, could be used to explain anonymous testing.
All the respondents thought that prior notification was the right theoretical approach. However, some were concerned that, in practice, it might not be working: "It's a reasonable thing in practical terms provided people are informed. So … if people don't know about their option for opting out and their option for finding out then the system hasn't been working." (UK11)
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
12
Page 13 of 26
Some suggested that the approach needed to be more "active" with information displayed visibly on clinic walls or in leaflets given out to each patient. One clinician, however, expressed his discomfort with his role in the programme. He described filling out the form for UAT blood testing in front of the patient, without explaining what he was doing, and sometimes after the patient had declined a named HIV test:
"You're doing something that sort of seems slightly underhand…you're not explaining it to the patient. Yeah, I'm not comfortable with that, I think one wants to be, to
rp Fo
communicate well and to be as open as possible with patients and that's slightly in conflict with that." (UK3) USA
Consent was not deemed necessary initially in the USA for participation in UAT as patients
ee
had consented to the blood being collected in the first place, and this was seen as adequate. USA1 suggested that surveillance is a public good, meaning there are strong reasons for
rr
making it a universal obligation on citizens – and that notification that surveillance is occurring is different from asking for consent.
ev
The blood spot survey continued until the publication of the first trials showing that vertical
ie
transmission of HIV could be dramatically reduced by treating mothers. When it became clear that infected women could be offered treatment, and therefore should know their HIV
w
status, “it became impossible for the US Centre for Communicable Disease Control and Prevention (CDC) to continue that programme and it was dropped in the early 1990s along
on
with all the other ones” (USA3). It had become difficult to reconcile public health benefits of UAT with individual rights. If CDC had not voluntarily withdrawn funding from UAT, the
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
same respondent indicated that it was likely that Congress would have mandated cessation, as there was a strong coalition in favour of this.
A related legal issue was brought by USA4, in that public health surveillance appears to go against the “fundamental legal principles” in the USA, and constitutional scholars are shocked when they hear what is done.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
13
BMJ Open
‘Once you begin to argue that the government can take your personally identifiable information for any reason it finds useful without your consent you've opened up and blown a hole in much of the fundamental legal principles in the country.’ (USA4) The Tuskegee episode (a study between 1932 and 1972 in Alabama in which black syphilitic Americans were not given treatment although thought they were) was mentioned by USA1 as a reason for US citizens, particularly Afro-Americans, to be suspicious of surveillance practices. In the context of opposition to named case reporting and a consequent lack of trust
rp Fo
in surveillance, USA3 argued that there are probably cultural dimensions to issues such as consent, and the acceptability of particular interventions may vary from country to country. “It [UAT] became more of a civil rights issue than a gay specific one. It was, the government knows something that you don’t and the attitude in the States, at least at
ee
the time, towards the government was much less trusting than in this country [i.e. UK]. People did not assume and still don’t assume that the government necessarily is
rr
looking out for your well being. And I think there’s a much more paranoid attitude towards the government than there is in this country.” (USA3)
ev
Related to these cultural dimensions, a particular ethical concern was that the US (and, in
ie
particular, CDC) supports UAT in developing countries, with the inherent contradiction that it has been rejected in the US.
w
‘It seems to me the CDC can’t at the one time say it’s unethical to do it in the US and
on
say it’s ethical to do it abroad. Different public agencies may come to different decisions, but the same agency it seems to me should be bound by some rule of consistency.’ (USA1)
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 26
USA2 felt very strongly that arrangements for UAT in developing countries are not ethical where pregnant women do not have the opportunity to have a named test, counselling and treatment. Netherlands The respondents felt that more extensive UAT surveys would produce more accurate information for public health surveillance. However, inter-linked practical and ethical doubts
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
14
Page 15 of 26
about UAT were presented. One concern was that UAT might deter high risk groups from seeking testing and treatment. Another concern was that, if UAT focused only on high risk groups, such groups would risk being stigmatised. It was suggested that reporting of findings from UAT surveys of high risk groups would need to be carried out sensitively. A particular issue was consent and maintaining trust with patients. One interviewee expressed that staff in HIV treatment centres and STI clinics demand a more stringent consent process than required by the research ethics committees in the Netherlands in order to promote trust
rp Fo
among patients. Norway
A UAT system was never thought relevant, necessary or appropriate for Norway. In Norway it is not legal to test or analyse blood or biological material for infectious diseases for non-
ee
diagnostic purposes without consent unless the purpose is surveillance of an ‘epidemic’. One respondent felt this displayed a ‘moral grey zones’, as the law hinges on the definition of ‘epidemic’ [Norway 1].
ev
rr
There is very little concern or awareness in Norway about ethical issues around UAT. However, there is a general willingness to provide information for public health purposes in
ie
Norway, and people are used to being surveyed. It was reported that many Norwegians have given blood to the Biobank, and give broad consent for their tissue/blood to be used for a
w
wide range of research. A respondent indicated that many are willing to waive their consent in Norway because ‘there is significant trust’ in the health care system. However, another
on
respondent expressed that there was a possible danger that UAT blood could be used to identify someone via the Home Office DNA database, and that researchers might be forced to
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
share information under a court order (the respondent mentioned that the murderer of the Swedish politician Anna Lindh in 2003 was convicted on the basis of a match with DNA held in the national biobank).
One respondent suggested that informed consent could be different in different contexts. In a country like Norway, with low prevalence and easy access to health care, the current surveillance system (non-UAT) was described as acceptable and functioning well. However, in a country in Africa (where this respondent was carrying out research), with a high rate of
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
15
BMJ Open
HIV positive pregnant women and a low rate of testing, routine testing with opt-out consent procedures would be morally acceptable.
DISCUSSION Since its inception, unlinked anonymous testing (UAT) has been a subject of epidemiological, ethical and policy debate.15 16 17 Different approaches to sero-surveillance
rp Fo
were initially adopted in different countries in response to the developing HIV/AIDS epidemic.4 Over time some of those approaches have changed. The research findings presented in this paper are part of a larger project exploring the reasons behind such differences. The first findings, a quantitative study of the views towards UAT of users of users of two genito-urinary clinics in England, have recently been published.14
ee
In this paper we have presented the results of qualitative interviews with key informants in
rr
four countries pragmatically selected because of differences in approach to testing. It is important to note the limitations of the research.18 Only four settings (countries) were
ev
selected, but these were chosen deliberately as they represent a range of approaches to UAT. Although more data would have been generated by adding further countries, this was not practical for this relatively small study. A total of 23 respondents were interviewed, a
ie
reasonable number for a study of this kind, but they were differentially distributed between
w
countries: 12 in the UK, four in the USA, four in Norway and three in the Netherlands. This distribution (as well as the make-up of respondents) was chosen purposefully to gain useful
on
insights in a qualitative piece of research of this nature, and was not intended to provide a quantitatively representative sample.19 Future study could be undertaken with expanded numbers.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 26
First a summary of key findings will be presented, followed by a discussion of how the findings from this paper fit into broader contemporary debates about public health ethics and policy.
Summary of key findings from interviews
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
16
Page 17 of 26
In the United Kingdom (UK), a balance of epidemiological and ethical considerations in the late 1980s led to the implementation of the ‘prior notification’ approach to UAT; this involves leaflets and posters informing clients that UAT is occurring, allowing opting out but not necessitating the gaining of informed consent.7 10 Respondents stressed that the rationale remains apt today, and that this approach provides data unrivalled in terms of epidemiological quality,9 20 21 and which is of use for health service planning.22 Ethical concerns were generally felt to be minimal today since clients are (also) usually offered voluntary (named) testing in settings where UAT is occurring. However, there was mention of concern about the
rp Fo
lack of full explanation, whether people do actually know UAT is happening, and ‘ownership’ of blood.23 It was indicated that an approach involving informed consent would be cumbersome, time-consuming and epidemiologically damaging.24 In the United States (USA), UAT of blood for HIV was extensively used in the 1980s and
ee
early 1990s,25 but was stopped in the late 1990s as evidence emerged that anti-viral therapy in pregnant women could help reduce the vertical transmission of HIV from pregnant mother to
rr
child – named testing needed to be maximised so that HIV-positive pregnant women could be identified and offered advice and therapy.26 The interviews indicated that data from UAT
ev
were useful epidemiologically but, since UAT has stopped, there has been sufficient other data to plan services accordingly. Other issues raised include a general suspicion around
ie
epidemiological surveillance post-Tuskegee, cultural dimensions of UAT with respect to minority groups, a lack of trust in the Government, and the related disconnect between
w
surveillance (done to all) and health care provision (variable, and not done to all).6 11 There was also comment about the apparent contradiction in the contemporary support of the US
on
Centre for Communicable Disease Control and Prevention (CDC) for UAT in low-income countries.27
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Interviews in the Netherlands showed that named (voluntary) testing, which has generally been widely taken up, has provided the main vehicle for surveillance of HIV.4 There is acknowledgement that large-scale UAT would provide additional epidemiological data, but this has been balanced against concerns of this leading to diminished voluntary testing – especially in at risk groups – through loss of trust in the healthcare system.28 Smaller UAT surveys have been intermittently undertaken in such groups (e.g. intravenous drug users, sex workers, migrants) as research studies, but have required individual consent (initially oral, now written).
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
17
BMJ Open
Interviews in Norway revealed that UAT was never felt to be necessary epidemiologically, as HIV has not been a significant issue and the uptake of voluntary testing has been high.13 Surveillance has thus tended to focus on the monitoring of the diagnosed population (anonymised diagnosed cases), which has been supported by the gay community; the undiagnosed population is interestingly referred to as ‘dark figure(s)’. In Norway it is not legal to test without consent unless there is an epidemic, although respondents reported few ethical concerns since people generally have trust in the health care system, and are willing to
rp Fo
provide information for public health purposes (many give blood to the Biobank for a range of research purposes).29
The findings of this research on the ethics of UAT of blood need consideration in the context of two broader contemporary debates within medical and public health ethics. The first
ee
debate centres on the boundaries of informed consent in health research, particularly in relation to the use of human tissue. The second debate focuses on the balance between
rr
individuals’ rights and the appropriate role of government3 in infringing on those rights for broader public health interests or concerns. These debates are inter-connected, although will be presented in turn.
Informed consent and ethics: a changing debate
w
ie
ev
Respect for autonomy is the bioethical principle that is most likely to be jeopardised by a
on
programme of unlinked anonymous blood testing, since informed consent is not usually deemed necessary for an individual’s blood to be included; this has tended to be the basis for
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 26
opposition to UAT.5 30 The findings from our research show, however, that such positions differ between countries, and have also changed with time.
Debates in the late 1980s and early 1990s demonstrated the opposing views of those who were concerned about patient consent,8 25 and those who felt that the priority was to control what was seen as an epidemic.15 UAT programmes did go ahead in the USA and UK, and other high income countries: there was no requirement for consent in the USA, whereas in the UK an approach of ‘prior notification’ was adopted through leaflets and posters in clinics – consent is presumed but individuals have the opportunity to opt out.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
18
Page 19 of 26
Over the past two decades informed consent has emerged as ‘the ethical touchstone of medical research, and is now enshrined in practice and a range of international guidelines as one of the main preconditions of medical research’.24 31 Some commentators argue that even anonymised tissue should not be used without consent. Trouet, for example, believes that the use of such material is in contravention of Article 22 of the Convention on Human Rights and Biomedicine. This states that ‘when in the course of an intervention any part of a human body is removed, it may be stored and used for a purpose other than that for which it was removed,
rp Fo
only if this is done in conformity with appropriate information and consent procedures’, and in accordance with Article 8 of the European Convention on Human Rights which protects the right to respect for private life.32 Our research indicates that, rather than the universality of human rights discourse, a form of
ee
weak moral relativism operates in policy-making and practice around UAT.33 In the USA, the change in social context and debates about individual rights, along with evidence of reduction
rr
in mother-to-child vertical transmission, resulted in the abandonment of UAT in 1996.12 Interviews in the UK showed that the prior notification approach that remains in use today
ev
still holds strong support – based on perception that there is minimal harm to the individual, and the value of high quality, unbiased epidemiological data on HIV (held to be the most
ie
robust and accurate worldwide) (Buehler et al., 1994).34 Respondents felt that any change to UAT requiring informed consent would be cumbersome, time-consuming, and costly.
w
According to the Nuffield Council on Bioethics it may also be ethically unnecessary: ‘... it is acceptable to collect and use anonymised data for assessing and predicting trends in
on
infectious disease without consent, as long as any invasion of privacy is reduced as far as possible.’.35
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
This position is not, however, cross-culturally held. In the Netherlands, UAT has not been felt to be acceptable without informed consent, initially verbal and more recently written. The societal emphasis on individual autonomy, including in the health research environment, has been the paramount consideration.5 Northern European and Scandinavian countries are often held up as having a strong community ethos, with an expectation on individual involvement in endeavours for the public good.36 This was not reflected in UAT practice in the Netherlands, although was indicated to be important in Norway, albeit in a setting where UAT has never been taken up.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
19
BMJ Open
These findings indicate the importance of social and political context in relation to medicoethical debates, and how the context can change with time and place.37 Differences exist between the moral evaluations and resultant UAT approaches between countries such as England and the Netherlands. Within the USA, in contrast, the practice of UAT has been reevaluated with time. In the late 1980s and early 1990s the feared AIDS epidemic was seen as a public health emergency, there was no treatment available, transmission from mother to unborn child was not preventable, and the stigma and discrimination associated with the
rp Fo
infection were more overt than today. As these factors shifted, so did the policy on UAT in the USA. The situation in low income countries today has created a perception of the acceptability of UAT akin to that in western countries two decades ago.
ee
Individual rights and public good
rr
There has been criticism in recent times of government for its ‘nanny state’ approach to interventions in, for example, the areas of smoking, alcohol and diet. More recently the
ev
notion of ‘stewardship’ in public health rejects these negative associations,38 and advocates that government should have a role to play in protecting the health of its citizens, preventing
ie
them from harming themselves or others, and promoting healthy behaviour.[4] In this view, public health interventions should not be thought of as unwanted restrictions imposed by a
w
‘nanny’ on powerless children, but as actions that favour both individuals and society more widely as prescribed by a ‘steward’ administering and protecting citizens’ interests.[5] The
on
Nuffield Council on Bioethics, for example, argues that the state should act as steward both to individuals and to the population as a whole, but that individuals’ personal choices should also be taken into account.35
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 26
The subject of UAT sits squarely within the debate about public health stewardship for social good. Recently, commentators have argued that patients of a publicly supported health care system like the NHS have a moral duty to participate in scientific research,39 40 although most do not go so far as to articulate that such participation should be mandatory in practice.41 Nevertheless, UAT could be understood as something that individuals should feel they have a moral obligation to participate in, regardless of ethical concerns about individual rights and personally held views. Our own survey of over 400 users of two genitor-urinary medicine
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
20
Page 21 of 26
clinics (a parallel part of this research project) has found that a large majority of respondents (89%) agree that people should have a responsibility to take part in research studies, and the same proportion would agree to the use of their blood in unlinked anonymous seroprevalence testing. However, when asked about the question of consent, 74% indicated that they should be asked to consent before their blood is used for public health surveillance.14 Notwithstanding this important latter point, such support for use of blood samples for prevalence surveys or for future research has not, however, been demonstrated in an earlier study in the UK42 or more recently in Egypt.43
rp Fo
The way forwards perhaps is to ensure that patients – and the public more widely – are informed of the existence, purposes and value of disease surveillance arrangements, as well as the importance of contributing to future health research, while investigating arrangements for gaining consent which are not unnecessarily cumbersome or likely to create bias amongst
ee
those who agree that their blood be used.44
rr
The issue of consent in seroprevalence surveillance, such as UAT, remains an important instrument to gauge the moral direction of public health policies more broadly.45 Debate
ev
continues around how public health programmes can embrace a Kantian stewardship relationship, rather than a nanny paternalistic one, while respecting autonomy and advocating
ie
social responsibilities. Further examination of why different countries have adopted different positions on seroprevalence surveillance will help illuminate our understanding of how social
w
morality is embedded in public health policies and practice.
ly
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
21
BMJ Open
Acknowledgments The authors would like to thank the participants in the study. We also acknowledge the valuable advice given by Vivian Hope, Will Nutland, Cicely Marston, and Graham Hart.
Conflicts of interest AK is also Director of Public Health Strategy and Medical Director of the Health Protection Agency. The views expressed in this paper are those of the authors, and are not intended to represent the views of the Health Protection Agency.
Funding
rp Fo
This research was supported by the Economic and Social Research Council (ESRC Grant Reference: RES-000-22-2096).
Ethical approval
ee
The study was approved by the South West Research Ethics Committee on August 16th 2007
rr
(reference 07/H0206/46), and by the London School of Hygiene & Tropical Medicine ethics committee on 3rd September 2007 (reference 5174).
ev
Notes
The England and Wales programme is currently managed by the Health Protection
w
[1]
ie
Agency (HPA) and the Institute of Child Health, University of London. The programme was previously managed by the Communicable Disease Surveillance Centre, part of the Public
on
Health Laboratory Service, which was subsumed into the Health Protection Agency in 2005. [2]
Health Protection Scotland (HPS) co-ordinates data collection in Scotland.
[3]
Those chosen had particular involvement or interest in UAT and ethics, but further
details cannot be provided to maintain their confidentiality. [4]
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 26
By government, the meaning here is health and public health policy makers within
government. [5]
It is interesting that in this discourse ‘nanny’ (female) is seen as negative and
‘steward’ (male) as positive, when both are servants employed to manage and protect.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
22
Page 23 of 26
References
1
Nicoll A, Gill ON, Peckham C et al. The public health applications of unlinked
anonymous seroprevalence monitoring of HIV in the United Kingdom. International Journal of Epidemiology 2000;29:1-10. 2
Department of Health. Prevalence of HIV in England and Wales in 1997: Annual
Report of the Unlinked Anonymous Prevalence Monitoring Programme. London: Department
rp Fo
of Health, 1998. 3
UK Collaborative Group for HIV and STI Surveillance. Testing Times: HIV and
Other Sexually Transmitted Infections in the United Kingdom. London: Health Protection Agency, 2007. 4
ee
Bayer R, Lumey LH, Wan L. The American, British and Dutch responses to unlinked
anonymous HIV seroprevalence studies: an international comparison. AIDS 1990;4(4):283-
rr
290. 5
ev
Kopelman LM. Informed consent and anonymous tissue samples: the case of HIV
seroprevalence studies. Journal of Medicine and Philosophy 1994;19:525-52. 6
ie
Fairchild AL, Bayer R. Ethics and the conduct of public health surveillance. Science
2004;303:631-632. 7
w
Gill ON., Adler MW, Day NE. Monitoring the prevalence of HIV: foundations for a
on
Programme of unlinked anonymous testing in England and Wales. British Medical Journal 1989:299:1295-7. 8
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Berridge V. AIDS in the UK: the Making of Policy 1981-1994. Oxford: Oxford
University Press, 1996. 9
Health Protection Agency. Overall HIV prevalence. Health Protection Agency 2008.
http://www.hpa.org.uk/web/HPAweb&Page&HPAwebAutoListName/Page/1201094588821. 10
Gillon R. Testing for HIV without permission. British Medical Journal 1987;294:821-
823.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
23
BMJ Open
11
Fairchild AL, Bayer R, Colgrove J. Searching Eyes: Privacy, the State, and Disease
Surveillance in America. Berkeley: University of California Press, 2007. 12
Stoto MA. Almario DA, McCormick MC. Reducing the Odds: Preventing Perinatal Transmission of HIV in the United States. Washington: National Academy Press, 1999.
13
Aavitsland P, Nilsen Ø, Lystad A. Anonymous reporting of HIV infection: an
evaluation of the HIV/AIDS surveillance system in Norway 1983-2000. European Journal of Epidemiology 2001;17:479-89. 14
rp Fo
Datta J, Kessel AS, Wellings K, Nanchahal K, Marks D, Kinghorn G. The views of
genitourinary medicine (GUM) clinic users on unlinked anonymous testing of blood for HIV: evidence from a pilot study of clinics in two English cities. Journal of Medical Ethics 2011 doi:10.1136/jme.2011.042705. 15
Doll R. A proposal for doing prevalence studies of AIDS. British Medical Journal
rr
1987;294:443-444. 16
ee
Avins AL, Lo B. To tell or not to tell: the ethical dilemmas of HIV test notification in
ev
epidemiologic research. American Journal of Public Health 1989;79/11:1544-7. 17
Datta J, Kessel AS. Unlinked anonymous testing for blood purposes: an ethical
ie
dilemma? In: Peckham S, Hann A, eds. Public Health Ethics and Practice. Bristol: Policy Press, 2010:101-116.
w
18
Bowling A. Research Methods in Health. Buckingham: Open University Press, 2000.
19
Green J, Thorogood N. Qualitative Methods for Health Research. London: Sage,
2005.
ly
20
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 24 of 26
Duong T, Ades AE, Rogers P, et al. Non-participation bias in unlinked anonymous
HIV-prevalence surveys in England and Wales. Epidemiology and Infection 1999;122:26772. 21
Pulver WP, Glebatis D, Wade N, et al. Trends from an HIV seroprevalence study
among childbearing women in New York State from 1988 through 2000: a valuable epidemiologic tool. Archives of Pediatric and Adolescent Medicine 2004;158(5):443-8.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
24
Page 25 of 26
22
Kessel AS, Watts CJ. Usefulness of information from the Unlinked Anonymous
Prevalence Monitoring Programme for HIV in England and Wales: survey of planners of HIV/AIDS services. International Journal of STD and AIDS 1999;10:808-11. 23
Pfeffer N, Laws S. 'It's only a blood test': what people know and think about
venepuncture and blood. Social Science & Medicine 2006;62:3011-3023. 24
Cassell J, Young A. Why we should not seek individual informed consent for
rp Fo
participation in health services research. Journal of Medical Ethics 2002;23:313-317. 25
Bayer R. The ethics of blinded HIV surveillance testing. American Journal of Public
Health 1993;83/4:496-7. 26
de Zulueta P. The ethics of anonymised HIV testing of pregnant women: a
ee
reappraisal. Journal of Medical Ethics 2000;26:16-21. 27
Rennie S, Norris Turner A, Mupenda B, et al. Conducting unlinked anonymous HIV
rr
surveillance in developing countries: ethical, epidemiological, and public health concerns. PLoS Medicine 2009;6(1):e1000004. 28
ev
Postema EJ, Willems PWJM, de Ridder MAJ, et al. Comparison of patients refusing
ie
with patients accepting unlinked anonymous HIV testing in an outpatient STD department in the Netherlands. International Journal of STD and AIDS 1997;8:368-72. 29
w
Skolbekken J-A, Ursin LO, Solberg B, et al. Not worth the paper it’s written on?
on
Informed consent and Biobank research in a Norwegian context. Critical Public Health 2005;15(4):335-347. 30
Beauchamp TL and Childress JF. Principles of Biomedical Ethics. New York: Oxford
University Press, 1994. 31
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Singleton P, Wadsworth M. Consent for the use of personal medical data in research.
British Medical Journal 2006;333:255-8. 32
Trouet C. New European guidelines for the use of stored human biological materials
in biomedical research. Journal of Medical Ethics 2004;30:99-103. 33
Rachels J. The Elements of Moral Philosophy. New York: McGraw-Hill, 1993.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
25
BMJ Open
34
Beuhler JW, Petersen LR, Ward JW, et al. Defending HIV seroprevalence surveys.
American Journal of Public Health 1994;84:319-320. 35
Nuffield Council on Bioethics. Public Health: Ethical Issues. London: Nuffield
Council on Bioethics, 2007 36
Cappelen AW, Norheim OF. Responsibility in health care: a liberal egalitarian
approach. Journal of Medical Ethics 2005;31:476-80. 37
rp Fo
Mariner WK. Mission creep: public health surveillance and medical privacy. Boston
University Law Review 2007;87/2:347-95. 38
Jochelson K. Nanny or Steward? The Role of Government in Public Health. King's
ee
Fund Working Paper. London: King's Fund 2005. 39
Harris J. Scientific research is a moral duty. Journal of Medical Ethics
2005;31(4):242-248.
rr
40
Evans HM. Do patients have duties? Journal of Medical Ethics 2007;33(12):689-694.
41
Shapshay S, Pimple KD. Participation in biomedical research is an imperfect moral
ie
ev
duty: a response to John Harris. Journal of Medical Ethics 2007;33:414-417. 42
w
Kessel AS, Watts C, Weiss HA. Bad blood? Survey of public's views on unlinked
anonymous testing of blood for HIV and other diseases. British Medical Journal 2000;320:
on
90-91. 43
Abou-Zeid A, Silverman H, Shehata M, et al. Collection, storage and use of blood
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 26 of 26
samples for future research: views of Egyptian patients expressed in a cross-sectional survey. Journal of Medical Ethics 2010;36:539-547. 44
Nicholls SG. Knowledge or understanding? Informed choice in the context of
newborn bloodspot screening. Public Health Ethics 2010;3:128-136. 45
Haimes E. What can the social sciences contribute to the study of ethics? Theoretical,
empirical and substantive considerations. Bioethics 2002;16/2:89-113.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
26
BMJ Open
The ethics of unlinked anonymous testing of blood: views from in-depth interviews with key informants in four countries
rp Fo Journal:
Manuscript ID: Article Type:
Date Submitted by the Author:
Complete List of Authors:
BMJ Open bmjopen-2012-001427.R1 Research 15-Aug-2012
Primary Subject Heading:
Keywords:
Public health
Ethics, Infectious diseases
ev
Secondary Subject Heading:
rr
ee
Kessel, Anthony; London School of Hygiene and Tropical Medicine, Social and environmental health research Datta, Jessica; London School of Hygiene and Tropical Medicine, Wellings, Kaye; London School of Hygiene and Tropical Medicine, Perman, Sarah; Health Protection Agency, Public Health Strategy
PUBLIC HEALTH, MEDICAL ETHICS, Health policy < HEALTH SERVICES ADMINISTRATION & MANAGEMENT
w
ie ly
on For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 1 of 53
Original paper
The ethics of unlinked anonymous testing of blood: views from in-depth interviews with key informants in four countries Unlinked anonymous testing of blood (running title)
Anthony S Kessel (corresponding author and guarantor of study), Honorary Professor, London
rp Fo
School of Hygiene & Tropical Medicine, Faculty of Public Health and Policy, 15-17 Tavistock Place, London WC1H 9SH (Tel: 020 7759 2715; email:
[email protected])
Jessica Datta, Research Fellow, London School of Hygiene & Tropical Medicine
ee
Kaye Wellings, Professor, London School of Hygiene & Tropical Medicine
rr
Sarah Perman, Health Protection Agency
Key words
w
ie
ev Unlinked, anonymous, blood, ethics, public health, policy, HIV
Word count
ly
6437 words
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
1
BMJ Open
Abstract In this study we explore the ethical issues around unlinked anonymous testing (UAT) of blood, a method of sero-prevalence surveillance for infectious diseases. Our study focused on UAT for HIV, although UAT can be used for other infectious diseases.
Objectives The objectives of the research were to gain a better understanding of the views of key informants in countries adopting different UAT testing strategies, and to use the findings of the research to inform
rp Fo
health policy
Design
Qualitative study using in-depth interviews and ethical analysis.
ee
Setting
Four countries using different strategies around unlinked anonymous testing of blood for HIV (United
rr
Kingdom, United States, Netherlands and Norway).
Participants
ev
Twenty-three key informants in the four countries.
ie
Results
Participants from the four countries have different views on UAT of blood, and the approaches and
w
policies on UAT adopted by different countries have been historically and culturally determined. We use our findings to explore the relationship between public health policy and ethics, framing our
on
discussion in relation to two important contemporary debates: informed consent for participation in medical and public health research; and the balance between individual good and the public good.
Conclusions
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 53
Qualitative research and ethical analysis of unlinked anonymous testing of blood in different countries has yielded important findings for consideration by policy makers. The policy of unlinked anonymous testing of blood for HIV and other diseases in the UK needs re-consideration in light of these findings.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
2
Page 3 of 53
The ethics of unlinked anonymous testing of blood: views from in-depth interviews with key informants in four countries
In this paper we explore unlinked anonymous testing (UAT) of blood for HIV and other diseases, and examine the ethical issues associated with this method of sero-prevalence surveillance for infectious diseases. We report original results from a study funded by the UK Economic and Social Research Council (ESRC) – qualitative interviews with stakeholders in
rp Fo
four countries adopting different UAT policies – and use the subject matter, and our findings, to explore the relationship between public health policy, context and ethics.
BACKGROUND
ee
Unlike voluntary testing for clinical purposes, unlinked anonymous testing (UAT) of blood is
rr
carried out in order to provide information about the prevalence of particular infectious diseases in a population. The use of UAT began in the late 1980s for HIV, and remains most
ev
commonly associated with this infection, but can be applied to other infectious diseases such as hepatitis C.1 In this paper we use the term UAT synonymously with sero-prevalence surveys, or sero-prevalence surveillance.
w
ie
The principles behind UAT can be explained with reference to the approach used in the UK for HIV. Leftover blood from samples taken for diagnostic or routine screening purposes is
on
made anonymous, irreversibly unlinked from the source, and then tested for HIV.2 3 This means that individuals whose blood tests positive for HIV cannot be contacted or informed.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
The data derived from UAT are then used for monitoring purposes and also to inform the planning and evaluation of health services and health promotion activities.
Historically, UAT has developed in different ways in different settings, yet exploration of the moral dimensions of such developments has been relatively limited.4 5 6 Further, dominant attention given to individual rights and autonomy – in the health and healthcare context – make it an appropriate time to reassess moral considerations around UAT. A brief summary
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
3
BMJ Open
of developments in UAT in the four countries used in our study is provided as a necessary introduction to the research project.
History of UAT in different countries
A national sero-prevalence surveillance programme using UAT to estimate the prevalence of HIV has been in place in the UK since 1990.11 The programme was first developed in
Formatted: Endn
rp Fo
response to concerns about the spread of HIV during the 1980s7 and, despite early opposition, the methodology eventually gained political support.8 The programme was funded initially as a pilot project by the Medical Research Council, and has since been funded by the Department of Health. [1] [2] The programme currently uses data collected from three ongoing surveys (newborn babies, [maternal antibodies], users of genitourinary medicine
ee
(GUM) clinics, and intravenous drug users [saliva]) to measure the proportion of individuals within the general population, and within subgroups of the population, infected with HIV.
rr
Since the programme’s inception, around ten million blood samples have been tested in this way.9
ev
A system of ‘prior notification’ operates in locations where testing goes on, involving the
ie
display of leaflets or posters explaining that blood may be tested anonymously. Consent to participation is implicit rather than explicit; provision is made to opt out of testing, although this naturally depends on awareness.99 10
w
Formatted: Endn
on
In the USA, UAT was extensively used in the late 1980s and early 1990s using discarded blood samples. The largest survey was national and used newborn heel prick tests. Other
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 53
UAT surveys focused on population groups, for instance a survey of young people volunteering for job corps, although these were regional or local. In contrast to the UK the prior notification system, with the opportunity to opt out, was not used. UAT essentially stopped in the USA in 1996, as evidence emerged in relation to reducing vertical transmission of the HIV virus from a pregnant mother to the foetus through use of anti-viral medicine.11 The interplay at the time of clinical, social and political factors was complex.12
The main method for HIV surveillance in the Netherlands is named voluntary testing rather than UAT (these are ‘opt-out’ programmes, such as the antenatal testing programme and the
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
4
Page 5 of 53
testing of users of STI clinics, as well as the extensive ‘opt-in’ testing carried out by general practitioners). The UAT that does occur in the Netherlands is carried out using blood samples from intravenous drug users (IDUs) and saliva samples from other groups. Unlike the UK or USA, informed consent is required from those surveyed by UAT, either verbal or, more recently, written consent. From 2002 to 2006, other high risk groups haven been included in UAT research surveys in the Netherlands (e.g. migrants from the Cape Verde islands, Surinam and the Netherlands Antilles, as well as commercial sex workers), because of concerns of a higher proportion of undiagnosed HIV in such groups, who were also felt to be
rp Fo
accessing health services less than others.
In Norway, public health efforts have focused on diagnosing all those infected, as it was thought (certainly in earlier stages) that those who were diagnosed would be less likely than those undiagnosed to transmit infection to others. Mandatory named AIDS reporting was
ee
introduced in Norway in 1983 and mandatory anonymous HIV case reporting in 1986, both including cases diagnosed before those dates.13 The Norwegian approach has therefore
rr
centred on individual testing and clinician reporting, and UAT has not been deemed epidemiologically necessary and has never been undertaken.
ie
ev AIMS AND OBJECTIVES OF THE RESEARCH STUDY
w
The ESRC-funded project had three components: a questionnaire survey of 471 patients
on
attending two inner-urban genito-urinary medicine (GUM) clinics in two English cities, exploring their views of, and attitudes towards, UAT;14 qualitative in-depth interviews with a sub-sample (20) of these patients; and in-depth interviews with key informants in four
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
countries chosen to represent different approaches towards sero-surveillance for HIV. In this paper, we present findings from this third component, the qualitative interviews with key informants.
The over-arching aim of this component of the research study was to use the subject of UAT, and our findings, to explore the relationship between public health policy, context and ethics. There were two more specific objectives of the research. The first was to gain a better understanding of how UAT has developed in different settings over time, with particular
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
5
BMJ Open
attention to the ethical considerations. The second was to use UAT as an instrument to gain empirical understanding of the ethical tensions between surveillance and patients’ rights, and to explore how these tensions may have developed and been handled in different contexts.
METHODS
Four countries were selected purposively to participate in this study component (the key
rp Fo
informant interviews), based on their representing a range of approaches, or positions taken, with regard to UAT: the UK (specifically England), USA, the Netherlands and Norway. Purposive sampling was used to select the key informants from these countries, informed by the following criteria: range of stakeholder perspectives (e.g. epidemiology, clinical, public health policy, charity or pressure group, and academia (public health or medical ethics));
ee
involvement in UAT and with appropriate knowledge or expertise, historically and/or current; existing links with the research team or advisory group members; and practical issues (e.g.
rr
availability and likelihood of participation). A pragmatic approach meant that more informants were interviewed in the UK than the other countries.
ev
A topic guide was developed that allowed for a semi-structured interview. Elements within
ie
the topic guide included: introduction and background to the purpose and nature of the interview (prior information had been provided); knowledge and views on unlinked
w
anonymous testing of blood (and national programme as appropriate), including advantages and disadvantages; ethical considerations including consent, prior notification,
on
confidentiality, anonymity, individual rights/liberties, social responsibilities, and any changes in ethical position over time; the value to public health of different surveillance methods;
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 53
blood as a tissue; legal considerations; and any other relevant issues the respondent wished to discuss. The topic guide was initially tailored for the different countries based on our own understanding, but then applied flexibly during the interview.
A researcher, with a background and skills in qualitative interviewing, was employed for the duration of the project, and undertook all the interviews. Interviews were carried out, in English, either face-to-face or by telephone between January and September 2008. All nonUK informants spoke good English, without evidence to the research team of any obvious language barrier issues. Interviews were recorded (except for two, when notes were taken: the
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
6
Page 7 of 53
first due to unsuitable location and the second to mechanical failure) with the participant’s permission, and transcribed verbatim. Interviews lasted between 20 minutes and one hour, the length directed by the contributions of the informants. In total 21 interviews were completed with 23 respondents (two interviews were with two respondents together). Of these, 12 were with UK respondents (three clinicians in genito-urinary medicine/HIV; one person working in government health policy; two HIV epidemiologists; two ethics academics; two charity representatives [MedFash, Terence Higgins Trust]; one person working at the British Medical Association; one editor of a medical journal), four with respondents in the USA (one health
rp Fo
policy academic; one academic in health law; one academic in public health; and one epidemiologist), three with respondents from the Netherlands (two epidemiologists; and one academic in medical ethics), and four with respondents in Norway (one epidemiologist; three academics in medical ethics – one of whom is also a hospital clinician). There were three interview refusals, all in the UK. [3]
ee
Two researchers (primary researcher, JD, and a second researcher, SP) coded the interview
rr
data using a thematic content analysis approach, looking for emergent themes (open coding in the context of the aims and objectives of the research). There were no discrepancies between
ev
the two researchers that required further discussion and consensus. Content was analysed for description of UAT developments in the country concerned, the context in which UAT may
ie
have changed, ethical tensions in UAT between surveillance and human rights, and broader connections between UAT and public health policy. Inevitably some of the themes inter-link
w
with each other (e.g. ‘value of UAT’ and ‘ethical issues’). Coding frames and analyses were shared with the wider research team.
on
Interview data pertaining to the history and state of sero-surveillance in different countries
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
were triangulated with literature sources. It was not appropriate to triangulate informants’ views on sero-surveillance.
RESULTS
The results, which represent the views and perceptions of informants, are presented by themes and by country. The labels describe the themes that emerged, in the context of informants’ awareness of the aims and objectives of the study.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
7
BMJ Open
Value of UAT
UK The programme is largely seen by respondents as indispensable in providing an unbiased measure of HIV prevalence and risk factors in the UK population. The most important outcome from the surveys has been information on undiagnosed HIV infection which, respondents suggested, has led to changes in policy on named testing; for example, HIV
rp Fo
testing in pregnant women has changed from opt in to opt out, and sexual health clinics have moved from targeted to universal voluntary testing. The neonatal dried blood spot survey is used as a check on the performance of local antenatal screening. The public health rationale for the programme has not changed since 1990, and respondents cited the continuing high infection rate, particularly among men who have sex with men, as justification:
ee
"There was far more feeling of a public health emergency (then) than there is today,
rr
although you could argue … that it's actually much worse today because of where we see the disease." (UK4)
ev
Several respondents suggested that a review of UAT would be timely, including
ie
consideration of extending the surveys to other health care settings and other diseases. UK6 discussed how advances in HIV testing, including self-testing, might alter the questions asked
w
by the programme. Some suggested that if named testing was normalised, there would be less need for anonymous testing.
on
A particular theme was whether the British public has a duty to participate in public health
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 53
monitoring, possibly as part-and-parcel of receiving the benefits of a national health service (NHS):
"There is a very good case for really exploring with the public, and then pushing very hard to get public agreement, that in general terms that sort of tissue, extra bits of tissue or blood, should be available for use." (UK4)
USA
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
8
Page 9 of 53
According to one respondent, the surveillance programme was valuable in terms of planning, resource allocation, advocacy and interventions to prevent transmission. The heel prick survey provided accurate national data on prevalence in childbearing women and was seen as extremely valuable, with nothing since that has given as accurate a picture of national prevalence (USA 2).
However, given the questionable ethics of UAT (see later), as well as the resultant lack of good quality prevalence data, USA3 argued that enough became known about prevalence to
rp Fo
target the most vulnerable groups for health promotion.
“And how much more detail do you want from that, so what would you do differently if you knew that in the next level of detail.” (USA3)
Netherlands
ee
Unlinked anonymous surveys have been used in the Netherlands to measure HIV prevalence
rr
among high risk populations, but with informed consent (verbal or written). The populations surveyed by UAT have changed over time. Injecting Drug Users (IDUs) have been surveyed
ev
bi-annually since 1994 in different cities in the Netherlands, and are recruited via methadone care; sex workers and migrant groups have also been subject to UAT surveys.
ie
There are two reasons, indicated by respondents, why UAT in Netherlands has been of
w
limited value. First, the uptake of voluntary testing is generally high in the Netherlands, meaning the quality of that data are generally good. And second, UAT has only been deemed
on
ethically permissible in the Netherlands with informed consent, making large-scale UAT surveys impracticable, so restricting UAT to specific population groups and/or the research context.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Norway The unknown, undiagnosed segment of the population is known as the ‘dark figure(s)’ in Norwegian, although neither the ‘dark figure’ nor the number of cases of HIV has been perceived as significant issues in Norway. Uptake of voluntary testing for HIV was described as very high in intravenous drug users and antenatal clinic users, meaning that UAT has never been perceived as to be of value epidemiologically. The system of surveillance for HIV in Norway instead covers all diagnosed cases (i.e. not undiagnosed cases as with UAT), and was
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
9
BMJ Open
reported as being simple, inexpensive and capable of providing a representative and timely overview that guides prevention, while respecting confidentiality and adhering to ethical principles.
Quality and nature of data
UK
rp Fo
The epidemiological data yielded from unlinked anonymous testing are considered by respondents to be of unrivalled quality internationally. It was suggested, however, that the apparent high number of undiagnosed HIV infections from UAT surveys might be inflated by double-counting:
ee
"You can quite understand that if you've got gonorrhoea and you've been told by your doctor not to have unsafe sex, you might well decide to go to a different clinic to have
rr
your gonorrhoea treated." (UK1)
ev
Others defended the statistics and argued that they were explained by a resurgence of high risk behaviour, particularly among gay men. UK1 argued that people who refuse a named test
ie
and who are subsequently found to be HIV positive through UAT are likely to be people who know that they are positive.
w
USA
on
There was little comment from USA interviewees about the quality of UAT data, since UAT surveys stopped in 1996. However, a theme that emerged in relation to UAT data was around the difference between research and surveillance.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 53
The fact that there is no ‘right’ to health care in the USA was felt to have an effect on people’s views on surveillance – in other words, the ‘disconnect’ between health care provision (not available to all) and surveillance (done to all) (USA 1). The differences between surveillance, research and practice, with perhaps different ethical standards required for each, emerged as a theme from the interviews (USA 4).
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
10
Page 11 of 53
One point of view was that surveillance (because it is named and part of routine care) is part of clinical care rather than purely research (USA 2). However, while some of the surveillance undertakings at state level in the USA were considered to be practice, at the level of CDC they were treated as research (USA 1). The same respondent commented that cancer sufferers advocated in favour of surveillance for the purposes of research on the disease, since “they were willing to trade privacy and consent for a good which was surveillance that they saw as serving their interests”.
rp Fo
Norway and the Netherlands
There were no comments from Norway respondents about the quality of UAT data since such surveys are not carried out in Norway, in part because of the good quality of data on HIV and AIDS from other sources.
ee
Two respondents from the Netherlands judged that large-scale UAT programmes were not necessary because named HIV testing is widespread. The quality of information obtained
rr
from named testing was felt to be robust and could be used to model estimates of HIV prevalence in the population. Coverage of pregnant women is very high and this group is
ev
considered a valid proxy for the heterosexual population. Named HIV testing among users of STI clinics was also presented as being a valid method for estimating prevalence. It was
ie
argued that, as long as the uptake of named testing by pregnant women and STI clinic users remains high (above 90%) then a national programme of UAT was not thought to be
w
necessary. However, notwithstanding the above point it was acknowledged that current methods of surveillance do not allow undiagnosed HIV infection to be measured accurately.
Ethical issues
ly
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
UK Respondents felt that, on balance, no harm was caused by the programme. A recurrent argument for this was that, in settings where anonymous testing takes place, people are offered a named HIV test:
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
11
BMJ Open
"I think the ethics are not in fact, are not a problem because everybody is now very strongly advised to have a test and therefore the residual rump who don't have a test, it's important to know what their rate of HIV positivity is." (UK1)
Some respondents, however, said there would be ethical concerns if people in these settings were not routinely offered a named test. Several respondents considered the ethics of UAT to be relatively unimportant, for instance in the context of more significant health-related problems in Africa:
rp Fo
"There are these enormous ethical, real ethical issues, and to me, this one … I just can't understand why one should put so much emphasis on it." (UK9)
UK11 suggested that the question of ownership of the sample was relevant, but that this could be argued both ways:
rr
ee
"Is it yours or is it somebody else's? ... If it's waste that's going to be chucked away then why shouldn't people use it if they find it useful? On the other hand, 'I gave
ev
permission for this test for my own benefit and I didn't give you permission to do things to help others.'" (UK11)
ie
There was universal agreement that changing to a system of informed consent would damage
w
the programme. There was a fear that introducing information about UAT into a clinical consultation could confuse patients. Patients might think they were getting a named test:
on
"You would lose someone's confidence and trust because they might not understand
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 12 of 53
the words or the reason behind why we're doing it, and they would think that you could get back to them and you did know." (UK12)
Other respondents argued that it would reduce time for discussing important clinical matters:
"When the midwife turns up, does she pull out a leaflet and say by the way, I'm putting this beside you here? … it would be a perversion of the consultation to divert it into discussing this ..." (UK5)
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
12
Page 13 of 53
Further concerns were that a system of informed consent would increase refusals and lead to bias in the surveys, and also the time and cost of gaining consent from patients:
"You know we've got ten minute appointments. It could take you five minutes, half of your time just to explain what this is all about. So doing it in a consultation is not realistic." (UK3) The respondent suggested that a system of clinic self-registration, using computers, could be used to explain anonymous testing.
rp Fo
All the respondents thought that prior notification was the right theoretical approach. However, some were concerned that, in practice, it might not be working:
"It's a reasonable thing in practical terms provided people are informed. So … if people don't know about their option for opting out and their option for finding out
ee
then the system hasn't been working." (UK11)
rr
Some suggested that the approach needed to be more "active" with information displayed visibly on clinic walls or in leaflets given out to each patient. One clinician, however,
ev
expressed his discomfort with his role in the programme. He described filling out the form for UAT blood testing in front of the patient, without explaining what he was doing, and
ie
sometimes after the patient had declined a named HIV test:
w
"You're doing something that sort of seems slightly underhand…you're not explaining it to the patient. Yeah, I'm not comfortable with that, I think one wants to be, to
on
communicate well and to be as open as possible with patients and that's slightly in conflict with that." (UK3)
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
USA
Consent was not deemed necessary initially in the USA for participation in UAT as patients had consented to the blood being collected in the first place, and this was seen as adequate. USA1 suggested that surveillance is a public good, meaning there are strong reasons for making it a universal obligation on citizens – and that notification that surveillance is occurring is different from asking for consent.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
13
BMJ Open
The blood spot survey continued until the publication of the first trials showing that vertical transmission of HIV could be dramatically reduced by treating mothers. When it became clear that infected women could be offered treatment, and therefore should know their HIV status, “it became impossible for the US Centre for Communicable Disease Control and Prevention (CDC) to continue that programme and it was dropped in the early 1990s along with all the other ones” (USA3). It had become difficult to reconcile public health benefits of UAT with individual rights. If CDC had not voluntarily withdrawn funding from UAT, the same respondent indicated that it was likely that Congress would have mandated cessation, as
rp Fo
there was a strong coalition in favour of this.
A related legal issue was brought by USA4, in that public health surveillance appears to go against “fundamental legal principles” in the USA, and constitutional scholars are shocked when they hear what is done.
ee
‘Once you begin to argue that the government can take your personally identifiable
rr
information for any reason it finds useful without your consent you've opened up and blown a hole in much of the fundamental legal principles in the country.’ (USA4)
ev
The Tuskegee episode (a study between 1932 and 1972 in Alabama in which black syphilitic
ie
Americans were not given treatment although thought they were) was mentioned by USA1 as a reason for USA citizens, particularly Afro-Americans, to be suspicious of surveillance
w
practices. In the context of opposition to named case reporting and a consequent lack of trust in surveillance, USA3 argued that there are probably cultural dimensions to issues such as
on
consent, and the acceptability of particular interventions may vary from country to country.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 53
“It [UAT] became more of a civil rights issue than a gay specific one. It was, the government knows something that you don’t and the attitude in the States, at least at the time, towards the government was much less trusting than in this country [i.e. UK]. People did not assume and still don’t assume that the government necessarily is looking out for your well being. And I think there’s a much more paranoid attitude towards the government than there is in this country.” (USA3)
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
14
Page 15 of 53
Related to these cultural dimensions, a particular ethical concern was that the US (and, in particular, CDC) supports UAT in developing countries, with the inherent contradiction that it has been rejected in the USA.
‘It seems to me the CDC can’t at the one time say it’s unethical to do it in the US and say it’s ethical to do it abroad. Different public agencies may come to different decisions, but the same agency it seems to me should be bound by some rule of consistency.’ (USA1)
rp Fo
USA2 felt very strongly that arrangements for UAT in developing countries are not ethical where pregnant women do not have the opportunity to have a named test, counselling and treatment.
Netherlands
ee
The respondents felt that more extensive UAT surveys would produce more accurate
rr
information for public health surveillance. However, inter-linked practical and ethical doubts about UAT were presented. One concern was that UAT might deter high risk groups from
ev
seeking testing and treatment. Another concern was that, if UAT focused only on high risk groups, such groups would risk being stigmatised. It was suggested that reporting of findings
ie
from UAT surveys of high risk groups would need to be carried out sensitively.
w
A particular issue was consent and maintaining trust with patients. One interviewee expressed that staff in HIV treatment centres and STI clinics demand a more stringent consent process
on
than required by the research ethics committees in the Netherlands in order to promote trust among patients.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Norway
A UAT system was never thought relevant, necessary or appropriate for Norway. In Norway it is not legal to test or analyse blood or biological material for infectious diseases for nondiagnostic purposes without consent unless the purpose is surveillance of an ‘epidemic’. One respondent felt this displayed a ‘moral grey zones’, as the law hinges on the definition of ‘epidemic’ [Norway 1].
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
15
BMJ Open
There is very little concern or awareness in Norway about ethical issues around UAT. However, there is a general willingness to provide information for public health purposes in Norway, and people are used to being surveyed. It was reported that many Norwegians have given blood to the Biobank, and give broad consent for their tissue/blood to be used for a wide range of research. A respondent indicated that many are willing to waive their consent in Norway because ‘there is significant trust’ in the health care system. However, another respondent expressed that there was a possible danger that UAT blood could be used to identify someone via the Home Office DNA database, and that researchers might be forced to
rp Fo
share information under a court order (the respondent mentioned that the murderer of the Swedish politician Anna Lindh in 2003 was convicted on the basis of a match with DNA held in the national biobank).
One respondent suggested that informed consent could be different in different contexts. In a
ee
country like Norway, with low prevalence and easy access to health care, the current surveillance system (non-UAT) was described as acceptable and functioning well. However,
rr
in a country in Africa (where this respondent was carrying out research), with a high rate of HIV positive pregnant women and a low rate of testing, routine testing with opt-out consent procedures would be morally acceptable.
w
DISCUSSION
ie
ev
Since its inception, unlinked anonymous testing (UAT) has been a subject of
on
epidemiological, ethical and policy debate.15 16 17 In this paper we have presented the results of qualitative interviews with key informants in four countries pragmatically selected because they represent a range of approaches to UAT. 18 A total of 23 respondents were interviewed,
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 53
differentially distributed between countries: 12 in the UK, four in the USA, four in Norway and three in the Netherlands. First a summary of key findings is presented (in the context of historical development of UAT in different countries, as well as what is already know of the subject) , followed by a discussion of how the findings from this paper fit into broader contemporary debates about public health ethics and policy.
Summary of key findings from interviews
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
16
Page 17 of 53
In the United Kingdom (UK), a balance of epidemiological and ethical considerations in the late 1980s led to the implementation of the ‘prior notification’ approach to UAT; this involves leaflets and posters informing clients that UAT is occurring, allowing opting out but not necessitating the gaining of informed consent.77 1010 Respondents stressed that the
Formatted: Endn
Formatted: Endn
rationale remains apt today, and that this approach provides data unrivalled in terms of epidemiological quality,99 19 20 and which is of use for health service planning.21 Ethical
Formatted: Endn
concerns were generally felt to be minimal today since clients are (also) usually offered
rp Fo
voluntary (named) testing in settings where UAT is occurring. However, there was mention of concern about the lack of full explanation, whether people do actually know UAT is happening, and ‘ownership’ of blood.22 It was indicated that an approach involving informed consent would be cumbersome, time-consuming and epidemiologically damaging.23
ee
In the United States (USA), UAT of blood for HIV was extensively used in the 1980s and early 1990s,24 but was stopped in the late 1990s as evidence emerged that anti-viral therapy in
rr
pregnant women could help reduce the vertical transmission of HIV from pregnant mother to child – named testing needed to be maximised so that HIV-positive pregnant women could be
ev
identified and offered advice and therapy.25 The interviews indicated that data from UAT were useful epidemiologically but, since UAT has stopped, there has been sufficient other
ie
data to plan services accordingly. Other issues raised include a general suspicion around epidemiological surveillance post-Tuskegee, cultural dimensions of UAT with respect to
w
minority groups, a lack of trust in the Government, and the related disconnect between surveillance (done to all) and health care provision (variable, and not done to all).66 1111 There
on
Formatted: Endn
Formatted: Endn
was also comment about the apparent contradiction in the contemporary support of the US Centre for Communicable Disease Control and Prevention (CDC) for UAT in low-income countries.26
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Interviews in the Netherlands showed that named (voluntary) testing, which has generally been widely taken up, has provided the main vehicle for surveillance of HIV.44 There is
Formatted: Endn
acknowledgement that large-scale UAT would provide additional epidemiological data, but this has been balanced against concerns of this leading to diminished voluntary testing – especially in at risk groups – through loss of trust in the healthcare system.27 Smaller UAT surveys have been intermittently undertaken in such groups (e.g. intravenous drug users, sex
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
17
BMJ Open
workers, migrants) as research studies, but have required individual consent (initially oral, now written).
Interviews in Norway revealed that UAT was never felt to be necessary epidemiologically, as HIV has not been a significant issue and the uptake of voluntary testing has been high.1313
Formatted: Endn
Surveillance has thus tended to focus on the monitoring of the diagnosed population (anonymised diagnosed cases), which has been supported by the gay community; the undiagnosed population is interestingly referred to as ‘dark figure(s)’. In Norway it is not
rp Fo
legal to test without consent unless there is an epidemic, although respondents reported few ethical concerns since people generally have trust in the health care system, and are willing to provide information for public health purposes (many give blood to the Biobank for a range of research purposes).28
ee
The findings of this research on the ethics of UAT of blood need consideration in the context of two broader contemporary debates within medical and public health ethics. The first
rr
debate centres on the boundaries of informed consent in health research, particularly in relation to the use of human tissue. The second debate focuses on the balance between
ev
individuals’ rights and the appropriate role of government3 in infringing on those rights for broader public health interests or concerns. These debates are inter-connected, although will
ie
be presented in turn.
w Informed consent and ethics: a changing debate
on
Respect for autonomy is the bioethical principle that is most likely to be jeopardised by a
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 53
programme of unlinked anonymous blood testing, since informed consent is not usually deemed necessary for an individual’s blood to be included; this has tended to be the basis for opposition to UAT.55 29 The findings from our research show, however, that such positions
Formatted: Endn
differ between countries, and have also changed with time.
Debates in the late 1980s and early 1990s demonstrated the opposing views of those who were concerned about patient consent,88 2425 and those who felt that the priority was to control what was seen as an epidemic.
1515
Formatted: Endn
Formatted: Endn
UAT programmes did go ahead in the USA and UK, and
Formatted: Endn
other high income countries: there was no requirement for consent in the USA, whereas in the
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
18
Page 19 of 53
UK an approach of ‘prior notification’ was adopted through leaflets and posters in clinics – consent is presumed but individuals have the opportunity to opt out.
Over the past two decades informed consent has emerged as ‘the ethical touchstone of medical research, and is now enshrined in practice and a range of international guidelines as one of the main preconditions of medical research’.2324 30 Some commentators argue that
Formatted: Endn
even anonymised tissue should not be used without consent. Trouet, for example, believes that the use of such material is in contravention of Article 22 of the Convention on Human
rp Fo
Rights and Biomedicine. This states that ‘when in the course of an intervention any part of a human body is removed, it may be stored and used for a purpose other than that for which it was removed, only if this is done in conformity with appropriate information and consent procedures’, and in accordance with Article 8 of the European Convention on Human Rights which protects the right to respect for private life.31
ee
Our research indicates that, rather than the universality of human rights discourse, a form of
rr
weak moral relativism operates in policy-making and practice around UAT.32 In the USA, the change in social context and debates about individual rights, along with evidence of reduction
ev
in mother-to-child vertical transmission, resulted in the abandonment of UAT in 1996.1212
Formatted: Endn
The fact that the USA continued to support UAT in international research overseas illustrates
ie
further an attitude of moral relativism, although it could be argued that differential health needs justified different approaches.
w
Interviews in the UK showed that the prior notification approach that remains in use today
on
still holds strong support – based on perception that there is minimal harm to the individual, and the value of high quality, unbiased epidemiological data on HIV (held to be the most
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
robust and accurate worldwide).33 Respondents felt that any change to UAT requiring informed consent would be cumbersome, time-consuming, and costly. According to the Nuffield Council on Bioethics it may also be ethically unnecessary: ‘... it is acceptable to collect and use anonymised data for assessing and predicting trends in infectious disease without consent, as long as any invasion of privacy is reduced as far as possible.’.34
This position is not, however, cross-culturally held. In the Netherlands, UAT has not been felt to be acceptable without informed consent, initially verbal and more recently written. The societal emphasis on individual autonomy, including in the health research environment, has
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
19
BMJ Open
been the paramount consideration.55 Northern European and Scandinavian countries are often
Formatted: Endn
held up as having a strong community ethos, with an expectation on individual involvement in endeavours for the public good.35 This was not reflected in UAT practice in the Netherlands, although was indicated to be important in Norway, albeit in a setting where UAT has never been taken up.
These findings indicate the importance of social and political context in relation to medicoethical debates, and how attitudes and policies can change in time within any given
rp Fo
jurisdiction.36 Differences exist between the moral evaluations and resultant UAT approaches between countries such as England and the Netherlands. Within the USA, in contrast, the practice of UAT has been re-evaluated with time. In the late 1980s and early 1990s the feared AIDS epidemic was seen as a public health emergency, there was no treatment available, transmission from mother to unborn child was not preventable, and the stigma and
ee
discrimination associated with the infection were more overt than today. As these factors shifted, so did the policy on UAT in the USA. The situation in low income countries today
rr
has created a perception of the acceptability of UAT akin to that in western countries two decades ago.
ie
ev Individual good and community (public) good
w
There has been criticism in recent times of government for its ‘nanny state’ approach to interventions in, for example, the areas of smoking, alcohol and diet. More recently the
on
notion of ‘stewardship’ in public health rejects these negative associations,37 and advocates that government should have a role to play in protecting the health of its citizens, preventing
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 53
them from harming themselves or others, and promoting healthy behaviour.[4] In this view, public health interventions should not be thought of as unwanted restrictions imposed by a ‘nanny’ on powerless children, but as actions that favour both individuals and society more widely as prescribed by a ‘steward’ administering and protecting citizens’ interests. The Nuffield Council on Bioethics, for example, argues that the state should act as steward both to individuals and to the population as a whole, but that individuals’ personal choices should also be taken into account.3435
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Formatted: Endn
20
Page 21 of 53
The subject of UAT sits squarely within the debate about public health stewardship for social good. Recently, commentators have argued that patients of a publicly supported health care system like the NHS have a moral duty to participate in scientific research,38 39 although most do not go so far as to articulate that such participation should be mandatory in practice.40 Nevertheless, UAT could be understood as something that individuals should feel they have a moral obligation to participate in, regardless of ethical concerns about individual rights and personally held views. Our own survey of over 400 users of two genitor-urinary medicine clinics (a parallel part of this research project) has found that a large majority of respondents
rp Fo
(89%) agree that people should have a responsibility to take part in research studies, and the same proportion would agree to the use of their blood in unlinked anonymous seroprevalence testing. However, when asked about the question of consent, 74% indicated that they should be asked to consent before their blood is used for public health surveillance.1414
Formatted: Endn
Notwithstanding this important latter point, such support for use of blood samples for
ee
prevalence surveys or for future research has not, however, been demonstrated in an earlier study in the UK41 or more recently in Egypt.42
rr
The way forwards perhaps is to ensure that patients – and the public more widely – are
ev
informed of the existence, purposes and value of disease surveillance arrangements, as well as the importance of contributing to future health research, while investigating arrangements
ie
for gaining consent which are not unnecessarily cumbersome or likely to create bias amongst those who agree that their blood be used.43
w
The issue of consent in seroprevalence surveillance, such as UAT, remains an important
on
instrument to gauge the moral direction of public health policies more broadly.44 Debate continues around how public health programmes can embrace a stewardship relationship,
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
rather than a nanny paternalistic one, while respecting autonomy and advocating social responsibilities. Further examination of why different countries have adopted different positions on seroprevalence surveillance will help illuminate our understanding of how social morality is embedded in public health policies and practice.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
21
BMJ Open
Acknowledgments The authors would like to thank the participants in the study. We also acknowledge the valuable advice given by Vivian Hope, Will Nutland, Cicely Marston, and Graham Hart.
Conflicts of interest AK is also Director of Public Health Strategy and Medical Director of the Health Protection Agency. The views expressed in this paper are those of the authors, and are not intended to
rp Fo
represent the views of the Health Protection Agency.
Funding
This research was supported by the Economic and Social Research Council (ESRC Grant Reference: RES-000-22-2096).
rr
Ethical approval
ee
The study was approved by the South West Research Ethics Committee on August 16th 2007 (reference 07/H0206/46), and by the London School of Hygiene & Tropical Medicine ethics
ev
committee on 3rd September 2007 (reference 5174).
ie
Notes [1]
w
The England and Wales programme is currently managed by the Health Protection
Agency (HPA) and the Institute of Child Health, University of London. The programme was
on
previously managed by the Communicable Disease Surveillance Centre, part of the Public Health Laboratory Service, which was subsumed into the Health Protection Agency in 2005.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 53
[2]
Health Protection Scotland (HPS) co-ordinates data collection in Scotland.
[3]
Those chosen had particular involvement or interest in UAT and ethics, but further
details cannot be provided to maintain their confidentiality. [4]
By government, the meaning here is health and public health policy makers within
government.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
22
Page 23 of 53
References
1
Nicoll A, Gill ON, Peckham C et al. The public health applications of unlinked
anonymous seroprevalence monitoring of HIV in the United Kingdom. International Journal of Epidemiology 2000;29:1-10. 2
Department of Health. Prevalence of HIV in England and Wales in 1997: Annual
Report of the Unlinked Anonymous Prevalence Monitoring Programme. London: Department
rp Fo
of Health, 1998. 3
UK Collaborative Group for HIV and STI Surveillance. Testing Times: HIV and
Other Sexually Transmitted Infections in the United Kingdom. London: Health Protection Agency, 2007. 4
ee
Bayer R, Lumey LH, Wan L. The American, British and Dutch responses to unlinked
anonymous HIV seroprevalence studies: an international comparison. AIDS 1990;4(4):283-
rr
290. 5
ev
Kopelman LM. Informed consent and anonymous tissue samples: the case of HIV
seroprevalence studies. Journal of Medicine and Philosophy 1994;19:525-52. 6
ie
Fairchild AL, Bayer R. Ethics and the conduct of public health surveillance. Science
2004;303:631-632. 7
w
Gill ON., Adler MW, Day NE. Monitoring the prevalence of HIV: foundations for a
on
Programme of unlinked anonymous testing in England and Wales. British Medical Journal 1989:299:1295-7. 8
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Berridge V. AIDS in the UK: the Making of Policy 1981-1994. Oxford: Oxford
University Press, 1996. 9
Health Protection Agency. Overall HIV prevalence. Health Protection Agency 2008.
http://www.hpa.org.uk/web/HPAweb&Page&HPAwebAutoListName/Page/1201094588821. 10
Gillon R. Testing for HIV without permission. British Medical Journal 1987;294:821-
823.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
23
BMJ Open
11
Fairchild AL, Bayer R, Colgrove J. Searching Eyes: Privacy, the State, and Disease
Surveillance in America. Berkeley: University of California Press, 2007. 12
Stoto MA. Almario DA, McCormick MC. Reducing the Odds: Preventing Perinatal Transmission of HIV in the United States. Washington: National Academy Press, 1999.
13
Aavitsland P, Nilsen Ø, Lystad A. Anonymous reporting of HIV infection: an
evaluation of the HIV/AIDS surveillance system in Norway 1983-2000. European Journal of Epidemiology 2001;17:479-89. 14
rp Fo
Datta J, Kessel AS, Wellings K, Nanchahal K, Marks D, Kinghorn G. The views of
genitourinary medicine (GUM) clinic users on unlinked anonymous testing of blood for HIV: evidence from a pilot study of clinics in two English cities. Journal of Medical Ethics 2011 doi:10.1136/jme.2011.042705. 15
Doll R. A proposal for doing prevalence studies of AIDS. British Medical Journal
rr
1987;294:443-444. 16
ee
Avins AL, Lo B. To tell or not to tell: the ethical dilemmas of HIV test notification in
ev
epidemiologic research. American Journal of Public Health 1989;79/11:1544-7. 17
Datta J, Kessel AS. Unlinked anonymous testing for blood purposes: an ethical
ie
dilemma? In: Peckham S, Hann A, eds. Public Health Ethics and Practice. Bristol: Policy Press, 2010:101-116.
w
18
Bowling A. Research Methods in Health. Buckingham: Open University Press, 2000.
19
Duong T, Ades AE, Rogers P, et al. Non-participation bias in unlinked anonymous
ly
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 24 of 53
HIV-prevalence surveys in England and Wales. Epidemiology and Infection 1999;122:26772. 20
Pulver WP, Glebatis D, Wade N, et al. Trends from an HIV seroprevalence study
among childbearing women in New York State from 1988 through 2000: a valuable epidemiologic tool. Archives of Pediatric and Adolescent Medicine 2004;158(5):443-8. 21
Kessel AS, Watts CJ. Usefulness of information from the Unlinked Anonymous
Prevalence Monitoring Programme for HIV in England and Wales: survey of planners of HIV/AIDS services. International Journal of STD and AIDS 1999;10:808-11.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
24
Page 25 of 53
22
Pfeffer N, Laws S. 'It's only a blood test': what people know and think about
venepuncture and blood. Social Science & Medicine 2006;62:3011-3023. 23
Cassell J, Young A. Why we should not seek individual informed consent for
participation in health services research. Journal of Medical Ethics 2002;23:313-317. 24
Bayer R. The ethics of blinded HIV surveillance testing. American Journal of Public
Health 1993;83/4:496-7. 25
rp Fo
de Zulueta P. The ethics of anonymised HIV testing of pregnant women: a
reappraisal. Journal of Medical Ethics 2000;26:16-21. 26
Rennie S, Norris Turner A, Mupenda B, et al. Conducting unlinked anonymous HIV
ee
surveillance in developing countries: ethical, epidemiological, and public health concerns. PLoS Medicine 2009;6(1):e1000004. 27
rr
Postema EJ, Willems PWJM, de Ridder MAJ, et al. Comparison of patients refusing
with patients accepting unlinked anonymous HIV testing in an outpatient STD department in
ev
the Netherlands. International Journal of STD and AIDS 1997;8:368-72. 28
Skolbekken J-A, Ursin LO, Solberg B, et al. Not worth the paper it’s written on?
ie
Informed consent and Biobank research in a Norwegian context. Critical Public Health 2005;15(4):335-347. 29
on
Beauchamp TL and Childress JF. Principles of Biomedical Ethics. New York: Oxford
University Press, 1994.
Singleton P, Wadsworth M. Consent for the use of personal medical data in research.
British Medical Journal 2006;333:255-8. 31
ly
30
w
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Trouet C. New European guidelines for the use of stored human biological materials
in biomedical research. Journal of Medical Ethics 2004;30:99-103. 32
Rachels J. The Elements of Moral Philosophy. New York: McGraw-Hill, 1993.
33
Beuhler JW, Petersen LR, Ward JW, et al. Defending HIV seroprevalence surveys.
American Journal of Public Health 1994;84:319-320.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
25
BMJ Open
34
Nuffield Council on Bioethics. Public Health: Ethical Issues. London: Nuffield
Council on Bioethics, 2007 35
Cappelen AW, Norheim OF. Responsibility in health care: a liberal egalitarian
approach. Journal of Medical Ethics 2005;31:476-80. 36
Mariner WK. Mission creep: public health surveillance and medical privacy. Boston
University Law Review 2007;87/2:347-95. 37
rp Fo
Jochelson K. Nanny or Steward? The Role of Government in Public Health. King's
Fund Working Paper. London: King's Fund 2005. 38
Harris J. Scientific research is a moral duty. Journal of Medical Ethics
2005;31(4):242-248.
ee
39
Evans HM. Do patients have duties? Journal of Medical Ethics 2007;33(12):689-694.
40
Shapshay S, Pimple KD. Participation in biomedical research is an imperfect moral
rr
duty: a response to John Harris. Journal of Medical Ethics 2007;33:414-417. 41
ev
Kessel AS, Watts C, Weiss HA. Bad blood? Survey of public's views on unlinked
ie
anonymous testing of blood for HIV and other diseases. British Medical Journal 2000;320: 90-91. 42
w
Abou-Zeid A, Silverman H, Shehata M, et al. Collection, storage and use of blood
on
samples for future research: views of Egyptian patients expressed in a cross-sectional survey. Journal of Medical Ethics 2010;36:539-547. 43
Nicholls SG. Knowledge or understanding? Informed choice in the context of
newborn bloodspot screening. Public Health Ethics 2010;3:128-136. 44
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 26 of 53
Haimes E. What can the social sciences contribute to the study of ethics? Theoretical,
empirical and substantive considerations. Bioethics 2002;16/2:89-113.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
26
Page 27 of 53
Original paper
The ethics of unlinked anonymous testing of blood: views from in-depth interviews with key informants in four countries Unlinked anonymous testing of blood (running title)
rp Fo
Anthony S Kessel (corresponding author and guarantor of study), Honorary Professor, London School of Hygiene & Tropical Medicine, Faculty of Public Health and Policy, 15-17 Tavistock Place, London WC1H 9SH (Tel: 020 7759 2715; email:
[email protected])
Jessica Datta, Research Fellow, London School of Hygiene & Tropical Medicine
ee
Kaye Wellings, Professor, London School of Hygiene & Tropical Medicine
Sarah Perman, Health Protection Agency
Key words
6437 words
ly
on
Word count
w
Unlinked, anonymous, blood, ethics, public health, policy, HIV
ie
ev
rr
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
Abstract In this study we explore the ethical issues around unlinked anonymous testing (UAT) of blood, a method of sero-prevalence surveillance for infectious diseases. Our study focused on UAT for HIV, although UAT can be used for other infectious diseases.
Objectives
rp Fo
The objectives of the research were to gain a better understanding of the views of key informants in countries adopting different UAT testing strategies, and to use the findings of the research to inform health policy
Design
ee
Qualitative study using in-depth interviews and ethical analysis.
Setting
rr
Four countries using different strategies around unlinked anonymous testing of blood for HIV (United Kingdom, United States, Netherlands and Norway).
Participants Twenty-three key informants in the four countries.
w
ie
Results
ev
Participants from the four countries have different views on UAT of blood, and that the approaches and policies on UAT adopted by different countries have been historically and culturally determined. We use our findings to explore the relationship between public health policy and ethics, framing our
on
discussion in relation to two important contemporary debates: informed consent for participation in medical and public health research; and the balance between individual rights good and the public good.
Conclusions
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 28 of 53
Qualitative research and ethical analysis of unlinked anonymous testing of blood in different countries has yielded important findings for consideration by policy makers. The policy of unlinked anonymous testing of blood for HIV and other diseases in the UK needs re-consideration in light of these findings.
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 29 of 53
Public health policy, context and ethics: the example ofThe ethics of unlinked anonymous testing of blood: views from in-depth interviews with key informants in four countries unlinked anonymous testing of blood
In this paper we explore unlinked anonymous testing (UAT) of blood for HIV and other
rp Fo
diseases, and examine the ethical issues associated with this method of sero-prevalence surveillance for infectious diseases. We report original results from a study funded by the UK Economic and Social Research Council (ESRC) – qualitative interviews with stakeholders in four countries adopting different UAT policies – and use the subject matter, and our findings, to explore the relationship between public health policy, context and ethics.
BACKGROUND
rr
ee
Unlike voluntary testing for clinical purposes, unlinked anonymous testing (UAT) of blood is carried out in order to provide information about the prevalence of particular infectious
ev
diseases in a population. The use of UAT began in the late 1980s for HIV, and remains most commonly associated with this infection, but can be applied to other infectious diseases such
ie
as hepatitis C.1 In this paper we use the term UAT synonymously with sero-prevalence surveys, or sero-prevalence surveillance.
w
The principles behind UAT can be explained with reference to the approach used in the UK
on
for HIV. Leftover blood from samples taken for diagnostic or routine screening purposes is made anonymous, irreversibly unlinked from the source, and then tested for HIV.2 3 This
means that individuals whose blood tests positive for HIV cannot be contacted or informed.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
The data derived from UAT are then used for monitoring purposes and also to inform the planning and evaluation of health services and health promotion activities.
Historically, UAT has developed in different ways in different settings, yet exploration of the moral dimensions of such developments has been relatively limited.4 5 6 Further, recent dominant attention given to individual rights and autonomy – in the health and healthcare context – make it an appropriate time to reassess moral considerations around UAT. A brief
3
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
summary of developments in UAT in the four countries used in our study is provided as a necessary introduction to the research project.
History of UAT in different countries
A national sero-prevalence surveillance programme using UAT to estimate the prevalence of
rp Fo
HIV has been in place in the UK since 1990.111 The programme was first developed in
Formatted: Endnote Reference
7
response to concerns about the spread of HIV during the 1980s and, despite early opposition,
Formatted: Endnote Reference
the methodology eventually gained political support.8 The programme was funded initially as a pilot project by the Medical Research Council, and has since been funded by the Department of Health. [1] [2] The programme currently uses data collected from three
ee
ongoing surveys (newborn babies, [maternal antibodies], users of genitourinary medicine (GUM) clinics, and intravenous drug users [saliva]) to measure the proportion of individuals within the general population, and within subgroups of the population, infected with HIV.
rr
Since the programme’s inception, around ten million blood samples have been tested in this way.9
ev
A system of ‘prior notification’ operates in locations where testing goes on, involving the display of leaflets or posters explaining that blood may be tested anonymously. Consent to
ie
participation is implicit rather than explicit; provision is made to opt out of testing, although this naturally depends on awareness.999 10
w
Formatted: Endnote Reference Formatted: Endnote Reference
In the USA, UAT was extensively used in the late 1980s and early 1990s using discarded
on
blood samples. The largest survey was national and used newborn heel prick tests. Other UAT surveys focused on population groups, for instance a survey of young people
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 30 of 53
volunteering for job corps, although these were regional or local. In contrast to the UK the prior notification system, with the opportunity to opt out, was not used. UAT essentially stopped in the USA in 1996, as evidence emerged in relation to reducing vertical
transmission of the HIV virus from a pregnant mother to the foetus through use of anti-viral medicine.11 The interplay at the time of clinical, social and political factors was complex.12
The main method for HIV surveillance in the Netherlands is named voluntary testing rather than UAT (these are ‘opt-out’ programmes, such as the antenatal testing programme and the 4
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 31 of 53
testing of users of STI clinics, as well as the extensive ‘opt-in’ testing carried out by general practitioners). The UAT that does occur in the Netherlands is carried out using blood samples from intravenous drug users (IDUs) and saliva samples from other groups. Unlike the UK or USA, informed consent is required from those surveyed by UAT, either verbal or, more recently, written consent. From 2002 to 2006, other high risk groups haven been included in UAT research surveys in the Netherlands (e.g. migrants from the Cape Verde islands, Surinam and the Netherlands Antilles, as well as commercial sex workers), because of
rp Fo
concerns of a higher proportion of undiagnosed HIV in such groups, who were also felt to be accessing health services less than others.
In Norway, public health efforts have focused on diagnosing all those infected, as it was thought (certainly in earlier stages) that those who were diagnosed would be less likely than
ee
those undiagnosed to transmit infection to others. Mandatory named AIDS reporting was introduced in Norway in 1983 and mandatory anonymous HIV case reporting in 1986, both including cases diagnosed before those dates.13 The Norwegian approach has therefore
rr
centred on individual testing and clinician reporting, and UAT has not been deemed epidemiologically necessary and has never been undertaken.
AIMS AND OBJECTIVES OF THE RESEARCH STUDY
w
ie
ev The ESRC-funded project had three components: a questionnaire survey of 471 patients attending two inner-urban genito-urinary medicine (GUM) clinics in two English cities,
on
exploring their views of, and attitudes towards, UAT;14 qualitative in-depth interviews with a sub-sample (20) of these patients; and in-depth interviews with key informants in four
countries chosen to represent different approaches towards sero-surveillance for HIV. In this
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
paper, we present findings from this third component, the qualitative interviews with key informants.
The over-arching aim of this component of the research study was to use the subject of UAT, and our findings, to explore the relationship between public health policy, context and ethics. There were two more specific objectives of the research. The first was to gain a better understanding of how UAT has developed in different settings over time, with particular
5
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
attention to the ethical considerations. The second was to use UAT as an instrument to gain empirical understanding of the ethical tensions between surveillance and patients’ rights, and to explore how these tensions may have developed and been handled in different contexts.
METHODS
rp Fo
Four countries were selected purposively to participate in this study component (the key informant interviews), based on their representing a range of approaches, or positions taken, with regard to UAT: the UK (specifically England), USA, the Netherlands and Norway. Purposive sampling was used to select the key informants from these countries, informed by the following criteria: range of stakeholder perspectives (e.g. epidemiology, clinical, public
ee
health policy, charity or pressure group, and academia (public health or medical ethics)); involvement in UAT and with appropriate knowledge or expertise, historically and/or current; existing links with the research team or advisory group members; and practical issues (e.g.
rr
availability and likelihood of participation). A pragmatic approach meant that more informants were interviewed in the UK than the other countries.
ev
A topic guide was developed that allowed for a semi-structured interview. Elements within the topic guide included: introduction and background to the purpose and nature of the
ie
interview (prior information had been provided); knowledge and views on unlinked
w
anonymous testing of blood (and national programme as appropriate), including advantages and disadvantages; ethical considerations including consent, prior notification,
confidentiality, anonymity, individual rights/liberties, social responsibilities, and any changes
on
in ethical position over time; the value to public health of different surveillance methods;
blood as a tissue; legal considerations; and any other relevant issues the respondent wished to
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 32 of 53
discuss. The topic guide was initially tailored for the different countries based on our own understanding, but then applied flexibly during the interview.
A researcher, with a background and skills in qualitative interviewing, was employed for the duration of the project, and undertook all the interviews. Interviews were carried out, in English, either face-to-face or by telephone between January and September 2008. All nonUK informants spoke good English, without evidence to the research team of any obvious language barrier issues. Interviews were recorded (except for two, when notes were taken: the 6
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 33 of 53
first due to unsuitable location and the second to mechanical failure) with the participant’s permission, and transcribed verbatim. Interviews lasted between 20 minutes and one hour, the length directed by the contributions of the informants. In total 21 interviews were completed with 23 respondents (two interviews were with two respondents together). Of these, 12 were with UK respondents (three clinicians in genito-urinary medicine/HIV; one person working in government health policy; two HIV epidemiologists; two ethics academics; two charity representatives [MedFash, Terence Higgins Trust]; one person working at the British Medical
rp Fo
Association; one editor of a medical journal), four with respondents in the USA (one health policy academic; one academic in health law; one academic in public health; and one epidemiologist), three with respondents from the Netherlands (two epidemiologists; and one academic in medical ethics), and four with respondents in Norway (one epidemiologist; three academics in medical ethics – one of whom is also a hospital clinician). There were three interview refusals, all in the UK. [3]
ee
Interview data were coded by Two the researchers (primary researcher, JD, and a second
rr
researcher, SP) coded the interview data using a thematic content analysis approach, and looking for emergent themes emergent themes assessed by the same researcher (open coding
ev
in the context of the aims and objectives of the research); a second researcher assisted with coding and analysis. There were no discrepancies between the two researchers that required further discussion and consensus. Inevitably some of the themes inter-link with each other
ie
(e.g. ‘value of UAT’ and ‘ethical issues’). Content was analysed for description of UAT
w
developments in the country concerned, the context in which UAT may have changed, ethical tensions in UAT between surveillance and human rights, and broader connections between UAT and public health policy. Inevitably some of the themes inter-link with each other (e.g.
on
‘value of UAT’ and ‘ethical issues’). Coding frames and analyses were shared with the wider research team.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Interview data pertaining to the history and state of sero-surveillance in different countries
Formatted: Font: Not Bold
were triangulated with literature sources. It was not appropriate to triangulate informants’ views on sero-surveillance.
7
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
RESULTS Formatted: Font: Not Bold
The results, which represent the views and perceptions of informants, are presented by
Formatted: Font: Not Bold
themes and by country. The labels describe the themes that emerged, in the context of informants’ awareness of the aims and objectives of the study.
Value of UAT
UK
rp Fo
The programme is largely seen by respondents as indispensable in providing an unbiased measure of HIV prevalence and risk factors in the UK population. The most important outcome from the surveys has been information on undiagnosed HIV infection which,
ee
respondents suggested, has led to changes in policy on named testing; for example, HIV testing in pregnant women has changed from opt in to opt out, and sexual health clinics have moved from targeted to universal voluntary testing. The neonatal dried blood spot survey is
rr
used as a check on the performance of local antenatal screening. The public health rationale for the programme has not changed since 1990, and respondents cited the continuing high
ev
infection rate, particularly among men who have sex with men, as justification:
"There was far more feeling of a public health emergency (then) than there is today,
ie
although you could argue … that it's actually much worse today because of where we see the disease." (UK4)
w
Several respondents suggested that a review of UAT would be timely, including
on
consideration of extending the surveys to other health care settings and other diseases. UK6
discussed how advances in HIV testing, including self-testing, might alter the questions asked
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 34 of 53
by the programme. Some suggested that if named testing was normalised, there would be less need for anonymous testing.
A particular theme was whether the British public has a duty to participate in public health monitoring, possibly as part-and-parcel of receiving the benefits of a national health service (NHS):
8
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 35 of 53
"There is a very good case for really exploring with the public, and then pushing very hard to get public agreement, that in general terms that sort of tissue, extra bits of tissue or blood, should be available for use." (UK4)
USA According to one respondent, the surveillance programme was valuable in terms of planning, resource allocation, advocacy and interventions to prevent transmission. The heel prick
rp Fo
survey provided accurate national data on prevalence in childbearing women and was seen as extremely valuable, with nothing since that has given as accurate a picture of national prevalence (USA 2).
However, given the questionable ethics of UAT (see later), as well as the resultant lack of
ee
good quality prevalence data, USA3 argued that enough became known about prevalence to target the most vulnerable groups for health promotion.
rr
“And how much more detail do you want from that, so what would you do differently if you knew that in the next level of detail.” (USA3)
Netherlands
ev
Unlinked anonymous surveys have been used in the Netherlands to measure HIV prevalence
ie
among high risk populations, but with informed consent (verbal orf written). The populations
w
surveyed by UAT have changed over time. Injecting Drug Users (IDUs) have been surveyed bi-annually since 1994 in different cities in the Netherlands, and are recruited via methadone care; sex workers and migrant groups have also been subject to UAT surveys.
on
There are two reasons, indicated by respondents, why UAT in Netherlands has been of
limited value. First, the uptake of voluntary testing is generally high in the Netherlands,
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
meaning the quality of that data are generally good. And second, UAT has only been deemed ethically permissible in the Netherlands with informed consent, making large-scale UAT surveys impracticable, so restricting UAT to specific population groups and/or the research context.
Norway
9
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
The unknown, undiagnosed segment of the population is known as the ‘dark figure(s)’ in Norwegian, although neither the ‘dark figure’ nor the number of cases of HIV has been perceived as significant issues in Norway. Uptake of voluntary testing for HIV was described as very high in intravenous drug users and antenatal clinic users, meaning that UAT has never been perceived as to be of value epidemiologically. The system of surveillance for HIV in Norway instead covers all diagnosed cases (i.e. not undiagnosed cases as with UAT), and was reported as being simple, inexpensive and capable of providing a representative and timely
rp Fo
overview that guides prevention, while respecting confidentiality and adhering to ethical principles.
Quality and nature of data
ee
UK
The epidemiological data yielded from unlinked anonymous testing are considered by
rr
respondents to be of unrivalled quality internationally. It was suggested, however, that the apparent high number of undiagnosed HIV infections from UAT surveys might be inflated by
ev
double-counting:
"You can quite understand that if you've got gonorrhoea and you've been told by your
ie
doctor not to have unsafe sex, you might well decide to go to a different clinic to have your gonorrhoea treated." (UK1)
w
Others defended the statistics and argued that they were explained by a resurgence of high
on
risk behaviour, particularly among gay men. UK1 argued that people who refuse a named test and who are subsequently found to be HIV positive through UAT are likely to be people who know that they are positive.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 36 of 53
USA There was little comment from USA interviewees about the quality of UAT data, since UAT surveys stopped in 1996. However, a theme that emerged in relation to UAT data was around the difference between research and surveillance.
10
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 37 of 53
The fact that there is no ‘right’ to health care in the USA was felt to have an effect on people’s views on surveillance – in other words, the ‘disconnect’ between health care provision (not available to all) and surveillance (done to all) (USA 1). The differences between surveillance, research and practice, with perhaps different ethical standards required for each, emerged as a theme from the interviews (USA 4).
One point of view was that surveillance (because it is named and part of routine care) is part
rp Fo
of clinical care rather than purely research (USA 2). However, while some of the surveillance undertakings at state level in the USUSA were considered to be practice, at the level of CDC they were treated as research (USA 1). The same respondent commented that cancer sufferers advocated in favour of surveillance for the purposes of research on the disease, since “they were willing to trade privacy and consent for a good which was
ee
surveillance that they saw as serving their interests”.
Norway and the Netherlands
rr
There were no comments from Norway respondents about the quality of UAT data since such surveys are not carried out in Norway, in part because of the good quality of data on HIV and AIDS from other sources.
ev
Two respondents from the Netherlands judged that large-scale UAT programmes were not
ie
necessary because named HIV testing is widespread. The quality of information obtained
w
from named testing was felt to be robust and could be used to model estimates of HIV prevalence in the population. Coverage of pregnant women is very high and this group is considered a valid proxy for the heterosexual population. Named HIV testing among users of
on
STI clinics was also presented as being a valid method for estimating prevalence. It was
argued that, as long as the uptake of named testing by pregnant women and STI clinic users remains high (above 90%) then a national programme of UAT was not thought to be necessary. However, notwithstanding the above point it was acknowledged that current
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
methods of surveillance do not allow undiagnosed HIV infection to be measured accurately.
Ethical issues
UK 11
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
Respondents felt that, on balance, no harm was caused by the programme. A recurrent argument for this was that, in settings where anonymous testing takes place, people are offered a named HIV test:
"I think the ethics are not in fact, are not a problem because everybody is now very strongly advised to have a test and therefore the residual rump who don't have a test, it's important to know what their rate of HIV positivity is." (UK1)
rp Fo
Some respondents, however, said there would be ethical concerns if people in these settings were not routinely offered a named test. Several respondents considered the ethics of UAT to be relatively unimportant, for instance in the context of more significant health-related problems in Africa:
ee
"There are these enormous ethical, real ethical issues, and to me, this one … I just can't understand why one should put so much emphasis on it." (UK9)
rr
UK11 suggested that the question of ownership of the sample was relevant, but that this could be argued both ways:
ev
"Is it yours or is it somebody else's? ... If it's waste that's going to be chucked away
ie
then why shouldn't people use it if they find it useful? On the other hand, 'I gave
w
permission for this test for my own benefit and I didn't give you permission to do things to help others.'" (UK11)
on
There was universal agreement that changing to a system of informed consent would damage the programme. There was a fear that introducing information about UAT into a clinical
consultation could confuse patients. Patients might think they were getting a named test:
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 38 of 53
"You would lose someone's confidence and trust because they might not understand the words or the reason behind why we're doing it, and they would think that you could get back to them and you did know." (UK12)
Other respondents argued that it would reduce time for discussing important clinical matters:
12
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 39 of 53
"When the midwife turns up, does she pull out a leaflet and say by the way, I'm putting this beside you here? … it would be a perversion of the consultation to divert it into discussing this ..." (UK5)
Further concerns were that a system of informed consent would increase refusals and lead to bias in the surveys, and also the time and cost of gaining consent from patients:
rp Fo
"You know we've got ten minute appointments. It could take you five minutes, half of your time just to explain what this is all about. So doing it in a consultation is not realistic." (UK3) The respondent suggested that a system of clinic self-registration, using computers, could be used to explain anonymous testing.
ee
All the respondents thought that prior notification was the right theoretical approach. However, some were concerned that, in practice, it might not be working:
rr
"It's a reasonable thing in practical terms provided people are informed. So … if people don't know about their option for opting out and their option for finding out then the system hasn't been working." (UK11)
ev
Some suggested that the approach needed to be more "active" with information displayed
ie
visibly on clinic walls or in leaflets given out to each patient. One clinician, however,
w
expressed his discomfort with his role in the programme. He described filling out the form for UAT blood testing in front of the patient, without explaining what he was doing, and sometimes after the patient had declined a named HIV test:
on
"You're doing something that sort of seems slightly underhand…you're not explaining it to the patient. Yeah, I'm not comfortable with that, I think one wants to be, to
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
communicate well and to be as open as possible with patients and that's slightly in conflict with that." (UK3)
USA Consent was not deemed necessary initially in the USA for participation in UAT as patients had consented to the blood being collected in the first place, and this was seen as adequate. USA1 suggested that surveillance is a public good, meaning there are strong reasons for 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
making it a universal obligation on citizens – and that notification that surveillance is occurring is different from asking for consent.
The blood spot survey continued until the publication of the first trials showing that vertical transmission of HIV could be dramatically reduced by treating mothers. When it became clear that infected women could be offered treatment, and therefore should know their HIV status, “it became impossible for the US Centre for Communicable Disease Control and
rp Fo
Prevention (CDC) to continue that programme and it was dropped in the early 1990s along with all the other ones” (USA3). It had become difficult to reconcile public health benefits of UAT with individual rights. If CDC had not voluntarily withdrawn funding from UAT, the same respondent indicated that it was likely that Congress would have mandated cessation, as there was a strong coalition in favour of this.
ee
A related legal issue was brought by USA4, in that public health surveillance appears to go against the “fundamental legal principles” in the USA, and constitutional scholars are shocked when they hear what is done.
ev
rr
‘Once you begin to argue that the government can take your personally identifiable information for any reason it finds useful without your consent you've opened up and blown a hole in much of the fundamental legal principles in the country.’ (USA4)
w
ie
The Tuskegee episode (a study between 1932 and 1972 in Alabama in which black syphilitic Americans were not given treatment although thought they were) was mentioned by USA1 as a reason for USA citizens, particularly Afro-Americans, to be suspicious of surveillance
on
practices. In the context of opposition to named case reporting and a consequent lack of trust in surveillance, USA3 argued that there are probably cultural dimensions to issues such as
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 40 of 53
consent, and the acceptability of particular interventions may vary from country to country.
“It [UAT] became more of a civil rights issue than a gay specific one. It was, the government knows something that you don’t and the attitude in the States, at least at the time, towards the government was much less trusting than in this country [i.e. UK]. People did not assume and still don’t assume that the government necessarily is looking out for your well being. And I think there’s a much more paranoid attitude towards the government than there is in this country.” (USA3) 14
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 41 of 53
Related to these cultural dimensions, a particular ethical concern was that the US (and, in particular, CDC) supports UAT in developing countries, with the inherent contradiction that it has been rejected in the USUSA.
‘It seems to me the CDC can’t at the one time say it’s unethical to do it in the US and say it’s ethical to do it abroad. Different public agencies may come to different
rp Fo
decisions, but the same agency it seems to me should be bound by some rule of consistency.’ (USA1)
USA2 felt very strongly that arrangements for UAT in developing countries are not ethical where pregnant women do not have the opportunity to have a named test, counselling and
ee
treatment.
Netherlands
rr
The respondents felt that more extensive UAT surveys would produce more accurate information for public health surveillance. However, inter-linked practical and ethical doubts
ev
about UAT were presented. One concern was that UAT might deter high risk groups from seeking testing and treatment. Another concern was that, if UAT focused only on high risk groups, such groups would risk being stigmatised. It was suggested that reporting of findings
ie
from UAT surveys of high risk groups would need to be carried out sensitively.
w
A particular issue was consent and maintaining trust with patients. One interviewee expressed that staff in HIV treatment centres and STI clinics demand a more stringent consent process
on
than required by the research ethics committees in the Netherlands in order to promote trust among patients.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Norway
A UAT system was never thought relevant, necessary or appropriate for Norway. In Norway it is not legal to test or analyse blood or biological material for infectious diseases for nondiagnostic purposes without consent unless the purpose is surveillance of an ‘epidemic’. One respondent felt this displayed a ‘moral grey zones’, as the law hinges on the definition of ‘epidemic’ [Norway 1].
15
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
There is very little concern or awareness in Norway about ethical issues around UAT. However, there is a general willingness to provide information for public health purposes in Norway, and people are used to being surveyed. It was reported that many Norwegians have given blood to the Biobank, and give broad consent for their tissue/blood to be used for a wide range of research. A respondent indicated that many are willing to waive their consent in Norway because ‘there is significant trust’ in the health care system. However, another respondent expressed that there was a possible danger that UAT blood could be used to
rp Fo
identify someone via the Home Office DNA database, and that researchers might be forced to share information under a court order (the respondent mentioned that the murderer of the Swedish politician Anna Lindh in 2003 was convicted on the basis of a match with DNA held in the national biobank).
ee
One respondent suggested that informed consent could be different in different contexts. In a country like Norway, with low prevalence and easy access to health care, the current surveillance system (non-UAT) was described as acceptable and functioning well. However,
rr
in a country in Africa (where this respondent was carrying out research), with a high rate of HIV positive pregnant women and a low rate of testing, routine testing with opt-out consent procedures would be morally acceptable.
w
ie
DISCUSSION
ev
Since its inception, unlinked anonymous testing (UAT) has been a subject of
epidemiological, ethical and policy debate.15 16 17 Different approaches to sero-surveillance
on
were initially adopted in different countries in response to the developing HIV/AIDS
epidemic.4 Over time some of those approaches have changed. The research findings presented in this paper are part of a larger project exploring the reasons behind such
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 42 of 53
differences. The first findings, a quantitative study of the views towards UAT of users of users of two genito-urinary clinics in England, have recently been published.14
In this paper we have presented the results of qualitative interviews with key informants in four countries pragmatically selected because they represent a range of approaches to UATof differences in approach to testing. It is important to note the limitations of the research.18 Only four settings (countries) were selected, but these were chosen deliberately as they
16
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 43 of 53
represent a range of approaches to UAT. Although more data would have been generated by adding further countries, this was not practical for this relatively small study. A total of 23 respondents were interviewed, a reasonable number for a study of this kind, but they were differentially distributed between countries: 12 in the UK, four in the USA, four in Norway and three in the Netherlands. This distribution (as well as the make-up of respondents) was chosen purposefully to gain useful insights in a qualitative piece of research of this nature, and was not intended to provide a quantitatively representative sample.19 Future study could
rp Fo
be undertaken with expanded numbers.
First a summary of key findings willis be presented (in the context of historical development of UAT in different countries, as well as what is already know of the subject) , followed by a discussion of how the findings from this paper fit into broader contemporary debates about public health ethics and policy.
Summary of key findings from interviews
ev
rr
ee
In the United Kingdom (UK), a balance of epidemiological and ethical considerations in the late 1980s led to the implementation of the ‘prior notification’ approach to UAT; this involves leaflets and posters informing clients that UAT is occurring, allowing opting out but
ie
not necessitating the gaining of informed consent.777 101010 Respondents stressed that the
Formatted: Endnote Reference Formatted: Endnote Reference
w
rationale remains apt today, and that this approach provides data unrivalled in terms of
Formatted: Endnote Reference
epidemiological quality,999 20 21 and which is of use for health service planning.22 Ethical
Formatted: Endnote Reference
concerns were generally felt to be minimal today since clients are (also) usually offered
on
voluntary (named) testing in settings where UAT is occurring. However, there was mention
Formatted: Endnote Reference Formatted: Endnote Reference
of concern about the lack of full explanation, whether people do actually know UAT is
happening, and ‘ownership’ of blood.23 It was indicated that an approach involving informed consent would be cumbersome, time-consuming and epidemiologically damaging.24
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
In the United States (USA), UAT of blood for HIV was extensively used in the 1980s and early 1990s,25 but was stopped in the late 1990s as evidence emerged that anti-viral therapy in pregnant women could help reduce the vertical transmission of HIV from pregnant mother to child – named testing needed to be maximised so that HIV-positive pregnant women could be identified and offered advice and therapy.26 The interviews indicated that data from UAT 17
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
were useful epidemiologically but, since UAT has stopped, there has been sufficient other data to plan services accordingly. Other issues raised include a general suspicion around epidemiological surveillance post-Tuskegee, cultural dimensions of UAT with respect to minority groups, a lack of trust in the Government, and the related disconnect between surveillance (done to all) and health care provision (variable, and not done to all).666 111111
Formatted: Endnote Reference
There was also comment about the apparent contradiction in the contemporary support of the
Formatted: Endnote Reference Formatted: Endnote Reference
US Centre for Communicable Disease Control and Prevention (CDC) for UAT in lowincome countries.27
rp Fo
Formatted: Endnote Reference
Interviews in the Netherlands showed that named (voluntary) testing, which has generally been widely taken up, has provided the main vehicle for surveillance of HIV.444 There is
Formatted: Endnote Reference
acknowledgement that large-scale UAT would provide additional epidemiological data, but
Formatted: Endnote Reference
ee
this has been balanced against concerns of this leading to diminished voluntary testing – especially in at risk groups – through loss of trust in the healthcare system.28 Smaller UAT surveys have been intermittently undertaken in such groups (e.g. intravenous drug users, sex
rr
workers, migrants) as research studies, but have required individual consent (initially oral, now written).
ev
Interviews in Norway revealed that UAT was never felt to be necessary epidemiologically, as HIV has not been a significant issue and the uptake of voluntary testing has been high.131313
ie
Formatted: Endnote Reference Formatted: Endnote Reference
Surveillance has thus tended to focus on the monitoring of the diagnosed population
w
(anonymised diagnosed cases), which has been supported by the gay community; the undiagnosed population is interestingly referred to as ‘dark figure(s)’. In Norway it is not legal to test without consent unless there is an epidemic, although respondents reported few
on
ethical concerns since people generally have trust in the health care system, and are willing to provide information for public health purposes (many give blood to the Biobank for a range of research purposes).29
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 44 of 53
The findings of this research on the ethics of UAT of blood need consideration in the context of two broader contemporary debates within medical and public health ethics. The first debate centres on the boundaries of informed consent in health research, particularly in relation to the use of human tissue. The second debate focuses on the balance between individuals’ rights and the appropriate role of government3 in infringing on those rights for
18
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 45 of 53
broader public health interests or concerns. These debates are inter-connected, although will be presented in turn.
Informed consent and ethics: a changing debate
Respect for autonomy is the bioethical principle that is most likely to be jeopardised by a
rp Fo
programme of unlinked anonymous blood testing, since informed consent is not usually deemed necessary for an individual’s blood to be included; this has tended to be the basis for opposition to UAT.555 30 The findings from our research show, however, that such positions
Formatted: Endnote Reference Formatted: Endnote Reference
differ between countries, and have also changed with time.
ee
Debates in the late 1980s and early 1990s demonstrated the opposing views of those who were concerned about patient consent,888 252525 and those who felt that the priority was to control what was seen as an epidemic.
151515
Formatted: Endnote Reference
UAT programmes did go ahead in the USA and
rr
Formatted: Endnote Reference Formatted: Endnote Reference
UK, and other high income countries: there was no requirement for consent in the USA,
Formatted: Endnote Reference
whereas in the UK an approach of ‘prior notification’ was adopted through leaflets and
Formatted: Endnote Reference
ev
Formatted: Endnote Reference
posters in clinics – consent is presumed but individuals have the opportunity to opt out.
Over the past two decades informed consent has emerged as ‘the ethical touchstone of
ie
medical research, and is now enshrined in practice and a range of international guidelines as one of the main preconditions of medical research’.242424 31 Some commentators argue that
w
Formatted: Endnote Reference Formatted: Endnote Reference
even anonymised tissue should not be used without consent. Trouet, for example, believes that the use of such material is in contravention of Article 22 of the Convention on Human
on
Rights and Biomedicine. This states that ‘when in the course of an intervention any part of a human body is removed, it may be stored and used for a purpose other than that for which it
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
was removed, only if this is done in conformity with appropriate information and consent
procedures’, and in accordance with Article 8 of the European Convention on Human Rights which protects the right to respect for private life.32
Our research indicates that, rather than the universality of human rights discourse, a form of weak moral relativism operates in policy-making and practice around UAT.33 In the USA, the change in social context and debates about individual rights, along with evidence of reduction in mother-to-child vertical transmission, resulted in the abandonment of UAT in 1996.121212
Formatted: Endnote Reference Formatted: Endnote Reference
19
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
The fact that the USA continued to support UAT in international research overseas illustrates further an attitude of moral relativism, although it could be argued that differential health needs justified different approaches.
Interviews in the UK showed that the prior notification approach that remains in use today still holds strong support – based on perception that there is minimal harm to the individual, and the value of high quality, unbiased epidemiological data on HIV (held to be the most
rp Fo
robust and accurate worldwide) (Buehler et al., 1994).34 Respondents felt that any change to UAT requiring informed consent would be cumbersome, time-consuming, and costly. According to the Nuffield Council on Bioethics it may also be ethically unnecessary: ‘... it is acceptable to collect and use anonymised data for assessing and predicting trends in infectious disease without consent, as long as any invasion of privacy is reduced as far as
ee
possible.’.35
This position is not, however, cross-culturally held. In the Netherlands, UAT has not been felt
rr
to be acceptable without informed consent, initially verbal and more recently written. The societal emphasis on individual autonomy, including in the health research environment, has
ev
been the paramount consideration.555 Northern European and Scandinavian countries are
Formatted: Endnote Reference Formatted: Endnote Reference
often held up as having a strong community ethos, with an expectation on individual involvement in endeavours for the public good.36 This was not reflected in UAT practice in
ie
the Netherlands, although was indicated to be important in Norway, albeit in a setting where UAT has never been taken up.
w
These findings indicate the importance of social and political context in relation to medico-
on
ethical debates, and how attitudes and policies can change in time within any given
jurisdiction the context can change with time and place.37 Differences exist between the
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 46 of 53
moral evaluations and resultant UAT approaches between countries such as England and the Netherlands. Within the USA, in contrast, the practice of UAT has been re-evaluated with
time. In the late 1980s and early 1990s the feared AIDS epidemic was seen as a public health emergency, there was no treatment available, transmission from mother to unborn child was not preventable, and the stigma and discrimination associated with the infection were more overt than today. As these factors shifted, so did the policy on UAT in the USA. The situation in low income countries today has created a perception of the acceptability of UAT akin to that in western countries two decades ago. 20
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 47 of 53
Individual rights good and public community (public) good
There has been criticism in recent times of government for its ‘nanny state’ approach to interventions in, for example, the areas of smoking, alcohol and diet. More recently the notion of ‘stewardship’ in public health rejects these negative associations,38 and advocates
rp Fo
that government should have a role to play in protecting the health of its citizens, preventing them from harming themselves or others, and promoting healthy behaviour.[4] In this view, public health interventions should not be thought of as unwanted restrictions imposed by a ‘nanny’ on powerless children, but as actions that favour both individuals and society more widely as prescribed by a ‘steward’ administering and protecting citizens’ interests.[5] The
ee
Nuffield Council on Bioethics, for example, argues that the state should act as steward both to individuals and to the population as a whole, but that individuals’ personal choices should also be taken into account.353535
Formatted: Endnote Reference
rr
Formatted: Endnote Reference
The subject of UAT sits squarely within the debate about public health stewardship for social
ev
good. Recently, commentators have argued that patients of a publicly supported health care system like the NHS have a moral duty to participate in scientific research,39 40 although most do not go so far as to articulate that such participation should be mandatory in practice.41
ie
Nevertheless, UAT could be understood as something that individuals should feel they have a
w
moral obligation to participate in, regardless of ethical concerns about individual rights and personally held views. Our own survey of over 400 users of two genitor-urinary medicine clinics (a parallel part of this research project) has found that a large majority of respondents
on
(89%) agree that people should have a responsibility to take part in research studies, and the
same proportion would agree to the use of their blood in unlinked anonymous seroprevalence
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
testing. However, when asked about the question of consent, 74% indicated that they should be asked to consent before their blood is used for public health surveillance.141414
Formatted: Endnote Reference Formatted: Endnote Reference
Notwithstanding this important latter point, such support for use of blood samples for prevalence surveys or for future research has not, however, been demonstrated in an earlier study in the UK42 or more recently in Egypt.43
The way forwards perhaps is to ensure that patients – and the public more widely – are informed of the existence, purposes and value of disease surveillance arrangements, as well 21
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
as the importance of contributing to future health research, while investigating arrangements for gaining consent which are not unnecessarily cumbersome or likely to create bias amongst those who agree that their blood be used.44
The issue of consent in seroprevalence surveillance, such as UAT, remains an important instrument to gauge the moral direction of public health policies more broadly.45 Debate continues around how public health programmes can embrace a Kantian stewardship
rp Fo
relationship, rather than a nanny paternalistic one, while respecting autonomy and advocating social responsibilities. Further examination of why different countries have adopted different positions on seroprevalence surveillance will help illuminate our understanding of how social morality is embedded in public health policies and practice.
Acknowledgments
w
ie
ev
rr
ee The authors would like to thank the participants in the study. We also acknowledge the
on
valuable advice given by Vivian Hope, Will Nutland, Cicely Marston, and Graham Hart.
Conflicts of interest
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 48 of 53
AK is also Director of Public Health Strategy and Medical Director of the Health Protection Agency. The views expressed in this paper are those of the authors, and are not intended to represent the views of the Health Protection Agency.
Funding This research was supported by the Economic and Social Research Council (ESRC Grant Reference: RES-000-22-2096).
22
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 49 of 53
Ethical approval The study was approved by the South West Research Ethics Committee on August 16th 2007 (reference 07/H0206/46), and by the London School of Hygiene & Tropical Medicine ethics committee on 3rd September 2007 (reference 5174).
rp Fo
Notes [1]
The England and Wales programme is currently managed by the Health Protection
Agency (HPA) and the Institute of Child Health, University of London. The programme was previously managed by the Communicable Disease Surveillance Centre, part of the Public Health Laboratory Service, which was subsumed into the Health Protection Agency in 2005.
ee
[2]
Health Protection Scotland (HPS) co-ordinates data collection in Scotland.
[3]
Those chosen had particular involvement or interest in UAT and ethics, but further
rr
details cannot be provided to maintain their confidentiality. [4]
By government, the meaning here is health and public health policy makers within
government. [5]
ev
It is interesting that in this discourse ‘nanny’ (female) is seen as negative and
‘steward’ (male) as positive, when both are servants employed to manage and protect.
w
ie
References
1
on
Nicoll A, Gill ON, Peckham C et al. The public health applications of unlinked
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
anonymous seroprevalence monitoring of HIV in the United Kingdom. International Journal of Epidemiology 2000;29:1-10. 2
Department of Health. Prevalence of HIV in England and Wales in 1997: Annual
Report of the Unlinked Anonymous Prevalence Monitoring Programme. London: Department of Health, 1998.
23
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
3
UK Collaborative Group for HIV and STI Surveillance. Testing Times: HIV and
Other Sexually Transmitted Infections in the United Kingdom. London: Health Protection Agency, 2007. 4
Bayer R, Lumey LH, Wan L. The American, British and Dutch responses to unlinked
anonymous HIV seroprevalence studies: an international comparison. AIDS 1990;4(4):283290. 5
rp Fo
Kopelman LM. Informed consent and anonymous tissue samples: the case of HIV
seroprevalence studies. Journal of Medicine and Philosophy 1994;19:525-52. 6
Fairchild AL, Bayer R. Ethics and the conduct of public health surveillance. Science
2004;303:631-632. 7
ee
Gill ON., Adler MW, Day NE. Monitoring the prevalence of HIV: foundations for a
Programme of unlinked anonymous testing in England and Wales. British Medical Journal
rr
1989:299:1295-7. 8
Berridge V. AIDS in the UK: the Making of Policy 1981-1994. Oxford: Oxford
University Press, 1996.
Health Protection Agency. Overall HIV prevalence. Health Protection Agency 2008.
ie
9
ev
http://www.hpa.org.uk/web/HPAweb&Page&HPAwebAutoListName/Page/1201094588821. 10
w
Gillon R. Testing for HIV without permission. British Medical Journal 1987;294:821-
823. 11
on
Fairchild AL, Bayer R, Colgrove J. Searching Eyes: Privacy, the State, and Disease
Surveillance in America. Berkeley: University of California Press, 2007.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 50 of 53
12
Stoto MA. Almario DA, McCormick MC. Reducing the Odds: Preventing Perinatal Transmission of HIV in the United States. Washington: National Academy Press, 1999.
13
Aavitsland P, Nilsen Ø, Lystad A. Anonymous reporting of HIV infection: an
evaluation of the HIV/AIDS surveillance system in Norway 1983-2000. European Journal of Epidemiology 2001;17:479-89.
24
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 51 of 53
14
Datta J, Kessel AS, Wellings K, Nanchahal K, Marks D, Kinghorn G. The views of
genitourinary medicine (GUM) clinic users on unlinked anonymous testing of blood for HIV: evidence from a pilot study of clinics in two English cities. Journal of Medical Ethics 2011 doi:10.1136/jme.2011.042705. 15
Doll R. A proposal for doing prevalence studies of AIDS. British Medical Journal
1987;294:443-444. 16
rp Fo
Avins AL, Lo B. To tell or not to tell: the ethical dilemmas of HIV test notification in
epidemiologic research. American Journal of Public Health 1989;79/11:1544-7. 17
Datta J, Kessel AS. Unlinked anonymous testing for blood purposes: an ethical
dilemma? In: Peckham S, Hann A, eds. Public Health Ethics and Practice. Bristol: Policy Press, 2010:101-116.
ee
18
Bowling A. Research Methods in Health. Buckingham: Open University Press, 2000.
19
Green J, Thorogood N. Qualitative Methods for Health Research. London: Sage,
2005.
ev
20
rr
Duong T, Ades AE, Rogers P, et al. Non-participation bias in unlinked anonymous
HIV-prevalence surveys in England and Wales. Epidemiology and Infection 1999;122:26772.
Pulver WP, Glebatis D, Wade N, et al. Trends from an HIV seroprevalence study
w
21
ie
among childbearing women in New York State from 1988 through 2000: a valuable epidemiologic tool. Archives of Pediatric and Adolescent Medicine 2004;158(5):443-8. 22
on
Kessel AS, Watts CJ. Usefulness of information from the Unlinked Anonymous
Prevalence Monitoring Programme for HIV in England and Wales: survey of planners of HIV/AIDS services. International Journal of STD and AIDS 1999;10:808-11. 23
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Pfeffer N, Laws S. 'It's only a blood test': what people know and think about
venepuncture and blood. Social Science & Medicine 2006;62:3011-3023. 24
Cassell J, Young A. Why we should not seek individual informed consent for
participation in health services research. Journal of Medical Ethics 2002;23:313-317.
25
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
25
Bayer R. The ethics of blinded HIV surveillance testing. American Journal of Public
Health 1993;83/4:496-7. 26
de Zulueta P. The ethics of anonymised HIV testing of pregnant women: a
reappraisal. Journal of Medical Ethics 2000;26:16-21. 27
Rennie S, Norris Turner A, Mupenda B, et al. Conducting unlinked anonymous HIV
rp Fo
surveillance in developing countries: ethical, epidemiological, and public health concerns. PLoS Medicine 2009;6(1):e1000004. 28
Postema EJ, Willems PWJM, de Ridder MAJ, et al. Comparison of patients refusing
with patients accepting unlinked anonymous HIV testing in an outpatient STD department in the Netherlands. International Journal of STD and AIDS 1997;8:368-72. 29
ee
Skolbekken J-A, Ursin LO, Solberg B, et al. Not worth the paper it’s written on?
Informed consent and Biobank research in a Norwegian context. Critical Public Health 2005;15(4):335-347.
Beauchamp TL and Childress JF. Principles of Biomedical Ethics. New York: Oxford
University Press, 1994.
Singleton P, Wadsworth M. Consent for the use of personal medical data in research.
British Medical Journal 2006;333:255-8.
w
32
ie
31
ev
30
rr
Trouet C. New European guidelines for the use of stored human biological materials
in biomedical research. Journal of Medical Ethics 2004;30:99-103.
on
33
Rachels J. The Elements of Moral Philosophy. New York: McGraw-Hill, 1993.
34
Beuhler JW, Petersen LR, Ward JW, et al. Defending HIV seroprevalence surveys.
American Journal of Public Health 1994;84:319-320. 35
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 52 of 53
Nuffield Council on Bioethics. Public Health: Ethical Issues. London: Nuffield
Council on Bioethics, 2007 36
Cappelen AW, Norheim OF. Responsibility in health care: a liberal egalitarian
approach. Journal of Medical Ethics 2005;31:476-80.
26
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 53 of 53
37
Mariner WK. Mission creep: public health surveillance and medical privacy. Boston
University Law Review 2007;87/2:347-95. 38
Jochelson K. Nanny or Steward? The Role of Government in Public Health. King's
Fund Working Paper. London: King's Fund 2005. 39
Harris J. Scientific research is a moral duty. Journal of Medical Ethics
rp Fo
2005;31(4):242-248. 40
Evans HM. Do patients have duties? Journal of Medical Ethics 2007;33(12):689-694.
41
Shapshay S, Pimple KD. Participation in biomedical research is an imperfect moral
duty: a response to John Harris. Journal of Medical Ethics 2007;33:414-417. 42
ee
Kessel AS, Watts C, Weiss HA. Bad blood? Survey of public's views on unlinked
anonymous testing of blood for HIV and other diseases. British Medical Journal 2000;320:
rr
90-91. 43
Abou-Zeid A, Silverman H, Shehata M, et al. Collection, storage and use of blood
ev
samples for future research: views of Egyptian patients expressed in a cross-sectional survey. Journal of Medical Ethics 2010;36:539-547. 44
ie
Nicholls SG. Knowledge or understanding? Informed choice in the context of
newborn bloodspot screening. Public Health Ethics 2010;3:128-136. 45
w
Haimes E. What can the social sciences contribute to the study of ethics? Theoretical,
empirical and substantive considerations. Bioethics 2002;16/2:89-113.
ly
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
27
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
The ethics of unlinked anonymous testing of blood: views from in-depth interviews with key informants in four countries
rp Fo Journal:
Manuscript ID: Article Type:
Date Submitted by the Author:
Complete List of Authors:
BMJ Open bmjopen-2012-001427.R2 Research 08-Nov-2012
Primary Subject Heading:
Keywords:
Public health
Ethics, Infectious diseases
ev
Secondary Subject Heading:
rr
ee
Kessel, Anthony; London School of Hygiene and Tropical Medicine, Social and environmental health research Datta, Jessica; London School of Hygiene and Tropical Medicine, Wellings, Kaye; London School of Hygiene and Tropical Medicine, Perman, Sarah; Health Protection Agency, Public Health Strategy
PUBLIC HEALTH, MEDICAL ETHICS, Health policy < HEALTH SERVICES ADMINISTRATION & MANAGEMENT
w
ie ly
on For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 1 of 53
Original paper
The ethics of unlinked anonymous testing of blood: views from in-depth interviews with key informants in four countries Unlinked anonymous testing of blood (running title) Anthony S Kessel (corresponding author and guarantor of study), Honorary Professor, London
rp Fo
School of Hygiene & Tropical Medicine, Faculty of Public Health and Policy, 15-17 Tavistock Place, London WC1H 9SH (Tel: 020 7759 2715; email:
[email protected])
Jessica Datta, Research Fellow, London School of Hygiene & Tropical Medicine
ee
Kaye Wellings, Professor, London School of Hygiene & Tropical Medicine
rr
Sarah Perman, Health Protection Agency
Key words
w
ie
ev Unlinked, anonymous, blood, ethics, public health, policy, HIV
Word count
ly
6608 words
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
1
BMJ Open
Abstract In this study we explore the ethical issues around unlinked anonymous testing (UAT) of blood, a method of sero-prevalence surveillance for infectious diseases. Our study focused on UAT for HIV, although UAT can be used for other infectious diseases. Objectives The objectives of the research were to gain a better understanding of the views of key informants in countries adopting different UAT testing strategies, and to use the findings of the research to inform
rp Fo
health policy Design
Qualitative study using in-depth interviews and ethical analysis.
ee
Setting
Four countries using different strategies around unlinked anonymous testing of blood for HIV (United
rr
Kingdom, United States, Netherlands and Norway). Participants
ev
Twenty-three key informants in the four countries.
ie
Results
Participants from the four countries have different views on UAT of blood, and the approaches and
w
policies on UAT adopted by different countries have been historically and culturally determined. We use our findings to explore the relationship between public health policy and ethics, framing our
on
discussion in relation to two important contemporary debates: informed consent for participation in medical and public health research; and the balance between individual good and the public good. Conclusions
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 53
Qualitative research and ethical analysis of unlinked anonymous testing of blood in different countries has yielded important findings for consideration by policy makers. The policy of unlinked anonymous testing of blood for HIV and other diseases in the UK needs re-consideration in light of these findings.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
2
Page 3 of 53
The ethics of unlinked anonymous testing of blood: views from in-depth interviews with key informants in four countries
In this paper we explore unlinked anonymous testing (UAT) of blood for HIV and other diseases, and examine the ethical issues associated with this method of sero-prevalence surveillance for infectious diseases. We report original results from a study funded by the UK Economic and Social Research Council (ESRC) – qualitative interviews with stakeholders in
rp Fo
four countries adopting different UAT policies – and use the subject matter, and our findings, to explore the relationship between public health policy, context and ethics.
BACKGROUND
ee
Unlike voluntary testing for clinical purposes, unlinked anonymous testing (UAT) of blood is
rr
carried out in order to provide information about the prevalence of particular infectious diseases in a population. The use of UAT began in the late 1980s for HIV, and remains most
ev
commonly associated with this infection, but can be applied to other infectious diseases such as hepatitis C.1 In this paper we use the term UAT synonymously with sero-prevalence surveys, or sero-prevalence surveillance.
w
ie
The principles behind UAT can be explained with reference to the approach used in the UK for HIV. Leftover blood from samples taken for diagnostic or routine screening purposes is
on
made anonymous, irreversibly unlinked from the source, and then tested for HIV.2 3 This means that individuals whose blood tests positive for HIV cannot be contacted or informed.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
The data derived from UAT are then used for monitoring purposes and also to inform the planning and evaluation of health services and health promotion activities.
Historically, UAT has developed in different ways in different settings, yet exploration of the moral dimensions of such developments has been relatively limited.4 5 6 Further, dominant attention given to individual rights and autonomy – in the health and healthcare context – make it an appropriate time to reassess moral considerations around UAT. A brief summary
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
3
BMJ Open
of developments in UAT in the four countries used in our study is provided as a necessary introduction to the research project.
History of UAT in different countries A national sero-prevalence surveillance programme using UAT to estimate the prevalence of HIV has been in place in the UK since 1990.1 The programme was first developed in
rp Fo
response to concerns about the spread of HIV during the 1980s7 and, despite early opposition, the methodology eventually gained political support.8 The programme was funded initially as a pilot project by the Medical Research Council, and has since been funded by the Department of Health. [1] [2] The programme currently uses data collected from three ongoing surveys (newborn babies, [maternal antibodies], users of genitourinary medicine
ee
(GUM) clinics, and intravenous drug users [saliva]) to measure the proportion of individuals within the general population, and within subgroups of the population, infected with HIV.
rr
Since the programme’s inception, around ten million blood samples have been tested in this way.9
ev
A system of ‘prior notification’ operates in locations where testing goes on, involving the
ie
display of leaflets or posters explaining that blood may be tested anonymously. Consent to participation is implicit rather than explicit; provision is made to opt out of testing, although this naturally depends on awareness.9 10
w
on
In the USA, UAT was extensively used in the late 1980s and early 1990s using discarded blood samples. The largest survey was national and used newborn heel prick tests. Other
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 53
UAT surveys focused on population groups, for instance a survey of young people volunteering for job corps, although these were regional or local. In contrast to the UK the prior notification system, with the opportunity to opt out, was not used. UAT essentially stopped in the USA in 1996, as evidence emerged in relation to reducing vertical transmission of the HIV virus from a pregnant mother to the foetus through use of anti-viral medicine.11 The interplay at the time of clinical, social and political factors was complex.12 The main method for HIV surveillance in the Netherlands is named voluntary testing rather than UAT (these are ‘opt-out’ programmes, such as the antenatal testing programme and the
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
4
Page 5 of 53
testing of users of STI clinics, as well as the extensive ‘opt-in’ testing carried out by general practitioners). The UAT that does occur in the Netherlands is carried out using blood samples from intravenous drug users (IDUs) and saliva samples from other groups. Unlike the UK or USA, informed consent is required from those surveyed by UAT, either verbal or, more recently, written consent. From 2002 to 2006, other high risk groups haven been included in UAT research surveys in the Netherlands (e.g. migrants from the Cape Verde islands, Surinam and the Netherlands Antilles, as well as commercial sex workers), because of concerns of a higher proportion of undiagnosed HIV in such groups, who were also felt to be
rp Fo
accessing health services less than others. In Norway, public health efforts have focused on diagnosing all those infected, as it was thought (certainly in earlier stages) that those who were diagnosed would be less likely than those undiagnosed to transmit infection to others. Mandatory named AIDS reporting was
ee
introduced in Norway in 1983 and mandatory anonymous HIV case reporting in 1986, both including cases diagnosed before those dates.13 The Norwegian approach has therefore
rr
centred on individual testing and clinician reporting, and UAT has not been deemed epidemiologically necessary and has never been undertaken.
ie
ev AIMS AND OBJECTIVES OF THE RESEARCH STUDY
w
The ESRC-funded project had three components: a questionnaire survey of 471 patients
on
attending two inner-urban genito-urinary medicine (GUM) clinics in two English cities, exploring their views of, and attitudes towards, UAT;14 qualitative in-depth interviews with a sub-sample (20) of these patients; and in-depth interviews with key informants in four
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
countries chosen to represent different approaches towards sero-surveillance for HIV. In this paper, we present findings from this third component, the qualitative interviews with key informants.
The over-arching aim of this component of the research study was to use the subject of UAT, and our findings, to explore the relationship between public health policy, context and ethics. There were two more specific objectives of the research. The first was to gain a better understanding of how UAT has developed in different settings over time, with particular
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
5
BMJ Open
attention to the ethical considerations. The second was to use UAT as an instrument to gain empirical understanding of the ethical tensions between surveillance and patients’ rights, and to explore how these tensions may have developed and been handled in different contexts.
METHODS Four countries were selected purposively to participate in this study component (the key
rp Fo
informant interviews), based on their representing a range of approaches, or positions taken, with regard to UAT: the UK (specifically England), USA, the Netherlands and Norway. Purposive sampling was used to select the key informants from these countries, informed by the following criteria: range of stakeholder perspectives (e.g. epidemiology, clinical, public health policy, charity or pressure group, and academia (public health or medical ethics));
ee
involvement in UAT and with appropriate knowledge or expertise, historically and/or current; existing links with the research team or advisory group members; and practical issues (e.g.
rr
availability and likelihood of participation). A pragmatic approach meant that more informants were interviewed in the UK than the other countries.
ev
A topic guide was developed that allowed for a semi-structured interview. Elements within
ie
the topic guide included: introduction and background to the purpose and nature of the interview (prior information had been provided); knowledge and views on unlinked
w
anonymous testing of blood (and national programme as appropriate), including advantages and disadvantages; ethical considerations including consent, prior notification,
on
confidentiality, anonymity, individual rights/liberties, social responsibilities, and any changes in ethical position over time; the value to public health of different surveillance methods;
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 53
blood as a tissue; legal considerations; and any other relevant issues the respondent wished to discuss. The topic guide was initially tailored for the different countries based on our own understanding, but then applied flexibly during the interview.
A researcher, with a background and skills in qualitative interviewing, was employed for the duration of the project, and undertook all the interviews. Interviews were carried out, in English, either face-to-face or by telephone between January and September 2008. All nonUK informants spoke good English, without evidence to the research team of any obvious language barrier issues. Interviews were recorded (except for two, when notes were taken: the
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
6
Page 7 of 53
first due to unsuitable location and the second to mechanical failure) with the participant’s permission, and transcribed verbatim. Interviews lasted between 20 minutes and one hour, the length directed by the contributions of the informants. In total 21 interviews were completed with 23 respondents (two interviews were with two respondents together). Of these, 12 were with UK respondents (three clinicians in genito-urinary medicine/HIV; one person working in government health policy; two HIV epidemiologists; two ethics academics; two charity representatives [MedFash, Terence Higgins Trust]; one person working at the British Medical Association; one editor of a medical journal), four with respondents in the USA (one health
rp Fo
policy academic; one academic in health law; one academic in public health; and one epidemiologist), three with respondents from the Netherlands (two epidemiologists; and one academic in medical ethics), and four with respondents in Norway (one epidemiologist; three academics in medical ethics – one of whom is also a hospital clinician). There were three interview refusals, all in the UK. [3]
ee
Two researchers (primary researcher, JD, and a second researcher, SP) coded the interview
rr
data using a thematic content analysis approach, looking for emergent themes (open coding in the context of the aims and objectives of the research). There were no discrepancies between
ev
the two researchers that required further discussion and consensus. Content was analysed for description of UAT developments in the country concerned, the context in which UAT may
ie
have changed, ethical tensions in UAT between surveillance and human rights, and broader connections between UAT and public health policy. Inevitably some of the themes inter-link
w
with each other (e.g. ‘value of UAT’ and ‘ethical issues’). Coding frames and analyses were shared with the wider research team. In a relatively small study of this kind it is not possible
on
to indicate definitively if data saturation was reached, and the focus was on the emergent themes.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Interview data pertaining to the history and factual state of sero-surveillance in different countries were triangulated, for the purpose of the discussion section, with literature sources. It was not appropriate to triangulate informants’ views on sero-surveillance, as such data were treated as informants’ opinions and perceptions.
RESULTS
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
7
BMJ Open
The results, which represent the views and perceptions of informants, are presented by themes and by country. The labels describe the themes that emerged, in the context of informants’ awareness of the aims and objectives of the study.
Value of UAT UK The programme is largely seen by respondents as indispensable in providing an unbiased
rp Fo
measure of HIV prevalence and risk factors in the UK population. The most important outcome from the surveys has been information on undiagnosed HIV infection which, respondents suggested, has led to changes in policy on named testing; for example, HIV testing in pregnant women has changed from opt in to opt out, and sexual health clinics have moved from targeted to universal voluntary testing. The neonatal dried blood spot survey is
ee
used as a check on the performance of local antenatal screening. The public health rationale for the programme has not changed since 1990, and respondents cited the continuing high
rr
infection rate, particularly among men who have sex with men, as justification:
ev
"There was far more feeling of a public health emergency (then) than there is today, although you could argue … that it's actually much worse today because of where we see the disease." (UK4)
w
ie
Several respondents suggested that a review of UAT would be timely, including consideration of extending the surveys to other health care settings and other diseases. UK6
on
discussed how advances in HIV testing, including self-testing, might alter the questions asked by the programme. Some suggested that if named testing was normalised, there would be less need for anonymous testing.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 53
A particular theme was whether the British public has a duty to participate in public health monitoring, possibly as part-and-parcel of receiving the benefits of a national health service (NHS): "There is a very good case for really exploring with the public, and then pushing very hard to get public agreement, that in general terms that sort of tissue, extra bits of tissue or blood, should be available for use." (UK4)
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
8
Page 9 of 53
USA According to one respondent, the surveillance programme was valuable in terms of planning, resource allocation, advocacy and interventions to prevent transmission. The heel prick survey provided accurate national data on prevalence in childbearing women and was seen as extremely valuable, with nothing since that has given as accurate a picture of national prevalence (USA 2).
rp Fo
However, given the questionable ethics of UAT (see later), as well as the resultant lack of good quality prevalence data, USA3 argued that enough became known about prevalence to target the most vulnerable groups for health promotion. “And how much more detail do you want from that, so what would you do differently
ee
if you knew that in the next level of detail.” (USA3) Netherlands
rr
Unlinked anonymous surveys have been used in the Netherlands to measure HIV prevalence
ev
among high risk populations, but with informed consent (verbal or written). The populations surveyed by UAT have changed over time. Injecting Drug Users (IDUs) have been surveyed
ie
bi-annually since 1994 in different cities in the Netherlands, and are recruited via methadone care; sex workers and migrant groups have also been subject to UAT surveys.
w
There are two reasons, indicated by respondents, why UAT in Netherlands has been of
on
limited value. First, the uptake of voluntary testing is generally high in the Netherlands, meaning the quality of that data are generally good. And second, UAT has only been deemed
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
ethically permissible in the Netherlands with informed consent, making large-scale UAT surveys impracticable, so restricting UAT to specific population groups and/or the research context.
Norway The unknown, undiagnosed segment of the population is known as the ‘dark figure(s)’ in Norwegian, although neither the ‘dark figure’ nor the number of cases of HIV has been perceived as significant issues in Norway. Uptake of voluntary testing for HIV was described as very high in intravenous drug users and antenatal clinic users, meaning that UAT has never
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
9
BMJ Open
been perceived as to be of value epidemiologically. The system of surveillance for HIV in Norway instead covers all diagnosed cases (i.e. not undiagnosed cases as with UAT), and was reported as being simple, inexpensive and capable of providing a representative and timely overview that guides prevention, while respecting confidentiality and adhering to ethical principles.
Quality and nature of data UK
rp Fo
The epidemiological data yielded from unlinked anonymous testing are considered by respondents to be of unrivalled quality internationally. It was suggested, however, that the apparent high number of undiagnosed HIV infections from UAT surveys might be inflated by double-counting:
rr
ee
"You can quite understand that if you've got gonorrhoea and you've been told by your doctor not to have unsafe sex, you might well decide to go to a different clinic to have your gonorrhoea treated." (UK1)
ie
ev
Others defended the statistics and argued that they were explained by a resurgence of high risk behaviour, particularly among gay men. UK1 argued that people who refuse a named test
w
and who are subsequently found to be HIV positive through UAT are likely to be people who know that they are positive. USA
ly
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 53
There was little comment from USA interviewees about the quality of UAT data, since UAT surveys stopped in 1996. However, a theme that emerged in relation to UAT data was around the difference between research and surveillance.
The fact that there is no ‘right’ to health care in the USA was felt to have an effect on people’s views on surveillance – in other words, the ‘disconnect’ between health care provision (not available to all) and surveillance (done to all) (USA 1). The differences between surveillance, research and practice, with perhaps different ethical standards required for each, emerged as a theme from the interviews (USA 4).
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
10
Page 11 of 53
One point of view was that surveillance (because it is named and part of routine care) is part of clinical care rather than purely research (USA 2). However, while some of the surveillance undertakings at state level in the USA were considered to be practice, at the level of CDC they were treated as research (USA 1). The same respondent commented that cancer sufferers advocated in favour of surveillance for the purposes of research on the disease, since “they were willing to trade privacy and consent for a good which was surveillance that they saw as serving their interests”.
rp Fo
Norway and the Netherlands
There were no comments from Norway respondents about the quality of UAT data since such surveys are not carried out in Norway, in part because of the good quality of data on HIV and AIDS from other sources.
ee
Two respondents from the Netherlands judged that large-scale UAT programmes were not
rr
necessary because named HIV testing is widespread. The quality of information obtained from named testing was felt to be robust and could be used to model estimates of HIV
ev
prevalence in the population. Coverage of pregnant women is very high and this group is considered a valid proxy for the heterosexual population. Named HIV testing among users of
ie
STI clinics was also presented as being a valid method for estimating prevalence. It was argued that, as long as the uptake of named testing by pregnant women and STI clinic users
w
remains high (above 90%) then a national programme of UAT was not thought to be necessary. However, notwithstanding the above point it was acknowledged that current
on
methods of surveillance do not allow undiagnosed HIV infection to be measured accurately.
Ethical issues
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
UK Respondents felt that, on balance, no harm was caused by the programme. A recurrent argument for this was that, in settings where anonymous testing takes place, people are offered a named HIV test:
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
11
BMJ Open
"I think the ethics are not in fact, are not a problem because everybody is now very strongly advised to have a test and therefore the residual rump who don't have a test, it's important to know what their rate of HIV positivity is." (UK1) Some respondents, however, said there would be ethical concerns if people in these settings were not routinely offered a named test. Several respondents considered the ethics of UAT to be relatively unimportant, for instance in the context of more significant health-related problems in Africa:
rp Fo
"There are these enormous ethical, real ethical issues, and to me, this one … I just can't understand why one should put so much emphasis on it." (UK9) UK11 suggested that the question of ownership of the sample was relevant, but that this could be argued both ways:
rr
ee
"Is it yours or is it somebody else's? ... If it's waste that's going to be chucked away then why shouldn't people use it if they find it useful? On the other hand, 'I gave
ev
permission for this test for my own benefit and I didn't give you permission to do things to help others.'" (UK11)
ie
There was universal agreement that changing to a system of informed consent would damage
w
the programme. There was a fear that introducing information about UAT into a clinical consultation could confuse patients. Patients might think they were getting a named test:
on
"You would lose someone's confidence and trust because they might not understand
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 12 of 53
the words or the reason behind why we're doing it, and they would think that you could get back to them and you did know." (UK12)
Other respondents argued that it would reduce time for discussing important clinical matters: "When the midwife turns up, does she pull out a leaflet and say by the way, I'm putting this beside you here? … it would be a perversion of the consultation to divert it into discussing this ..." (UK5)
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
12
Page 13 of 53
Further concerns were that a system of informed consent would increase refusals and lead to bias in the surveys, and also the time and cost of gaining consent from patients:
"You know we've got ten minute appointments. It could take you five minutes, half of your time just to explain what this is all about. So doing it in a consultation is not realistic." (UK3) The respondent suggested that a system of clinic self-registration, using computers, could be used to explain anonymous testing.
rp Fo
All the respondents thought that prior notification was the right theoretical approach. However, some were concerned that, in practice, it might not be working: "It's a reasonable thing in practical terms provided people are informed. So … if people don't know about their option for opting out and their option for finding out
ee
then the system hasn't been working." (UK11)
rr
Some suggested that the approach needed to be more "active" with information displayed visibly on clinic walls or in leaflets given out to each patient. One clinician, however,
ev
expressed his discomfort with his role in the programme. He described filling out the form for UAT blood testing in front of the patient, without explaining what he was doing, and
ie
sometimes after the patient had declined a named HIV test:
w
"You're doing something that sort of seems slightly underhand…you're not explaining it to the patient. Yeah, I'm not comfortable with that, I think one wants to be, to
on
communicate well and to be as open as possible with patients and that's slightly in conflict with that." (UK3)
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
USA
Consent was not deemed necessary initially in the USA for participation in UAT as patients had consented to the blood being collected in the first place, and this was seen as adequate. USA1 suggested that surveillance is a public good, meaning there are strong reasons for making it a universal obligation on citizens – and that notification that surveillance is occurring is different from asking for consent.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
13
BMJ Open
The blood spot survey continued until the publication of the first trials showing that vertical transmission of HIV could be dramatically reduced by treating mothers. When it became clear that infected women could be offered treatment, and therefore should know their HIV status, “it became impossible for the US Centre for Communicable Disease Control and Prevention (CDC) to continue that programme and it was dropped in the early 1990s along with all the other ones” (USA3). It had become difficult to reconcile public health benefits of UAT with individual rights. If CDC had not voluntarily withdrawn funding from UAT, the same respondent indicated that it was likely that Congress would have mandated cessation, as
rp Fo
there was a strong coalition in favour of this. A related legal issue was brought by USA4, in that public health surveillance appears to go against “fundamental legal principles” in the USA, and constitutional scholars are shocked when they hear what is done.
ee
‘Once you begin to argue that the government can take your personally identifiable
rr
information for any reason it finds useful without your consent you've opened up and blown a hole in much of the fundamental legal principles in the country.’ (USA4)
ev
The Tuskegee episode (a study between 1932 and 1972 in Alabama in which black syphilitic
ie
Americans were not given treatment although thought they were) was mentioned by USA1 as a reason for USA citizens, particularly Afro-Americans, to be suspicious of surveillance
w
practices. In the context of opposition to named case reporting and a consequent lack of trust in surveillance, USA3 argued that there are probably cultural dimensions to issues such as
on
consent, and the acceptability of particular interventions may vary from country to country.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 53
“It [UAT] became more of a civil rights issue than a gay specific one. It was, the government knows something that you don’t and the attitude in the States, at least at the time, towards the government was much less trusting than in this country [i.e. UK]. People did not assume and still don’t assume that the government necessarily is looking out for your well being. And I think there’s a much more paranoid attitude towards the government than there is in this country.” (USA3)
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
14
Page 15 of 53
Related to these cultural dimensions, a particular ethical concern was that the US (and, in particular, CDC) supports UAT in developing countries, with the inherent contradiction that it has been rejected in the USA. ‘It seems to me the CDC can’t at the one time say it’s unethical to do it in the US and say it’s ethical to do it abroad. Different public agencies may come to different decisions, but the same agency it seems to me should be bound by some rule of consistency.’ (USA1)
rp Fo
USA2 felt very strongly that arrangements for UAT in developing countries are not ethical where pregnant women do not have the opportunity to have a named test, counselling and treatment.
Netherlands
ee
The respondents felt that more extensive UAT surveys would produce more accurate
rr
information for public health surveillance. However, inter-linked practical and ethical doubts about UAT were presented. One concern was that UAT might deter high risk groups from
ev
seeking testing and treatment. Another concern was that, if UAT focused only on high risk groups, such groups would risk being stigmatised. It was suggested that reporting of findings
ie
from UAT surveys of high risk groups would need to be carried out sensitively.
w
A particular issue was consent and maintaining trust with patients. One interviewee expressed that staff in HIV treatment centres and STI clinics demand a more stringent consent process
on
than required by the research ethics committees in the Netherlands in order to promote trust among patients.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Norway
A UAT system was never thought relevant, necessary or appropriate for Norway. In Norway it is not legal to test or analyse blood or biological material for infectious diseases for nondiagnostic purposes without consent unless the purpose is surveillance of an ‘epidemic’. One respondent felt this displayed a ‘moral grey zones’, as the law hinges on the definition of ‘epidemic’ [Norway 1].
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
15
BMJ Open
There is very little concern or awareness in Norway about ethical issues around UAT. However, there is a general willingness to provide information for public health purposes in Norway, and people are used to being surveyed. It was reported that many Norwegians have given blood to the Biobank, and give broad consent for their tissue/blood to be used for a wide range of research. A respondent indicated that many are willing to waive their consent in Norway because ‘there is significant trust’ in the health care system. However, another respondent expressed that there was a possible danger that UAT blood could be used to identify someone via the Home Office DNA database, and that researchers might be forced to
rp Fo
share information under a court order (the respondent mentioned that the murderer of the Swedish politician Anna Lindh in 2003 was convicted on the basis of a match with DNA held in the national biobank).
One respondent suggested that informed consent could be different in different contexts. In a
ee
country like Norway, with low prevalence and easy access to health care, the current surveillance system (non-UAT) was described as acceptable and functioning well. However,
rr
in a country in Africa (where this respondent was carrying out research), with a high rate of HIV positive pregnant women and a low rate of testing, routine testing with opt-out consent procedures would be morally acceptable.
w
ie
DISCUSSION
ev
Since its inception, unlinked anonymous testing (UAT) has been a subject of
on
epidemiological, ethical and policy debate.15 16 17 In this paper we have presented the results of qualitative interviews with key informants in four countries pragmatically selected because they represent a range of approaches to UAT. 18 A total of 23 respondents were interviewed,
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 53
differentially distributed between countries: 12 in the UK, four in the USA, four in Norway and three in the Netherlands. There were limitations to the research: more key informants were interviewed in some countries than others; some interviews were face-to-face whereas others were by telephone; and two of the interviews relied on notes rather than recordings for later analysis.18 19 20 It is possible that such aspects introduced bias, but we believe that the balanced pragmatic approach taken to address the research questions was appropriate for this relatively small
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
16
Page 17 of 53
qualitative study. Interesting themes emerged in the analysis, and we were able to triangulate more fact-based comments with documentary sources in the literature.
First a summary of key findings is presented (in the context of historical development of UAT in different countries, as well as what is already known of the subject), followed by a discussion of how the findings from this paper fit into broader contemporary debates about public health ethics and policy.
rp Fo
Summary of key findings from interviews (in the context of what is currently known) In the United Kingdom (UK), a balance of epidemiological and ethical considerations in the late 1980s led to the implementation of the ‘prior notification’ approach to UAT; this
ee
involves leaflets and posters informing clients that UAT is occurring, allowing opting out but not necessitating the gaining of informed consent.7 10 Respondents stressed that the rationale
rr
remains apt today, and that this approach provides data unrivalled in terms of epidemiological quality,9 21 22 and which is of use for health service planning.23 Ethical concerns were
ev
generally felt to be minimal today since clients are (also) usually offered voluntary (named) testing in settings where UAT is occurring. However, there was mention of concern about the
ie
lack of full explanation, whether people do actually know UAT is happening, and ‘ownership’ of blood.24 It was indicated that an approach involving informed consent would
w
be cumbersome, time-consuming and epidemiologically damaging.25
on
In the United States (USA), UAT of blood for HIV was extensively used in the 1980s and early 1990s,26 but was stopped in the late 1990s as evidence emerged that anti-viral therapy in
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
pregnant women could help reduce the vertical transmission of HIV from pregnant mother to child – named testing needed to be maximised so that HIV-positive pregnant women could be identified and offered advice and therapy.27 The interviews indicated that data from UAT were useful epidemiologically but, since UAT has stopped, there has been sufficient other data to plan services accordingly. Other issues raised include a general suspicion around epidemiological surveillance post-Tuskegee, cultural dimensions of UAT with respect to minority groups, a lack of trust in the Government, and the related disconnect between surveillance (done to all) and health care provision (variable, and not done to all).6 11 There was also comment about the apparent contradiction in the contemporary support of the US
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
17
BMJ Open
Centre for Communicable Disease Control and Prevention (CDC) for UAT in low-income countries.28
Interviews in the Netherlands showed that named (voluntary) testing, which has generally been widely taken up, has provided the main vehicle for surveillance of HIV.4 There is acknowledgement that large-scale UAT would provide additional epidemiological data, but this has been balanced against concerns of this leading to diminished voluntary testing – especially in at risk groups – through loss of trust in the healthcare system.29 Smaller UAT
rp Fo
surveys have been intermittently undertaken in such groups (e.g. intravenous drug users, sex workers, migrants) as research studies, but have required individual consent (initially oral, now written).
Interviews in Norway revealed that UAT was never felt to be necessary epidemiologically, as
ee
HIV has not been a significant issue and the uptake of voluntary testing has been high.13 Surveillance has thus tended to focus on the monitoring of the diagnosed population
rr
(anonymised diagnosed cases), which has been supported by the gay community; the undiagnosed population is interestingly referred to as ‘dark figure(s)’. In Norway it is not
ev
legal to test without consent unless there is an epidemic, although respondents reported few ethical concerns since people generally have trust in the health care system, and are willing to
ie
provide information for public health purposes (many give blood to the Biobank for a range of research purposes).30
w
The findings of this research on the ethics of UAT of blood need consideration in the context
on
of two broader contemporary debates within medical and public health ethics. The first debate centres on the boundaries of informed consent in health research, particularly in
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 53
relation to the use of human tissue. The second debate focuses on the balance between individuals’ rights and the appropriate role of government3 in infringing on those rights for broader public health interests or concerns. These debates are inter-connected, although will be presented in turn.
Informed consent and ethics: a changing debate
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
18
Page 19 of 53
Respect for autonomy is the bioethical principle that is most likely to be jeopardised by a programme of unlinked anonymous blood testing, since informed consent is not usually deemed necessary for an individual’s blood to be included; this has tended to be the basis for opposition to UAT.5 31 The findings from our research show, however, that such positions differ between countries, and have also changed with time.
Debates in the late 1980s and early 1990s demonstrated the opposing views of those who were concerned about patient consent,8 26 and those who felt that the priority was to control
rp Fo
what was seen as an epidemic.15 UAT programmes did go ahead in the USA and UK, and other high income countries: there was no requirement for consent in the USA, whereas in the UK an approach of ‘prior notification’ was adopted through leaflets and posters in clinics – consent is presumed but individuals have the opportunity to opt out.
ee
Over the past two decades informed consent has emerged as ‘the ethical touchstone of medical research, and is now enshrined in practice and a range of international guidelines as
rr
one of the main preconditions of medical research’.25 32 Some commentators argue that even anonymised tissue should not be used without consent. Trouet, for example, believes that the
ev
use of such material is in contravention of Article 22 of the Convention on Human Rights and Biomedicine. This states that ‘when in the course of an intervention any part of a human body
ie
is removed, it may be stored and used for a purpose other than that for which it was removed, only if this is done in conformity with appropriate information and consent procedures’, and
w
in accordance with Article 8 of the European Convention on Human Rights which protects the right to respect for private life.33
on
Our research indicates that, rather than the universality of human rights discourse, a form of
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
weak moral relativism operates in policy-making and practice around UAT.34 In the USA, the change in social context and debates about individual rights, along with evidence of reduction in mother-to-child vertical transmission, resulted in the abandonment of UAT in 1996.12 The fact that the USA continued to support UAT in international research overseas illustrates further an attitude of moral relativism, although it could be argued that differential health needs justified different approaches. Interviews in the UK showed that the prior notification approach that remains in use today still holds strong support – based on perception that there is minimal harm to the individual,
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
19
BMJ Open
and the value of high quality, unbiased epidemiological data on HIV (held to be the most robust and accurate worldwide).35 Respondents felt that any change to UAT requiring informed consent would be cumbersome, time-consuming, and costly. According to the Nuffield Council on Bioethics it may also be ethically unnecessary: ‘... it is acceptable to collect and use anonymised data for assessing and predicting trends in infectious disease without consent, as long as any invasion of privacy is reduced as far as possible.’.36 This position is not, however, cross-culturally held. In the Netherlands, UAT has not been felt
rp Fo
to be acceptable without informed consent, initially verbal and more recently written. The societal emphasis on individual autonomy, including in the health research environment, has been the paramount consideration.5 Northern European and Scandinavian countries are often held up as having a strong community ethos, with an expectation on individual involvement in endeavours for the public good.37 This was not reflected in UAT practice in the
ee
Netherlands, although was indicated to be important in Norway, albeit in a setting where UAT has never been taken up.
rr
These findings indicate the importance of social and political context in relation to medico-
ev
ethical debates, and how attitudes and policies can change in time within any given jurisdiction.38 Differences exist between the moral evaluations and resultant UAT approaches
ie
between countries such as England and the Netherlands. Within the USA, in contrast, the practice of UAT has been re-evaluated with time. In the late 1980s and early 1990s the feared
w
AIDS epidemic was seen as a public health emergency, there was no treatment available, transmission from mother to unborn child was not preventable, and the stigma and
on
discrimination associated with the infection were more overt than today. As these factors shifted, so did the policy on UAT in the USA. The situation in low income countries today
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 53
has created a perception of the acceptability of UAT akin to that in western countries two decades ago.
Individual good and community (public) good There has been criticism in recent times of government for its ‘nanny state’ approach to interventions in, for example, the areas of smoking, alcohol and diet. More recently the notion of ‘stewardship’ in public health rejects these negative associations,39 and advocates
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
20
Page 21 of 53
that government should have a role to play in protecting the health of its citizens, preventing them from harming themselves or others, and promoting healthy behaviour.[4] In this view, public health interventions should not be thought of as unwanted restrictions imposed by a ‘nanny’ on powerless children, but as actions that favour both individuals and society more widely as prescribed by a ‘steward’ administering and protecting citizens’ interests. The Nuffield Council on Bioethics, for example, argues that the state should act as steward both to individuals and to the population as a whole, but that individuals’ personal choices should also be taken into account.36
rp Fo
The subject of UAT sits squarely within the debate about public health stewardship for social good. Recently, commentators have argued that patients of a publicly supported health care system like the NHS have a moral duty to participate in scientific research,40 41 although most do not go so far as to articulate that such participation should be mandatory in practice.42
ee
Nevertheless, UAT could be understood as something that individuals should feel they have a moral obligation to participate in, regardless of ethical concerns about individual rights and
rr
personally held views. Our own survey of over 400 users of two genitor-urinary medicine clinics (a parallel part of this research project) has found that a large majority of respondents
ev
(89%) agree that people should have a responsibility to take part in research studies, and the same proportion would agree to the use of their blood in unlinked anonymous seroprevalence
ie
testing. However, when asked about the question of consent, 74% indicated that they should be asked to consent before their blood is used for public health surveillance.14
w
Notwithstanding this important latter point, such support for use of blood samples for prevalence surveys or for future research has not, however, been demonstrated in an earlier study in the UK43 or more recently in Egypt.44
ly
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
The way forwards perhaps is to ensure that patients – and the public more widely – are informed of the existence, purposes and value of disease surveillance arrangements, as well as the importance of contributing to future health research, while investigating arrangements for gaining consent which are not unnecessarily cumbersome or likely to create bias amongst those who agree that their blood be used.45 The issue of consent in seroprevalence surveillance, such as UAT, remains an important instrument to gauge the moral direction of public health policies more broadly.46 Debate continues around how public health programmes can embrace a stewardship relationship,
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
21
BMJ Open
rather than a nanny paternalistic one, while respecting autonomy and advocating social responsibilities. Further examination of why different countries have adopted different positions on seroprevalence surveillance will help illuminate our understanding of how social morality is embedded in public health policies and practice.
Acknowledgments The authors would like to thank the participants in the study. We also acknowledge the
rp Fo
valuable advice given by Vivian Hope, Will Nutland, Cicely Marston, and Graham Hart.
Conflicts of interest
AK is also Director of Public Health Strategy and Medical Director of the Health Protection Agency. The views expressed in this paper are those of the authors, and are not intended to
ee
represent the views of the Health Protection Agency.
Funding
rr
This research was supported by the Economic and Social Research Council (ESRC Grant Reference: RES-000-22-2096).
ie
Ethical approval
ev
The study was approved by the South West Research Ethics Committee on August 16th 2007
w
(reference 07/H0206/46), and by the London School of Hygiene & Tropical Medicine ethics committee on 3rd September 2007 (reference 5174).
on
Notes [1] The England and Wales programme is currently managed by the Health Protection
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 53
Agency (HPA) and the Institute of Child Health, University of London. The programme was previously managed by the Communicable Disease Surveillance Centre, part of the Public Health Laboratory Service, which was subsumed into the Health Protection Agency in 2005. [2]
Health Protection Scotland (HPS) co-ordinates data collection in Scotland.
[3]
Those chosen had particular involvement or interest in UAT and ethics, but further
details cannot be provided to maintain their confidentiality. [4]
By government, the meaning here is health and public health policy makers within
government.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
22
Page 23 of 53
References
1
Nicoll A, Gill ON, Peckham C et al. The public health applications of unlinked
anonymous seroprevalence monitoring of HIV in the United Kingdom. International Journal of Epidemiology 2000;29:1-10. 2
Department of Health. Prevalence of HIV in England and Wales in 1997: Annual
Report of the Unlinked Anonymous Prevalence Monitoring Programme. London: Department
rp Fo
of Health, 1998. 3
UK Collaborative Group for HIV and STI Surveillance. Testing Times: HIV and
Other Sexually Transmitted Infections in the United Kingdom. London: Health Protection Agency, 2007. 4
ee
Bayer R, Lumey LH, Wan L. The American, British and Dutch responses to unlinked
anonymous HIV seroprevalence studies: an international comparison. AIDS 1990;4(4):283-
rr
290. 5
ev
Kopelman LM. Informed consent and anonymous tissue samples: the case of HIV
seroprevalence studies. Journal of Medicine and Philosophy 1994;19:525-52. 6
ie
Fairchild AL, Bayer R. Ethics and the conduct of public health surveillance. Science
2004;303:631-632. 7
w
Gill ON., Adler MW, Day NE. Monitoring the prevalence of HIV: foundations for a
on
Programme of unlinked anonymous testing in England and Wales. British Medical Journal 1989:299:1295-7. 8
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Berridge V. AIDS in the UK: the Making of Policy 1981-1994. Oxford: Oxford
University Press, 1996. 9
Health Protection Agency. Overall HIV prevalence. Health Protection Agency 2008.
http://www.hpa.org.uk/web/HPAweb&Page&HPAwebAutoListName/Page/1201094588821. 10
Gillon R. Testing for HIV without permission. British Medical Journal 1987;294:821-
823.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
23
BMJ Open
11
Fairchild AL, Bayer R, Colgrove J. Searching Eyes: Privacy, the State, and Disease
Surveillance in America. Berkeley: University of California Press, 2007. 12
Stoto MA. Almario DA, McCormick MC. Reducing the Odds: Preventing Perinatal Transmission of HIV in the United States. Washington: National Academy Press, 1999.
13
Aavitsland P, Nilsen Ø, Lystad A. Anonymous reporting of HIV infection: an
evaluation of the HIV/AIDS surveillance system in Norway 1983-2000. European Journal of Epidemiology 2001;17:479-89. 14
rp Fo
Datta J, Kessel AS, Wellings K, et al. The views of genitourinary medicine (GUM)
clinic users on unlinked anonymous testing of blood for HIV: evidence from a pilot study of clinics in two English cities. Journal of Medical Ethics 2011 doi:10.1136/jme.2011.042705. 15
Doll R. A proposal for doing prevalence studies of AIDS. British Medical Journal
1987;294:443-444.
rr
16
ee
Avins AL, Lo B. To tell or not to tell: the ethical dilemmas of HIV test notification in
epidemiologic research. American Journal of Public Health 1989;79/11:1544-7. 17
ev
Datta J, Kessel AS. Unlinked anonymous testing for blood purposes: an ethical
dilemma? In: Peckham S, Hann A, eds. Public Health Ethics and Practice. Bristol: Policy Press, 2010:101-116.
w
ie
18
Bowling A. Research Methods in Health. Buckingham: Open University Press, 2000.
19
Yin RK. Case study research: design and methods. London: Sage, 1994.
20
Green J, Thorogood N. Qualitative methods for health research. London: Sage, 2004.
21
Duong T, Ades AE, Rogers P, et al. Non-participation bias in unlinked anonymous
ly
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 24 of 53
HIV-prevalence surveys in England and Wales. Epidemiology and Infection 1999;122:26772. 22
Pulver WP, Glebatis D, Wade N, et al. Trends from an HIV seroprevalence study
among childbearing women in New York State from 1988 through 2000: a valuable epidemiologic tool. Archives of Pediatric and Adolescent Medicine 2004;158(5):443-8.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
24
Page 25 of 53
23
Kessel AS, Watts CJ. Usefulness of information from the Unlinked Anonymous
Prevalence Monitoring Programme for HIV in England and Wales: survey of planners of HIV/AIDS services. International Journal of STD and AIDS 1999;10:808-11. 24
Pfeffer N, Laws S. 'It's only a blood test': what people know and think about
venepuncture and blood. Social Science & Medicine 2006;62:3011-3023. 25
Cassell J, Young A. Why we should not seek individual informed consent for
rp Fo
participation in health services research. Journal of Medical Ethics 2002;23:313-317. 26
Bayer R. The ethics of blinded HIV surveillance testing. American Journal of Public
Health 1993;83/4:496-7. 27
de Zulueta P. The ethics of anonymised HIV testing of pregnant women: a
ee
reappraisal. Journal of Medical Ethics 2000;26:16-21. 28
Rennie S, Norris Turner A, Mupenda B, et al. Conducting unlinked anonymous HIV
rr
surveillance in developing countries: ethical, epidemiological, and public health concerns. PLoS Medicine 2009;6(1):e1000004. 29
ev
Postema EJ, Willems PWJM, de Ridder MAJ, et al. Comparison of patients refusing
ie
with patients accepting unlinked anonymous HIV testing in an outpatient STD department in the Netherlands. International Journal of STD and AIDS 1997;8:368-72. 30
w
Skolbekken J-A, Ursin LO, Solberg B, et al. Not worth the paper it’s written on?
on
Informed consent and Biobank research in a Norwegian context. Critical Public Health 2005;15(4):335-347. 31
Beauchamp TL and Childress JF. Principles of Biomedical Ethics. New York: Oxford
University Press, 1994. 32
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Singleton P, Wadsworth M. Consent for the use of personal medical data in research.
British Medical Journal 2006;333:255-8. 33
Trouet C. New European guidelines for the use of stored human biological materials
in biomedical research. Journal of Medical Ethics 2004;30:99-103. 34
Rachels J. The Elements of Moral Philosophy. New York: McGraw-Hill, 1993.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
25
BMJ Open
35
Beuhler JW, Petersen LR, Ward JW, et al. Defending HIV seroprevalence surveys.
American Journal of Public Health 1994;84:319-320. 36
Nuffield Council on Bioethics. Public Health: Ethical Issues. London: Nuffield
Council on Bioethics, 2007 37
Cappelen AW, Norheim OF. Responsibility in health care: a liberal egalitarian
approach. Journal of Medical Ethics 2005;31:476-80. 38
rp Fo
Mariner WK. Mission creep: public health surveillance and medical privacy. Boston
University Law Review 2007;87/2:347-95. 39
Jochelson K. Nanny or Steward? The Role of Government in Public Health. King's
ee
Fund Working Paper. London: King's Fund 2005. 40
Harris J. Scientific research is a moral duty. Journal of Medical Ethics
2005;31(4):242-248.
rr
41
Evans HM. Do patients have duties? Journal of Medical Ethics 2007;33(12):689-694.
42
Shapshay S, Pimple KD. Participation in biomedical research is an imperfect moral
ie
ev
duty: a response to John Harris. Journal of Medical Ethics 2007;33:414-417. 43
w
Kessel AS, Watts C, Weiss HA. Bad blood? Survey of public's views on unlinked
anonymous testing of blood for HIV and other diseases. British Medical Journal 2000;320:
on
90-91. 44
Abou-Zeid A, Silverman H, Shehata M, et al. Collection, storage and use of blood
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 26 of 53
samples for future research: views of Egyptian patients expressed in a cross-sectional survey. Journal of Medical Ethics 2010;36:539-547. 45
Nicholls SG. Knowledge or understanding? Informed choice in the context of
newborn bloodspot screening. Public Health Ethics 2010;3:128-136. 46
Haimes E. What can the social sciences contribute to the study of ethics? Theoretical,
empirical and substantive considerations. Bioethics 2002;16/2:89-113.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
26
Page 27 of 53
Revised manuscript, submitted to BMJ 15 Aug 2012 Marked
Original paper
The ethics of unlinked anonymous testing of blood: views from in-depth interviews with key informants in four countries Unlinked anonymous testing of blood (running title)
rp Fo
Anthony S Kessel (corresponding author and guarantor of study), Honorary Professor, London School of Hygiene & Tropical Medicine, Faculty of Public Health and Policy, 15-17 Tavistock Place, London WC1H 9SH (Tel: 020 7759 2715; email:
[email protected])
Jessica Datta, Research Fellow, London School of Hygiene & Tropical Medicine
ee
Kaye Wellings, Professor, London School of Hygiene & Tropical Medicine
Sarah Perman, Health Protection Agency
Key words
6608437 words
ly
on
Word count
w
Unlinked, anonymous, blood, ethics, public health, policy, HIV
ie
ev
rr
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
Abstract In this study we explore the ethical issues around unlinked anonymous testing (UAT) of blood, a method of sero-prevalence surveillance for infectious diseases. Our study focused on UAT for HIV, although UAT can be used for other infectious diseases.
Objectives
rp Fo
The objectives of the research were to gain a better understanding of the views of key informants in countries adopting different UAT testing strategies, and to use the findings of the research to inform health policy
Design
Qualitative study using in-depth interviews and ethical analysis.
Setting
ee
Four countries using different strategies around unlinked anonymous testing of blood for HIV (United Kingdom, United States, Netherlands and Norway).
Twenty-three key informants in the four countries.
ie
Results
ev
Participants
rr
Participants from the four countries have different views on UAT of blood, and the approaches and
w
policies on UAT adopted by different countries have been historically and culturally determined. We use our findings to explore the relationship between public health policy and ethics, framing our discussion in relation to two important contemporary debates: informed consent for participation in
on
medical and public health research; and the balance between individual good and the public good.
Conclusions
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 28 of 53
Qualitative research and ethical analysis of unlinked anonymous testing of blood in different countries has yielded important findings for consideration by policy makers. The policy of unlinked anonymous
testing of blood for HIV and other diseases in the UK needs re-consideration in light of these findings.
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 29 of 53
The ethics of unlinked anonymous testing of blood: views from in-depth interviews with key informants in four countries
In this paper we explore unlinked anonymous testing (UAT) of blood for HIV and other diseases, and examine the ethical issues associated with this method of sero-prevalence
rp Fo
surveillance for infectious diseases. We report original results from a study funded by the UK Economic and Social Research Council (ESRC) – qualitative interviews with stakeholders in four countries adopting different UAT policies – and use the subject matter, and our findings, to explore the relationship between public health policy, context and ethics.
BACKGROUND
rr
ee
Unlike voluntary testing for clinical purposes, unlinked anonymous testing (UAT) of blood is carried out in order to provide information about the prevalence of particular infectious diseases in a population. The use of UAT began in the late 1980s for HIV, and remains most
ev
commonly associated with this infection, but can be applied to other infectious diseases such as hepatitis C.1 In this paper we use the term UAT synonymously with sero-prevalence surveys, or sero-prevalence surveillance.
w
ie
The principles behind UAT can be explained with reference to the approach used in the UK for HIV. Leftover blood from samples taken for diagnostic or routine screening purposes is made anonymous, irreversibly unlinked from the source, and then tested for HIV.2 3 This
on
means that individuals whose blood tests positive for HIV cannot be contacted or informed. The data derived from UAT are then used for monitoring purposes and also to inform the planning and evaluation of health services and health promotion activities.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Historically, UAT has developed in different ways in different settings, yet exploration of the moral dimensions of such developments has been relatively limited.4 5 6 Further, dominant attention given to individual rights and autonomy – in the health and healthcare context – make it an appropriate time to reassess moral considerations around UAT. A brief summary
3
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
of developments in UAT in the four countries used in our study is provided as a necessary introduction to the research project.
History of UAT in different countries
A national sero-prevalence surveillance programme using UAT to estimate the prevalence of
rp Fo
HIV has been in place in the UK since 1990.111 The programme was first developed in
Formatted: Endnote Reference
7
response to concerns about the spread of HIV during the 1980s and, despite early opposition,
Formatted: Endnote Reference
the methodology eventually gained political support.8 The programme was funded initially as a pilot project by the Medical Research Council, and has since been funded by the Department of Health. [1] [2] The programme currently uses data collected from three
ee
ongoing surveys (newborn babies, [maternal antibodies], users of genitourinary medicine (GUM) clinics, and intravenous drug users [saliva]) to measure the proportion of individuals within the general population, and within subgroups of the population, infected with HIV.
rr
Since the programme’s inception, around ten million blood samples have been tested in this way.9
ev
A system of ‘prior notification’ operates in locations where testing goes on, involving the display of leaflets or posters explaining that blood may be tested anonymously. Consent to
ie
participation is implicit rather than explicit; provision is made to opt out of testing, although this naturally depends on awareness.999 10
w
Formatted: Endnote Reference Formatted: Endnote Reference
In the USA, UAT was extensively used in the late 1980s and early 1990s using discarded
on
blood samples. The largest survey was national and used newborn heel prick tests. Other UAT surveys focused on population groups, for instance a survey of young people
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 30 of 53
volunteering for job corps, although these were regional or local. In contrast to the UK the prior notification system, with the opportunity to opt out, was not used. UAT essentially stopped in the USA in 1996, as evidence emerged in relation to reducing vertical
transmission of the HIV virus from a pregnant mother to the foetus through use of anti-viral medicine.11 The interplay at the time of clinical, social and political factors was complex.12
The main method for HIV surveillance in the Netherlands is named voluntary testing rather than UAT (these are ‘opt-out’ programmes, such as the antenatal testing programme and the 4
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 31 of 53
testing of users of STI clinics, as well as the extensive ‘opt-in’ testing carried out by general practitioners). The UAT that does occur in the Netherlands is carried out using blood samples from intravenous drug users (IDUs) and saliva samples from other groups. Unlike the UK or USA, informed consent is required from those surveyed by UAT, either verbal or, more recently, written consent. From 2002 to 2006, other high risk groups haven been included in UAT research surveys in the Netherlands (e.g. migrants from the Cape Verde islands, Surinam and the Netherlands Antilles, as well as commercial sex workers), because of
rp Fo
concerns of a higher proportion of undiagnosed HIV in such groups, who were also felt to be accessing health services less than others.
In Norway, public health efforts have focused on diagnosing all those infected, as it was thought (certainly in earlier stages) that those who were diagnosed would be less likely than
ee
those undiagnosed to transmit infection to others. Mandatory named AIDS reporting was introduced in Norway in 1983 and mandatory anonymous HIV case reporting in 1986, both including cases diagnosed before those dates.13 The Norwegian approach has therefore
rr
centred on individual testing and clinician reporting, and UAT has not been deemed epidemiologically necessary and has never been undertaken.
AIMS AND OBJECTIVES OF THE RESEARCH STUDY
w
ie
ev The ESRC-funded project had three components: a questionnaire survey of 471 patients attending two inner-urban genito-urinary medicine (GUM) clinics in two English cities,
on
exploring their views of, and attitudes towards, UAT;14 qualitative in-depth interviews with a sub-sample (20) of these patients; and in-depth interviews with key informants in four
countries chosen to represent different approaches towards sero-surveillance for HIV. In this
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
paper, we present findings from this third component, the qualitative interviews with key informants.
The over-arching aim of this component of the research study was to use the subject of UAT, and our findings, to explore the relationship between public health policy, context and ethics. There were two more specific objectives of the research. The first was to gain a better understanding of how UAT has developed in different settings over time, with particular
5
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
attention to the ethical considerations. The second was to use UAT as an instrument to gain empirical understanding of the ethical tensions between surveillance and patients’ rights, and to explore how these tensions may have developed and been handled in different contexts.
METHODS
rp Fo
Four countries were selected purposively to participate in this study component (the key informant interviews), based on their representing a range of approaches, or positions taken, with regard to UAT: the UK (specifically England), USA, the Netherlands and Norway. Purposive sampling was used to select the key informants from these countries, informed by the following criteria: range of stakeholder perspectives (e.g. epidemiology, clinical, public
ee
health policy, charity or pressure group, and academia (public health or medical ethics)); involvement in UAT and with appropriate knowledge or expertise, historically and/or current; existing links with the research team or advisory group members; and practical issues (e.g.
rr
availability and likelihood of participation). A pragmatic approach meant that more informants were interviewed in the UK than the other countries.
ev
A topic guide was developed that allowed for a semi-structured interview. Elements within the topic guide included: introduction and background to the purpose and nature of the
ie
interview (prior information had been provided); knowledge and views on unlinked
w
anonymous testing of blood (and national programme as appropriate), including advantages and disadvantages; ethical considerations including consent, prior notification,
confidentiality, anonymity, individual rights/liberties, social responsibilities, and any changes
on
in ethical position over time; the value to public health of different surveillance methods;
blood as a tissue; legal considerations; and any other relevant issues the respondent wished to
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 32 of 53
discuss. The topic guide was initially tailored for the different countries based on our own understanding, but then applied flexibly during the interview.
A researcher, with a background and skills in qualitative interviewing, was employed for the duration of the project, and undertook all the interviews. Interviews were carried out, in English, either face-to-face or by telephone between January and September 2008. All nonUK informants spoke good English, without evidence to the research team of any obvious language barrier issues. Interviews were recorded (except for two, when notes were taken: the 6
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 33 of 53
first due to unsuitable location and the second to mechanical failure) with the participant’s permission, and transcribed verbatim. Interviews lasted between 20 minutes and one hour, the length directed by the contributions of the informants. In total 21 interviews were completed with 23 respondents (two interviews were with two respondents together). Of these, 12 were with UK respondents (three clinicians in genito-urinary medicine/HIV; one person working in government health policy; two HIV epidemiologists; two ethics academics; two charity representatives [MedFash, Terence Higgins Trust]; one person working at the British Medical
rp Fo
Association; one editor of a medical journal), four with respondents in the USA (one health policy academic; one academic in health law; one academic in public health; and one epidemiologist), three with respondents from the Netherlands (two epidemiologists; and one academic in medical ethics), and four with respondents in Norway (one epidemiologist; three academics in medical ethics – one of whom is also a hospital clinician). There were three interview refusals, all in the UK. [3]
ee
Two researchers (primary researcher, JD, and a second researcher, SP) coded the interview
rr
data using a thematic content analysis approach, looking for emergent themes (open coding in the context of the aims and objectives of the research). There were no discrepancies between
ev
the two researchers that required further discussion and consensus. Content was analysed for description of UAT developments in the country concerned, the context in which UAT may have changed, ethical tensions in UAT between surveillance and human rights, and broader
ie
connections between UAT and public health policy. Inevitably some of the themes inter-link
w
with each other (e.g. ‘value of UAT’ and ‘ethical issues’). Coding frames and analyses were shared with the wider research team. In a relatively small study of this kind it is not possible to indicate definitively if data saturation was reached, and the focus was on the emergent themes.
Interview data pertaining to the history and factual state of sero-surveillance in different
ly
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
countries were triangulated, for the purpose of the discussion section, with literature sources. It was not appropriate to triangulate informants’ views on sero-surveillance, as such data were treated as informants’ opinions and perceptions.
RESULTS
7
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
The results, which represent the views and perceptions of informants, are presented by themes and by country. The labels describe the themes that emerged, in the context of informants’ awareness of the aims and objectives of the study.
Value of UAT
UK
rp Fo
The programme is largely seen by respondents as indispensable in providing an unbiased measure of HIV prevalence and risk factors in the UK population. The most important outcome from the surveys has been information on undiagnosed HIV infection which, respondents suggested, has led to changes in policy on named testing; for example, HIV testing in pregnant women has changed from opt in to opt out, and sexual health clinics have
ee
moved from targeted to universal voluntary testing. The neonatal dried blood spot survey is used as a check on the performance of local antenatal screening. The public health rationale for the programme has not changed since 1990, and respondents cited the continuing high
rr
infection rate, particularly among men who have sex with men, as justification:
ev
"There was far more feeling of a public health emergency (then) than there is today, although you could argue … that it's actually much worse today because of where we see the disease." (UK4)
w
ie
Several respondents suggested that a review of UAT would be timely, including
consideration of extending the surveys to other health care settings and other diseases. UK6 discussed how advances in HIV testing, including self-testing, might alter the questions asked
on
by the programme. Some suggested that if named testing was normalised, there would be less need for anonymous testing.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 34 of 53
A particular theme was whether the British public has a duty to participate in public health
monitoring, possibly as part-and-parcel of receiving the benefits of a national health service (NHS):
"There is a very good case for really exploring with the public, and then pushing very hard to get public agreement, that in general terms that sort of tissue, extra bits of tissue or blood, should be available for use." (UK4) 8
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 35 of 53
USA According to one respondent, the surveillance programme was valuable in terms of planning, resource allocation, advocacy and interventions to prevent transmission. The heel prick survey provided accurate national data on prevalence in childbearing women and was seen as extremely valuable, with nothing since that has given as accurate a picture of national prevalence (USA 2).
rp Fo
However, given the questionable ethics of UAT (see later), as well as the resultant lack of good quality prevalence data, USA3 argued that enough became known about prevalence to target the most vulnerable groups for health promotion.
ee
“And how much more detail do you want from that, so what would you do differently if you knew that in the next level of detail.” (USA3)
Netherlands
rr
Unlinked anonymous surveys have been used in the Netherlands to measure HIV prevalence
ev
among high risk populations, but with informed consent (verbal or written). The populations surveyed by UAT have changed over time. Injecting Drug Users (IDUs) have been surveyed bi-annually since 1994 in different cities in the Netherlands, and are recruited via methadone
ie
care; sex workers and migrant groups have also been subject to UAT surveys.
w
There are two reasons, indicated by respondents, why UAT in Netherlands has been of limited value. First, the uptake of voluntary testing is generally high in the Netherlands,
on
meaning the quality of that data are generally good. And second, UAT has only been deemed ethically permissible in the Netherlands with informed consent, making large-scale UAT
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
surveys impracticable, so restricting UAT to specific population groups and/or the research context.
Norway The unknown, undiagnosed segment of the population is known as the ‘dark figure(s)’ in Norwegian, although neither the ‘dark figure’ nor the number of cases of HIV has been perceived as significant issues in Norway. Uptake of voluntary testing for HIV was described as very high in intravenous drug users and antenatal clinic users, meaning that UAT has never 9
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
been perceived as to be of value epidemiologically. The system of surveillance for HIV in Norway instead covers all diagnosed cases (i.e. not undiagnosed cases as with UAT), and was reported as being simple, inexpensive and capable of providing a representative and timely overview that guides prevention, while respecting confidentiality and adhering to ethical principles.
rp Fo
Quality and nature of data
UK
The epidemiological data yielded from unlinked anonymous testing are considered by respondents to be of unrivalled quality internationally. It was suggested, however, that the
ee
apparent high number of undiagnosed HIV infections from UAT surveys might be inflated by double-counting:
rr
"You can quite understand that if you've got gonorrhoea and you've been told by your doctor not to have unsafe sex, you might well decide to go to a different clinic to have your gonorrhoea treated." (UK1)
ev
Others defended the statistics and argued that they were explained by a resurgence of high
ie
risk behaviour, particularly among gay men. UK1 argued that people who refuse a named test
w
and who are subsequently found to be HIV positive through UAT are likely to be people who know that they are positive.
on
USA
There was little comment from USA interviewees about the quality of UAT data, since UAT
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 36 of 53
surveys stopped in 1996. However, a theme that emerged in relation to UAT data was around the difference between research and surveillance.
The fact that there is no ‘right’ to health care in the USA was felt to have an effect on people’s views on surveillance – in other words, the ‘disconnect’ between health care provision (not available to all) and surveillance (done to all) (USA 1). The differences between surveillance, research and practice, with perhaps different ethical standards required for each, emerged as a theme from the interviews (USA 4). 10
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 37 of 53
One point of view was that surveillance (because it is named and part of routine care) is part of clinical care rather than purely research (USA 2). However, while some of the surveillance undertakings at state level in the USA were considered to be practice, at the level of CDC they were treated as research (USA 1). The same respondent commented that cancer sufferers advocated in favour of surveillance for the purposes of research on the disease, since “they were willing to trade privacy and consent for a good which was surveillance that
rp Fo
they saw as serving their interests”.
Norway and the Netherlands
There were no comments from Norway respondents about the quality of UAT data since such surveys are not carried out in Norway, in part because of the good quality of data on HIV and AIDS from other sources.
ee
Two respondents from the Netherlands judged that large-scale UAT programmes were not
rr
necessary because named HIV testing is widespread. The quality of information obtained from named testing was felt to be robust and could be used to model estimates of HIV
ev
prevalence in the population. Coverage of pregnant women is very high and this group is considered a valid proxy for the heterosexual population. Named HIV testing among users of STI clinics was also presented as being a valid method for estimating prevalence. It was
ie
argued that, as long as the uptake of named testing by pregnant women and STI clinic users
w
remains high (above 90%) then a national programme of UAT was not thought to be necessary. However, notwithstanding the above point it was acknowledged that current methods of surveillance do not allow undiagnosed HIV infection to be measured accurately.
ly
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Ethical issues
UK Respondents felt that, on balance, no harm was caused by the programme. A recurrent argument for this was that, in settings where anonymous testing takes place, people are offered a named HIV test:
11
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
"I think the ethics are not in fact, are not a problem because everybody is now very strongly advised to have a test and therefore the residual rump who don't have a test, it's important to know what their rate of HIV positivity is." (UK1)
Some respondents, however, said there would be ethical concerns if people in these settings were not routinely offered a named test. Several respondents considered the ethics of UAT to be relatively unimportant, for instance in the context of more significant health-related problems in Africa:
rp Fo
"There are these enormous ethical, real ethical issues, and to me, this one … I just can't understand why one should put so much emphasis on it." (UK9)
ee
UK11 suggested that the question of ownership of the sample was relevant, but that this could be argued both ways:
rr
"Is it yours or is it somebody else's? ... If it's waste that's going to be chucked away then why shouldn't people use it if they find it useful? On the other hand, 'I gave
ev
permission for this test for my own benefit and I didn't give you permission to do things to help others.'" (UK11)
ie
There was universal agreement that changing to a system of informed consent would damage
w
the programme. There was a fear that introducing information about UAT into a clinical consultation could confuse patients. Patients might think they were getting a named test:
on
"You would lose someone's confidence and trust because they might not understand the words or the reason behind why we're doing it, and they would think that you could get back to them and you did know." (UK12)
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 38 of 53
Other respondents argued that it would reduce time for discussing important clinical matters:
"When the midwife turns up, does she pull out a leaflet and say by the way, I'm putting this beside you here? … it would be a perversion of the consultation to divert it into discussing this ..." (UK5)
12
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 39 of 53
Further concerns were that a system of informed consent would increase refusals and lead to bias in the surveys, and also the time and cost of gaining consent from patients:
"You know we've got ten minute appointments. It could take you five minutes, half of your time just to explain what this is all about. So doing it in a consultation is not realistic." (UK3) The respondent suggested that a system of clinic self-registration, using computers, could be used to explain anonymous testing.
rp Fo
All the respondents thought that prior notification was the right theoretical approach. However, some were concerned that, in practice, it might not be working:
"It's a reasonable thing in practical terms provided people are informed. So … if
ee
people don't know about their option for opting out and their option for finding out then the system hasn't been working." (UK11)
rr
Some suggested that the approach needed to be more "active" with information displayed visibly on clinic walls or in leaflets given out to each patient. One clinician, however,
ev
expressed his discomfort with his role in the programme. He described filling out the form for UAT blood testing in front of the patient, without explaining what he was doing, and sometimes after the patient had declined a named HIV test:
w
ie
"You're doing something that sort of seems slightly underhand…you're not explaining it to the patient. Yeah, I'm not comfortable with that, I think one wants to be, to communicate well and to be as open as possible with patients and that's slightly in conflict with that." (UK3)
ly
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
USA
Consent was not deemed necessary initially in the USA for participation in UAT as patients had consented to the blood being collected in the first place, and this was seen as adequate. USA1 suggested that surveillance is a public good, meaning there are strong reasons for making it a universal obligation on citizens – and that notification that surveillance is occurring is different from asking for consent.
13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
The blood spot survey continued until the publication of the first trials showing that vertical transmission of HIV could be dramatically reduced by treating mothers. When it became clear that infected women could be offered treatment, and therefore should know their HIV status, “it became impossible for the US Centre for Communicable Disease Control and Prevention (CDC) to continue that programme and it was dropped in the early 1990s along with all the other ones” (USA3). It had become difficult to reconcile public health benefits of UAT with individual rights. If CDC had not voluntarily withdrawn funding from UAT, the
rp Fo
same respondent indicated that it was likely that Congress would have mandated cessation, as there was a strong coalition in favour of this.
A related legal issue was brought by USA4, in that public health surveillance appears to go against “fundamental legal principles” in the USA, and constitutional scholars are shocked when they hear what is done.
ee
‘Once you begin to argue that the government can take your personally identifiable
rr
information for any reason it finds useful without your consent you've opened up and blown a hole in much of the fundamental legal principles in the country.’ (USA4)
ev
The Tuskegee episode (a study between 1932 and 1972 in Alabama in which black syphilitic Americans were not given treatment although thought they were) was mentioned by USA1 as
ie
a reason for USA citizens, particularly Afro-Americans, to be suspicious of surveillance
w
practices. In the context of opposition to named case reporting and a consequent lack of trust in surveillance, USA3 argued that there are probably cultural dimensions to issues such as consent, and the acceptability of particular interventions may vary from country to country.
on
“It [UAT] became more of a civil rights issue than a gay specific one. It was, the
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 40 of 53
government knows something that you don’t and the attitude in the States, at least at the time, towards the government was much less trusting than in this country [i.e.
UK]. People did not assume and still don’t assume that the government necessarily is looking out for your well being. And I think there’s a much more paranoid attitude towards the government than there is in this country.” (USA3)
14
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 41 of 53
Related to these cultural dimensions, a particular ethical concern was that the US (and, in particular, CDC) supports UAT in developing countries, with the inherent contradiction that it has been rejected in the USA.
‘It seems to me the CDC can’t at the one time say it’s unethical to do it in the US and say it’s ethical to do it abroad. Different public agencies may come to different decisions, but the same agency it seems to me should be bound by some rule of
rp Fo
consistency.’ (USA1)
USA2 felt very strongly that arrangements for UAT in developing countries are not ethical where pregnant women do not have the opportunity to have a named test, counselling and treatment.
ee
Netherlands
The respondents felt that more extensive UAT surveys would produce more accurate
rr
information for public health surveillance. However, inter-linked practical and ethical doubts about UAT were presented. One concern was that UAT might deter high risk groups from
ev
seeking testing and treatment. Another concern was that, if UAT focused only on high risk groups, such groups would risk being stigmatised. It was suggested that reporting of findings from UAT surveys of high risk groups would need to be carried out sensitively.
w
ie
A particular issue was consent and maintaining trust with patients. One interviewee expressed that staff in HIV treatment centres and STI clinics demand a more stringent consent process than required by the research ethics committees in the Netherlands in order to promote trust
on
among patients.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Norway
A UAT system was never thought relevant, necessary or appropriate for Norway. In Norway it is not legal to test or analyse blood or biological material for infectious diseases for non-
diagnostic purposes without consent unless the purpose is surveillance of an ‘epidemic’. One respondent felt this displayed a ‘moral grey zones’, as the law hinges on the definition of ‘epidemic’ [Norway 1].
15
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
There is very little concern or awareness in Norway about ethical issues around UAT. However, there is a general willingness to provide information for public health purposes in Norway, and people are used to being surveyed. It was reported that many Norwegians have given blood to the Biobank, and give broad consent for their tissue/blood to be used for a wide range of research. A respondent indicated that many are willing to waive their consent in Norway because ‘there is significant trust’ in the health care system. However, another respondent expressed that there was a possible danger that UAT blood could be used to
rp Fo
identify someone via the Home Office DNA database, and that researchers might be forced to share information under a court order (the respondent mentioned that the murderer of the Swedish politician Anna Lindh in 2003 was convicted on the basis of a match with DNA held in the national biobank).
ee
One respondent suggested that informed consent could be different in different contexts. In a country like Norway, with low prevalence and easy access to health care, the current surveillance system (non-UAT) was described as acceptable and functioning well. However,
rr
in a country in Africa (where this respondent was carrying out research), with a high rate of HIV positive pregnant women and a low rate of testing, routine testing with opt-out consent procedures would be morally acceptable.
ie
ev
DISCUSSION
w
Since its inception, unlinked anonymous testing (UAT) has been a subject of
epidemiological, ethical and policy debate.15 16 17 In this paper we have presented the results
on
of qualitative interviews with key informants in four countries pragmatically selected because they represent a range of approaches to UAT. 18 A total of 23 respondents were interviewed, differentially distributed between countries: 12 in the UK, four in the USA, four in Norway and three in the Netherlands.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 42 of 53
There were limitations to the research: more key informants were interviewed in some countries than others; some interviews were face-to-face whereas others were by telephone; and two of the interviews relied on notes rather than recordings for later analysis.18 19 20 It is
Formatted: Endnote Reference
possible that such aspects introduced bias, but we believe that the balanced pragmatic approach taken to address the research questions was appropriate for this relatively small
16
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 43 of 53
qualitative study. Interesting themes emerged in the analysis, and we were able to triangulate more fact-based comments with documentary sources in the literature.
First a summary of key findings is presented (in the context of historical development of UAT in different countries, as well as what is already known of the subject) , followed by a discussion of how the findings from this paper fit into broader contemporary debates about public health ethics and policy.
rp Fo
Summary of key findings from interviews (in the context of what is currently known)
In the United Kingdom (UK), a balance of epidemiological and ethical considerations in the
ee
late 1980s led to the implementation of the ‘prior notification’ approach to UAT; this involves leaflets and posters informing clients that UAT is occurring, allowing opting out but not necessitating the gaining of informed consent.777 101010 Respondents stressed that the
Formatted: Endnote Reference
rr
Formatted: Endnote Reference
rationale remains apt today, and that this approach provides data unrivalled in terms of
Formatted: Endnote Reference
epidemiological quality,999 21 22 and which is of use for health service planning.23 Ethical
Formatted: Endnote Reference
ev
concerns were generally felt to be minimal today since clients are (also) usually offered
Formatted: Endnote Reference
voluntary (named) testing in settings where UAT is occurring. However, there was mention
Formatted: Endnote Reference
of concern about the lack of full explanation, whether people do actually know UAT is
ie
happening, and ‘ownership’ of blood.24 It was indicated that an approach involving informed consent would be cumbersome, time-consuming and epidemiologically damaging.25
w
In the United States (USA), UAT of blood for HIV was extensively used in the 1980s and
on
early 1990s,26 but was stopped in the late 1990s as evidence emerged that anti-viral therapy in pregnant women could help reduce the vertical transmission of HIV from pregnant mother to
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
child – named testing needed to be maximised so that HIV-positive pregnant women could be identified and offered advice and therapy.27 The interviews indicated that data from UAT
were useful epidemiologically but, since UAT has stopped, there has been sufficient other data to plan services accordingly. Other issues raised include a general suspicion around epidemiological surveillance post-Tuskegee, cultural dimensions of UAT with respect to minority groups, a lack of trust in the Government, and the related disconnect between surveillance (done to all) and health care provision (variable, and not done to all).
Formatted: Endnote Reference Formatted: Endnote Reference
666 111111
Formatted: Endnote Reference
There was also comment about the apparent contradiction in the contemporary support of the
Formatted: Endnote Reference
17
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
US Centre for Communicable Disease Control and Prevention (CDC) for UAT in lowincome countries.28
Interviews in the Netherlands showed that named (voluntary) testing, which has generally been widely taken up, has provided the main vehicle for surveillance of HIV.444 There is
Formatted: Endnote Reference
acknowledgement that large-scale UAT would provide additional epidemiological data, but
Formatted: Endnote Reference
this has been balanced against concerns of this leading to diminished voluntary testing –
rp Fo
especially in at risk groups – through loss of trust in the healthcare system.29 Smaller UAT surveys have been intermittently undertaken in such groups (e.g. intravenous drug users, sex workers, migrants) as research studies, but have required individual consent (initially oral, now written).
ee
Interviews in Norway revealed that UAT was never felt to be necessary epidemiologically, as HIV has not been a significant issue and the uptake of voluntary testing has been high.131313
Formatted: Endnote Reference Formatted: Endnote Reference
Surveillance has thus tended to focus on the monitoring of the diagnosed population
rr
(anonymised diagnosed cases), which has been supported by the gay community; the undiagnosed population is interestingly referred to as ‘dark figure(s)’. In Norway it is not
ev
legal to test without consent unless there is an epidemic, although respondents reported few ethical concerns since people generally have trust in the health care system, and are willing to provide information for public health purposes (many give blood to the Biobank for a range of research purposes).30
w
ie
The findings of this research on the ethics of UAT of blood need consideration in the context of two broader contemporary debates within medical and public health ethics. The first
on
debate centres on the boundaries of informed consent in health research, particularly in relation to the use of human tissue. The second debate focuses on the balance between
individuals’ rights and the appropriate role of government3 in infringing on those rights for
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 44 of 53
broader public health interests or concerns. These debates are inter-connected, although will be presented in turn.
Informed consent and ethics: a changing debate
18
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 45 of 53
Respect for autonomy is the bioethical principle that is most likely to be jeopardised by a programme of unlinked anonymous blood testing, since informed consent is not usually deemed necessary for an individual’s blood to be included; this has tended to be the basis for opposition to UAT.555 31 The findings from our research show, however, that such positions
Formatted: Endnote Reference Formatted: Endnote Reference
differ between countries, and have also changed with time.
Debates in the late 1980s and early 1990s demonstrated the opposing views of those who
rp Fo
were concerned about patient consent,888 262625 and those who felt that the priority was to control what was seen as an epidemic.
151515
Formatted: Endnote Reference
UAT programmes did go ahead in the USA and
Formatted: Endnote Reference Formatted: Endnote Reference
UK, and other high income countries: there was no requirement for consent in the USA,
Formatted: Endnote Reference
whereas in the UK an approach of ‘prior notification’ was adopted through leaflets and
Formatted: Endnote Reference Formatted: Endnote Reference
posters in clinics – consent is presumed but individuals have the opportunity to opt out.
ee
Over the past two decades informed consent has emerged as ‘the ethical touchstone of medical research, and is now enshrined in practice and a range of international guidelines as
rr
one of the main preconditions of medical research’.252524 32 Some commentators argue that
Formatted: Endnote Reference Formatted: Endnote Reference
even anonymised tissue should not be used without consent. Trouet, for example, believes
ev
that the use of such material is in contravention of Article 22 of the Convention on Human Rights and Biomedicine. This states that ‘when in the course of an intervention any part of a human body is removed, it may be stored and used for a purpose other than that for which it
ie
was removed, only if this is done in conformity with appropriate information and consent
w
procedures’, and in accordance with Article 8 of the European Convention on Human Rights which protects the right to respect for private life.33
on
Our research indicates that, rather than the universality of human rights discourse, a form of
weak moral relativism operates in policy-making and practice around UAT.34 In the USA, the
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
change in social context and debates about individual rights, along with evidence of reduction in mother-to-child vertical transmission, resulted in the abandonment of UAT in 1996.121212
Formatted: Endnote Reference
The fact that the USA continued to support UAT in international research overseas illustrates
Formatted: Endnote Reference
further an attitude of moral relativism, although it could be argued that differential health needs justified different approaches.
Interviews in the UK showed that the prior notification approach that remains in use today still holds strong support – based on perception that there is minimal harm to the individual, 19
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
and the value of high quality, unbiased epidemiological data on HIV (held to be the most robust and accurate worldwide).35 Respondents felt that any change to UAT requiring informed consent would be cumbersome, time-consuming, and costly. According to the Nuffield Council on Bioethics it may also be ethically unnecessary: ‘... it is acceptable to collect and use anonymised data for assessing and predicting trends in infectious disease without consent, as long as any invasion of privacy is reduced as far as possible.’.36
rp Fo
This position is not, however, cross-culturally held. In the Netherlands, UAT has not been felt to be acceptable without informed consent, initially verbal and more recently written. The societal emphasis on individual autonomy, including in the health research environment, has been the paramount consideration.555 Northern European and Scandinavian countries are
Formatted: Endnote Reference Formatted: Endnote Reference
often held up as having a strong community ethos, with an expectation on individual
ee
involvement in endeavours for the public good.37 This was not reflected in UAT practice in the Netherlands, although was indicated to be important in Norway, albeit in a setting where UAT has never been taken up.
rr
These findings indicate the importance of social and political context in relation to medico-
ev
ethical debates, and how attitudes and policies can change in time within any given jurisdiction.38 Differences exist between the moral evaluations and resultant UAT approaches between countries such as England and the Netherlands. Within the USA, in contrast, the
ie
practice of UAT has been re-evaluated with time. In the late 1980s and early 1990s the feared
w
AIDS epidemic was seen as a public health emergency, there was no treatment available, transmission from mother to unborn child was not preventable, and the stigma and discrimination associated with the infection were more overt than today. As these factors
on
shifted, so did the policy on UAT in the USA. The situation in low income countries today has created a perception of the acceptability of UAT akin to that in western countries two
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 46 of 53
decades ago.
Individual good and community (public) good
There has been criticism in recent times of government for its ‘nanny state’ approach to interventions in, for example, the areas of smoking, alcohol and diet. More recently the notion of ‘stewardship’ in public health rejects these negative associations,39 and advocates 20
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 47 of 53
that government should have a role to play in protecting the health of its citizens, preventing them from harming themselves or others, and promoting healthy behaviour.[4] In this view, public health interventions should not be thought of as unwanted restrictions imposed by a ‘nanny’ on powerless children, but as actions that favour both individuals and society more widely as prescribed by a ‘steward’ administering and protecting citizens’ interests. The Nuffield Council on Bioethics, for example, argues that the state should act as steward both to individuals and to the population as a whole, but that individuals’ personal choices should
rp Fo
also be taken into account.363635
Formatted: Endnote Reference Formatted: Endnote Reference
The subject of UAT sits squarely within the debate about public health stewardship for social good. Recently, commentators have argued that patients of a publicly supported health care system like the NHS have a moral duty to participate in scientific research,40 41 although most
ee
do not go so far as to articulate that such participation should be mandatory in practice.42 Nevertheless, UAT could be understood as something that individuals should feel they have a moral obligation to participate in, regardless of ethical concerns about individual rights and
rr
personally held views. Our own survey of over 400 users of two genitor-urinary medicine clinics (a parallel part of this research project) has found that a large majority of respondents
ev
(89%) agree that people should have a responsibility to take part in research studies, and the same proportion would agree to the use of their blood in unlinked anonymous seroprevalence testing. However, when asked about the question of consent, 74% indicated that they should
ie
be asked to consent before their blood is used for public health surveillance.141414
Formatted: Endnote Reference Formatted: Endnote Reference
w
Notwithstanding this important latter point, such support for use of blood samples for prevalence surveys or for future research has not, however, been demonstrated in an earlier study in the UK43 or more recently in Egypt.44
on
The way forwards perhaps is to ensure that patients – and the public more widely – are
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
informed of the existence, purposes and value of disease surveillance arrangements, as well
as the importance of contributing to future health research, while investigating arrangements
for gaining consent which are not unnecessarily cumbersome or likely to create bias amongst those who agree that their blood be used.45
The issue of consent in seroprevalence surveillance, such as UAT, remains an important instrument to gauge the moral direction of public health policies more broadly.46 Debate continues around how public health programmes can embrace a stewardship relationship, 21
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
rather than a nanny paternalistic one, while respecting autonomy and advocating social responsibilities. Further examination of why different countries have adopted different positions on seroprevalence surveillance will help illuminate our understanding of how social morality is embedded in public health policies and practice.
Acknowledgments
rp Fo
The authors would like to thank the participants in the study. We also acknowledge the valuable advice given by Vivian Hope, Will Nutland, Cicely Marston, and Graham Hart.
Conflicts of interest
AK is also Director of Public Health Strategy and Medical Director of the Health Protection
ee
Agency. The views expressed in this paper are those of the authors, and are not intended to represent the views of the Health Protection Agency.
Funding
ev
rr
This research was supported by the Economic and Social Research Council (ESRC Grant Reference: RES-000-22-2096).
ie
Ethical approval
The study was approved by the South West Research Ethics Committee on August 16th 2007
w
(reference 07/H0206/46), and by the London School of Hygiene & Tropical Medicine ethics committee on 3rd September 2007 (reference 5174).
on
Notes
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 48 of 53
Formatted: Line spacing: single
[1]
The England and Wales programme is currently managed by the Health Protection
Agency (HPA) and the Institute of Child Health, University of London. The programme was previously managed by the Communicable Disease Surveillance Centre, part of the Public Health Laboratory Service, which was subsumed into the Health Protection Agency in 2005. [2]
Health Protection Scotland (HPS) co-ordinates data collection in Scotland.
[3]
Those chosen had particular involvement or interest in UAT and ethics, but further
details cannot be provided to maintain their confidentiality.
22
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 49 of 53
[4]
By government, the meaning here is health and public health policy makers within
government.
References
1
rp Fo
Nicoll A, Gill ON, Peckham C et al. The public health applications of unlinked
anonymous seroprevalence monitoring of HIV in the United Kingdom. International Journal of Epidemiology 2000;29:1-10. 2
Department of Health. Prevalence of HIV in England and Wales in 1997: Annual
ee
Report of the Unlinked Anonymous Prevalence Monitoring Programme. London: Department of Health, 1998. 3
rr
UK Collaborative Group for HIV and STI Surveillance. Testing Times: HIV and
Other Sexually Transmitted Infections in the United Kingdom. London: Health Protection
ev
Agency, 2007. 4
Bayer R, Lumey LH, Wan L. The American, British and Dutch responses to unlinked
ie
anonymous HIV seroprevalence studies: an international comparison. AIDS 1990;4(4):283290. 5
w
Kopelman LM. Informed consent and anonymous tissue samples: the case of HIV
seroprevalence studies. Journal of Medicine and Philosophy 1994;19:525-52. 6
Fairchild AL, Bayer R. Ethics and the conduct of public health surveillance. Science
ly
2004;303:631-632. 7
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
Gill ON., Adler MW, Day NE. Monitoring the prevalence of HIV: foundations for a
Programme of unlinked anonymous testing in England and Wales. British Medical Journal 1989:299:1295-7. 8
Berridge V. AIDS in the UK: the Making of Policy 1981-1994. Oxford: Oxford
University Press, 1996.
23
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
9
Health Protection Agency. Overall HIV prevalence. Health Protection Agency 2008.
http://www.hpa.org.uk/web/HPAweb&Page&HPAwebAutoListName/Page/1201094588821. 10
Gillon R. Testing for HIV without permission. British Medical Journal 1987;294:821-
823. 11
Fairchild AL, Bayer R, Colgrove J. Searching Eyes: Privacy, the State, and Disease
rp Fo
Surveillance in America. Berkeley: University of California Press, 2007. 12
Stoto MA. Almario DA, McCormick MC. Reducing the Odds: Preventing Perinatal Transmission of HIV in the United States. Washington: National Academy Press, 1999.
13
Aavitsland P, Nilsen Ø, Lystad A. Anonymous reporting of HIV infection: an
evaluation of the HIV/AIDS surveillance system in Norway 1983-2000. European Journal of Epidemiology 2001;17:479-89. 14
ee
Datta J, Kessel AS, Wellings K, Nanchahal K, Marks D, Kinghorn G. The views of
rr
genitourinary medicine (GUM) clinic users on unlinked anonymous testing of blood for HIV: evidence from a pilot study of clinics in two English cities. Journal of Medical Ethics 2011 doi:10.1136/jme.2011.042705. 15
Doll R. A proposal for doing prevalence studies of AIDS. British Medical Journal
w
ie
1987;294:443-444. 16
ev
Avins AL, Lo B. To tell or not to tell: the ethical dilemmas of HIV test notification in
epidemiologic research. American Journal of Public Health 1989;79/11:1544-7. 17
on
Datta J, Kessel AS. Unlinked anonymous testing for blood purposes: an ethical
dilemma? In: Peckham S, Hann A, eds. Public Health Ethics and Practice. Bristol: Policy Press, 2010:101-116.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 50 of 53
18
Bowling A. Research Methods in Health. Buckingham: Open University Press, 2000.
19
Yin RK. Case study research: design and methods. London: Sage, 1994.
20
Green J, Thorogood N. Qualitative methods for health research. London: Sage, 2004. Formatted: English (U.K.)
24
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 51 of 53
21
Duong T, Ades AE, Rogers P, et al. Non-participation bias in unlinked anonymous
HIV-prevalence surveys in England and Wales. Epidemiology and Infection 1999;122:26772. 22
Pulver WP, Glebatis D, Wade N, et al. Trends from an HIV seroprevalence study
among childbearing women in New York State from 1988 through 2000: a valuable
rp Fo
epidemiologic tool. Archives of Pediatric and Adolescent Medicine 2004;158(5):443-8. 23
Kessel AS, Watts CJ. Usefulness of information from the Unlinked Anonymous
Prevalence Monitoring Programme for HIV in England and Wales: survey of planners of HIV/AIDS services. International Journal of STD and AIDS 1999;10:808-11. 24
Pfeffer N, Laws S. 'It's only a blood test': what people know and think about
ee
venepuncture and blood. Social Science & Medicine 2006;62:3011-3023. 25
Cassell J, Young A. Why we should not seek individual informed consent for
rr
participation in health services research. Journal of Medical Ethics 2002;23:313-317. 26
Bayer R. The ethics of blinded HIV surveillance testing. American Journal of Public
Health 1993;83/4:496-7.
de Zulueta P. The ethics of anonymised HIV testing of pregnant women: a
reappraisal. Journal of Medical Ethics 2000;26:16-21.
w
28
ie
27
ev
Rennie S, Norris Turner A, Mupenda B, et al. Conducting unlinked anonymous HIV
surveillance in developing countries: ethical, epidemiological, and public health concerns. PLoS Medicine 2009;6(1):e1000004. 29
on
Postema EJ, Willems PWJM, de Ridder MAJ, et al. Comparison of patients refusing
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
with patients accepting unlinked anonymous HIV testing in an outpatient STD department in the Netherlands. International Journal of STD and AIDS 1997;8:368-72. 30
Skolbekken J-A, Ursin LO, Solberg B, et al. Not worth the paper it’s written on?
Informed consent and Biobank research in a Norwegian context. Critical Public Health 2005;15(4):335-347.
25
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
31
Beauchamp TL and Childress JF. Principles of Biomedical Ethics. New York: Oxford
University Press, 1994. 32
Singleton P, Wadsworth M. Consent for the use of personal medical data in research.
British Medical Journal 2006;333:255-8. 33
Trouet C. New European guidelines for the use of stored human biological materials
rp Fo
in biomedical research. Journal of Medical Ethics 2004;30:99-103. 34
Rachels J. The Elements of Moral Philosophy. New York: McGraw-Hill, 1993.
35
Beuhler JW, Petersen LR, Ward JW, et al. Defending HIV seroprevalence surveys.
American Journal of Public Health 1994;84:319-320. 36
Nuffield Council on Bioethics. Public Health: Ethical Issues. London: Nuffield
Council on Bioethics, 2007
rr
37
ee
Cappelen AW, Norheim OF. Responsibility in health care: a liberal egalitarian
approach. Journal of Medical Ethics 2005;31:476-80. 38
Mariner WK. Mission creep: public health surveillance and medical privacy. Boston
Jochelson K. Nanny or Steward? The Role of Government in Public Health. King's
w
Fund Working Paper. London: King's Fund 2005. 40
ie
University Law Review 2007;87/2:347-95. 39
ev
Harris J. Scientific research is a moral duty. Journal of Medical Ethics
2005;31(4):242-248.
on
41
Evans HM. Do patients have duties? Journal of Medical Ethics 2007;33(12):689-694.
42
Shapshay S, Pimple KD. Participation in biomedical research is an imperfect moral
duty: a response to John Harris. Journal of Medical Ethics 2007;33:414-417. 43
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 52 of 53
Kessel AS, Watts C, Weiss HA. Bad blood? Survey of public's views on unlinked
anonymous testing of blood for HIV and other diseases. British Medical Journal 2000;320: 90-91.
26
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
Page 53 of 53
44
Abou-Zeid A, Silverman H, Shehata M, et al. Collection, storage and use of blood
samples for future research: views of Egyptian patients expressed in a cross-sectional survey. Journal of Medical Ethics 2010;36:539-547. 45
Nicholls SG. Knowledge or understanding? Informed choice in the context of
newborn bloodspot screening. Public Health Ethics 2010;3:128-136. 46
rp Fo
Haimes E. What can the social sciences contribute to the study of ethics? Theoretical,
empirical and substantive considerations. Bioethics 2002;16/2:89-113.
w
ie
ev
rr
ee ly
on
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
BMJ Open
27
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml