Sexual abstinence only programmes to prevent HIV infection in high income countries: systematic review
BMJ 2007; 335 doi: https://doi.org/10.1136/bmj.39245.446586.BE (Published 02 August 2007) Cite this as: BMJ 2007;335:248
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Prof. De Irala’s measured and detailed analysis of Underhill’s recent
paper which purported to show that “abstinence doesn’t work”, does a very
comprehensive job of stating why the paper is not, and should not, be the
last word on the issue in the scientific community, even though the media
have swallowed wholesale the spin given on it. One further point not
raised by Prof. De Irala however is that, of the 13 interventions
included, “No trial assessed or reported outcomes by sexual orientation”
(1).
In the USA and UK, in common with most Western nations, new HIV
infections are still disproportionately acquired by men who have sex with
men - nearly a third of new cases in 2005 for the UK (2). The proportion
of new cases of HIV infection acquired in heterosexuals in the UK as the
country of acquisition remains comparatively low at 15% (2). If these
patterns are similar to those in the USA, then a statistically significant
effect of any sex education programme on HIV infection rates, whether it
was abstinence or condom-based, would be much more difficult to
demonstrate when outcomes are not reported by sexual orientation.
1) Underhill K, Montgomery P, Operando D. Sexual abstinence only
programmes to prevent HIV infection in high income countries: systematic
review. BMJ2007;335:248-252
2)http://www.hpa.org.uk/publications/2006/hiv_sti_2006/part2hiv.htm
Competing interests:
Trevor Stammers is the author of "Saving Sex: Answers to Teenagers' Questions on Relationships and Sex" and is a consultant to www.loveforlife.org.uk
Competing interests: No competing interests
Articles such as the one published by Underhill and colleagues (1),
where sex education programmes are reviewed, tend to be picked up by the
news media and make headlines. The problem is that conclusions are too
often made hastily in spite of complex data and analyses. Underhill and
colleagues have reviewed 13 trials of abstinence only programmes and have
done a great job putting together and summarising very heterogeneous
studies. Indeed, the methodological pitfalls they have encountered within
all studies are a sign of how difficult it is to achieve well designed
studies for evaluating the effect of educational programmes on human
behaviour, and/or to implement such studies. Most of the methodological
problems cited by Underhill and colleagues are crucial:
- lack of intention to treat analysis (intention to treat analysis tends to underestimate any
effect),
- heterogeneity in programme and trial designs making it impossible to
perform a meta-analysis (and this means it is impossible to obtain a valid
quantitative summary of programme performance),
- wide range of programme lengths (from one to 720 sessions and a median
duration of 8 sessions),
- control groups were mostly “usual care groups” but such groups were
rarely defined and reviewers affirm “they could have included any
programme type”,
- missing information, making the assessment of methodological quality
difficult,
- only four trials reported procedures for generating the sequence of
assignment to intervention or control group and no trial reported
procedures for concealing the randomisation process,
- attrition rates ranged from 5% to 45% (median 20%).
What puzzles me most, however, is the conclusion drawn by the authors
after such a meticulous job reviewing the trials. They bluntly state that
programmes that exclusively encourage abstinence are ineffective for
preventing HIV and, by implication, generally ineffective. This conclusion
is furthermore corroborated by the friendly editorial of Hawes and
colleagues (2) who speak of the studies being “remarkably consistent” in
suggesting that abstinence only programmes increased neither primary nor
secondary abstinence. The editorialists takes the opportunity to suggest
that “In contrast to abstinence only programmes, programmes that promote
the use of condoms greatly reduce the risk of acquiring HIV”, citing three
papers, two of which are from the late 1990s. The editorial finally argues
that money should not be spent on abstinence only programmes but rather on
condom promotion programmes. I am not sure under what criteria other
studies showing the contrary are being excluded before such statements are
made. For example, a trial found an increase in HIV risky behaviours in
the intervention group where condom use and supply was promoted (3). And a
meta-analysis by Dicenso and colleagues showed programmes, including
family planning clinics, were actually not impressively effective
improving contraceptive use, delaying sexual debut nor avoiding unwanted
pregnancies (4). No one then requested the elimination of funding to
family planning clinics.
In addition to the summary of problems stated above, the statistical
adjustment for possible confounding in these studies is by no means
clearly described. The reviewers strongly feel that “Internal and external
validity are further restricted by limitations in the primary trials, lack
of intention to treat analysis, and incomplete reporting of program
implementation”. In the absence of internal validity, extreme caution
should be exercised when interpreting any epidemiological results, even if
reviewers collect data using: “prereviewed Cochrane protocol”. What is
most worrisome, because it is not sufficiently described in the review, is
to what extent the attrition rates were differentially distributed between
the groups compared and to what extent this hinders any valid
interpretation. Furthermore, in the light of: (a) the statistical
adjustments likely performed, (b) some of the sample sizes described and
(c) the probable scarcity of outcomes in some age groups, either in the
intervention or in the control group, non-significant results can be
interpreted in different ways. Non significant results can suggest
“absence of differences” (which is what reviewers argue), only when the
study design and the statistical power of such studies are appropriate.
However, non-significant results may also suggest there is no evidence
for, or against, an existing difference due to lack of statistical power.
Among all comparisons made in the 13 studies reviewed what is remarkable
is that most studies find non-significant results. When this happens, we
epidemiologists usually have a look on whether non-significant differences
between groups favour the control or the intervention group. In the paper
of Underhill and colleagues, approximately half of the effect measures
they present tend to favour the control groups whereas the other half tend
to favour intervention groups.
I see no clear connection between the results of the 13 studies with their
numerous and severe methodological problems and the reviewers’ clear-cut
conclusion that programmes that exclusively encourage abstinence are
ineffective. I contend that other interpretations are as plausible. For
example, we could also conclude that abstinence only programmes seem to be
“at least as good” as their alternatives. One could similarly argue that
money could be allocated to any of the alternative programmes tested in
the trials, given they presented similar results; or, conversely, that
neither abstinence only nor their alternatives (some of which promoted
condoms) should receive money at all in the USA. Another plausible
conclusion is that the situation could have been worse had neither the
abstinence only programmes nor their alternatives been implemented.
Finally, it is very misleading and not methodologically sound, to sum up
the sample sizes of the trials reviewed and to speak of “more than 15900
participants” included in the studies, as both the reviewers and the
editorialists state on several occasions. If the data are not valid for a
proper meta-analysis, then no aggregate measure should be used to convey
any quantitative information. A smaller well-designed trial can be more
informative than an aggregate of studies presenting serious methodological
problems even if, overall, they have thousands of participants.
In the light of such data it might be safer to say there is no evidence
that these 13 particular abstinence only programmes reviewed have done any
better than the alternatives evaluated. This does not mean “abstinence
promotion does not work”, which is what some are trying to convey to the
public. For example, the Spanish paper “El País”, on August 12th, 2007.
But, irrespective of whether these programmes actually worked or not, the
problem with this controversial issue is that we might not be asking
ourselves the right questions, or we might not be looking for the right
evidence. Does any body really believe it is possible to change any human
behaviour with a dozen sessions in schools, if parents at home, television
programmes, movies, youth magazines, health and educational authorities
and society at large convey the opposite message? Think of gender
violence, sexism, discrimination, academic failure, lack of exercise,
unhealthy eating, the problem of drinking and driving, smoking and other
drug taking. Would a dozen of classes at school suffice to change these
behaviours if everywhere else the message was different? The question for
these issues is “how” can we convey the right message and not “whether” we
should convey them. If a programme aiming to prevent gender violence does
not succeed, it would be a terrible mistake to conclude that: “the
education against violence is not effective”. We would rather have to
think of a way to do it better given that this particular programme had
failed, or we would have to think of how we could help this program to
succeed. Let us not forget many anti-smoking programmes have little
success and no one doubts we should prevent smoking in youth. Do we really
expect that “abstinence promotion” during a few school sessions will work
in a society where the media are conveying exactly the opposite message?
(5). The question is: do we really believe abstinence is a good choice for
our youth and do we really want to promote abstinence?
I am not necessarily a defender of “abstinence only” programmes, at least
not for older adolescent age groups. I personally believe the whole truth
is the best we can give to our youth in order to help them make better and
healthier choices. But we should empower youth, to make the best choices
and, when behaviours are involved, this includes character education. It
is always best to “avoid risks” rather than to “reduce risks” and messages
should be tailored to specific target groups. There is sound
epidemiological evidence in favour of the ABC strategy of prevention.
Abstinence and being mutually faithful are best for avoiding risk whereas
condoms reduce risks in people who choose not to avoid risks with “A” nor
“B”.
The 2004 Lancet consensus (6) emphasised the importance of prioritising
messages by calling for a delay of sexual debut in youth or for the return
to abstinence in those who are having casual sex. When having sex is
chosen, the consensus prioritised the message of mutual monogamy. Those
who choose not to accept “A” nor “B”, should be advised they can reduce,
albeit never totally eliminate, the risk of infection. The Lancet
consensus signers do not believe it is sound public health policy to give
the exact same priority to one message (condom use) to teens who have not
began to be sexually active and to persons working in commercial sex. All
the truth should be conveyed but “abstinence plus” programmes have to be
“abstinence centred” and not just programs that add condom information to
abstinence promotion at the same level. This is crucial to avoid the
slippery slope of risk compensation (7, 8). There is evidence showing such
programs are helpful (9). No African country has succeeded in reducing HIV
incidence with programmes which relied exclusively on condom promotion
whereas the countries which have reduced HIV incidence have integrated “A”
and “B” in comprehensive national programmes (10).
Our major problem is to decide what we want to convey to our youth. It is
unlikely any limited and/or short programme will help change any risky
behaviour unless youth are given truthful information, and unless they are
empowered with life skills such as character education. This can hardly be
achieved, unless society at large and especially educational and health
authorities make the right effort to convey consistent messages to
specific target groups, thus helping parents do their job at home as well.
Are we ready to convey what is best for our children and rely on their
capability of making right decisions? Or, should we pessimistically and
patronisingly decide for them that they can not achieve risk avoidance and
that they have no other choice rather than reducing risks?
Jokin de Irala MD
Master of Public Health (University of Dundee, Scotland)
Ph.D. Medicine (University of Navarre, Spain)
Ph.D. Biostatistics and Epidemiology (University of Massachusetts)
References
(1) Underhill K, Montgomery P, Operando D. Sexual abstinence only
programmes to prevent HIV infection in high income countries: systematic
review. BMJ2007;335:248-
(2) Hawes S, Sow PS, Kiviat NB. Is there a role for abstinence only
programmes for HIV prevention in high income countries? BMJ 2007;335:217-
218
(3) Kajubi P, Kamya MR, Kamya S, Chen S, McFarland W, Hearst N. Increasing
condom use without reducing HIV risk: results of a controlled community
trial in Uganda. J Acquir Immune Defic Syndr 2005; 40: 77-82.
(4) DiCenso A, Guyatt G, Willan A, GriffithL. Interventions to reduce
unintended pregnancies among adolescents: systematic review of randomized
controlled trials. BMJ 2002;324:1426-1435
(5) Collins RL, Elliott MN, Berry SH, Kanouse DE, Kunkel D, Hunter SB, Miu
A. Watching sex on television predicts adolescent initiation of sexual
behavior. Pediatrics 2004;114:280
(6) Haleprin D, Steiner M, Cassel M, Green E, Hearts N, Kirby D, et al.
The time has come for common ground on preventing sexual transmission of
HIV. Lancet 2004; 364: 1913-1915.
(7) De Irala J, Alonso A. Changes in sexual behaviours to prevent HIV.
Lancet. 2006;368:1749-50.
(8) Cassell MM, Halperin DT, Shelton JD, Stanton D. Risk compensation: the
Achilles' heel of innovations in HIV prevention? BMJ 2006; 332: 605-7.
(9) Cabezon C, Vigil P, Rojas I, Leiva ME, Riquelme R, Aranda W, Garcia C.
Adolescent pregnancy prevention: An abstinence-centered randomized
controlled intervention in a Chilean public high school. J Adolesc Health.
2005;36:64-9.
(10) Stoneburner RL, Green T, Hearst N, McIlhaney J. Evidence that Demands
Action; Comparing Risk Avoidance and Risk Reduction Strategies for HIV
Prevention. In: Edited by Patricia Thickstun KH, editor: The Medical
Institute, 2004.
Competing interests:
None declared
Competing interests: No competing interests
To the Editor:
Re: Sexual abstinence only programmes to prevent HIV infection in
high income countries: systematic review, 4 August 2007, BMJ
2007;335:248-52.
Evidence-based and translational medicine--the idea that health
care delivery should incorporate best practices culled from medical
literature--is the current rage in the U.S. health care system. This
"bench to bedside" model operates on a basic principle: that
discoveries made in the scientific research arena should be quickly
applied to the everyday practice of medicine, allowing the benefits of
those findings to be felt as early as possible.
In theory, the United States' health care policy, whether domestic
or international in focus, also operates on this principle. When it comes
to U.S. efforts to address the global AIDS pandemic, however, we all too
often ignore science in the name of ideology.
The Bush administration has made considerable progress in the war on
AIDS through the President's Emergency Plan for AIDS Relief (PEPFAR),
domestic programs and support of the Global Fund. By requiring one-third
of all spending on AIDS prevention to go toward the promotion of
abstinence-only educational programming, however, the administration puts
valuable resources toward initiatives that are proven to be ineffective in
industrialized and developing countries alike, as indicated by the results
of the Underhill study.
President Bush should take a page out of our national health care
primer when developing future plans to combat AIDS, focusing on evidence-
based prevention programs (such as needle exhange programs and condom
distribution) that have a proven track record of success. For all the
current outcry on the state of American health care, relying on the
evidence is one example of where we're doing something right.
Ali Khan
Richmond, Va.
Third-year medical student, Virginia Commonwealth University School of
Medicine
Member, National Student Advisory Board, Physicians for Human Rights
Competing interests:
None declared
Competing interests: No competing interests
Sex education programme that teach abstenism, safe sex had very little impact on adolescence behavior & role of Microbicide in prevention of HIV in Rich countries
India is in the midst of an HIV and AIDS crisis with over 3.6 million
infected people up to August 2006 and 1,25,021 nos of full blown AIDS
cases(Male 82,335 Female 36,786) ,the highest burden in world after south
Africa[1]. The prevalence of HIV infection in India is heterogeneous, the
epidemic being mostly concentrated in southern states , while most of the
states has low rate of infection(<1% sero-prevalence). The states with
highest prevalence (>1% women presenting for antenatal check up when
found positive for HIV antibody ) are Mumbai, Tamilnadu, Maharastra,
Karnataka, West Bengal, Andhrapradesh, Manipur, & Nagaland. The
pattern of spread of HIV is also diverse & mostly heterosexual 86% in
some areas while IV drug users 2.6% in some others[2]. Heterosexual routes
of transmission of HIV is also increasing in rich countries and <10%
i.e 1 in 10 people is infected with HIV through this route. The number of
people with AIDS in UK also rose 144 in 1999 to 315 in 2003. Total nos of
AIDs in England, Wales, & northern Ireland between 1985-2003 is 21115
through heterosexual route[5]
So any attempt to reduce the heterosexual transmission of HIV
infection are highly one of the affective intervention that should be
widely implemented as measure for HIV/AIDs epidemic in a state/ country,
be it developed or developing or resource less. But the question remains
how? Reducing sexual transmission through change in sexual behavior of
population is mere a challenge and dream in modern society. Sexual
behavior is influenced by many factors and not always under an individual
control and depends on gender ness, social, cultural and economic
condition of an individual. Changing or maintaining of behavior results a
risk avoidance and risk reduction which must be cover stone of HIV
prevention. Abstainism, be faith full, reduce partner, use of condom
approach should play role in reducing the prevalence of HIV epidemic. All
these elements of the approach are essential to practice for reducing the
incidence at least theoretically
In the developing world( Indian scenario are different however) HIV rates
are higher in younger women then in men of same age and social, cultural
and economic factors compare the biological susceptibility to put women of
reproductive age group at particular risk. So targeting the young
population of abstainism or delay in sexual activity onset will emphasize
risk avoidance to prevent HIV & other STI as well as unwanted
pregnancy.
But what the practical experience tell us? Sex education programme
that teach abstainism ,safe sex had a very little impact on adolescence
behavior neither in India nor in US. Current UNIAIDS estimation of adult
HIV prevalence in India (0.5- 1.5%). Indian National Behavioral
surveillances (NBSS) reported extramarital sexual contact raised within
the previous years by 8.6% of men & 1.7% of married educated middle
aged women[3] Further a randomized community based study of adults in 28
chennai slums found that 2.9% of men and less then 1% of women reported
extramarital sexual contact on regular basis with 1% and 0.2% of HIV
infected respectively[4] 1% increase in AIDS prevalence in adults results
in an additional 5 million people infected with the virus.
Abstinence only sex education had been federally funded in the US
since 1999 and US congress did approve US $167 million for abstinence only
programme promotion for 2005 year. Abstinence education evaluation phase 5
technical report from Texas health dept also tell us that numbers of
adolescence who had sexual intercourse for more than once did not change
their behavior or numbers increased after they had received abstinence
only sex education[5]
For the developed countries or rich coutries without an effective
vaccine for HIV increasing attention should be paid to development of an
effective microbiocide to prevent the spread of HIV. Microbicides are
antimicrobials medications formulated for vaginal administration to
prevent the transmission of HIV. Ideally they should also protect other
sexually transmitted infections of STDS like gonorrhea, Chlamydia, HPV and
genetic ulcers Although no microbicide has yet been proven to block the
heterosexual transmission of HIV in humans but may be promising candidates
to prevent infection in the laboratory set up only and a dozen or so are
in advance stages of development. Several Microbiocide are today under
development 1) Those inactivate HIV by destabilizing or damaging viral
structure like SAVY-C31G -surfactants ,[surfactants ,disintegrates viral
lipid envelope] cyclodeytrins[alters envelope structure of virus] Buffer
gel and acid form lactobacillus suppositories[ Bacteria secrete bile acid
and H2O2] 2) Present HIV form dockers popery with target cells[ These are
various classes of lymphocytes that bear on their surface the virus
primary receptor CD4 and co-receptor CCR5 or CxCR4. These agents are
–carraguard, Pro 2000, Usher cell, anti HIV proteins Cynovirin N,
Monoclonal antibodies against component of docking mechanism 3) Anti
Retroviral drugs those prevents HIV from replicating inside the target
cells like Tenofovir(NRTI); TMC-120, UC-781; MIV-150(N-NRTIS) The
protective effectiveness and long term safety of these microbicides like
SAVY, Buffered Gel; Carraguard: Usher cell Pro-2000 are now being assessed
in large scale clinical trial in communities at high risk Of HIV
infection[6] Phase I and Phase II study on safety and acceptability of
Vaginal microbicide PRO-2000/5 was done for 14 days twice daily as intra-
vaginal microbicide (0.5% PRO2000gel) in HIV infected women in pune,
India. 42 eligible sero negative women ( 30 low risk and 12 in High risk)
enrolled this study and was given the gel for use. Local safety was
assessed via pelvic examination, colposcopy and laboratory monitoring and
systemic toxicity was judged based on whole blood count, liver, renal
function tests and plasma microbicide level. The Research findings
suggest that the gel is safe[7]34
Reference
1) Bhattacharya pranab, Bhattacharya Rupak, Bhattacharya Ritwik,
Bhattacharya Palash &Roy Arnab “HIV epidemic, &nature in West Bengal
State; India,: Role of male Circumcision Possible amongst Hindus as
additional preventive measure? Will state government start thinking of
It?” Rapid Responses published in BMJ for news byPeter Moszynski
Experts recommend circumcision to combat male HIV infections in Africa
BMJ 2007; 334: 712-b-713-b (10 April 2007)
2)http://www.nacoonline.org/facts htm
3) National AIDS control organization, Government of India, National
baseline general population behavioral surveillances survey 2001 htpp://
www naco online. Org/ publication htm
4) Srikrisnan AK, Sivram .S et al " The HIV epidemic in chennai 9 southern
India remains concentrated in high risk groups. Xv International AIDS
conference, Bangkok, Thailand, july11-16, 2004 MopeC3469
5) Hopkins. J BMJ.com news round up BMJ 330;12th feb 326;2005
6) Alan Stone, ShiboJiang “ Microbicides : Stopping HIv at the Gate- The
Lancet 368;August 5;431-33;2006
7) Annual Report of Indian Council Medical Research, New Delhi 2004-05
Edited by Sudha Chauhan DDG(SG) ICMR ; in Communicable disease HIV?AIDS
Page 75;2006
Authors-:1) Professor Pranab Kr Bhattacharya MD(Path) Cal, FIC path
(Ind.), Professor, Dept. of pathology, In charge of Histopathology Unit,
in charge of Cytogenetics, Ex-In charge of 24 hours Ronald Ross Malaria
clinic, Technical Supervisor In charge of Blood Bank, Institute of Post
Graduate Medical Education& Research (IPGMER)
244A AJC Bose Road, K0lkata-700020, West Bengal , India
Email= profpkb@yahoo.co.in
phone no- 91- 9231510435
2) Mr. Rupak Bhattacharya BSc(cal), MSc(JU)
7/51 Purbapalli Po= Sodepur ,Dist.- 24 parganas(North),West Bengal ,Pin
700110, India
Email= profpkb@yahoo.co.in
pranab@unipathos.com
3)Miss Upasana Bhattacharya
Delhi Public School, Rubipark, Kolkata
Email= profpkb@yahoo.co.in
pranab@unipathos.com
4) Dr. Srabani Chakraborty MD(cal) Path, Assistant Professor, Pathology,
Institute of Post Graduate Medical Education &Research, 244A AJC Bose
Road, kOlkata-700020, India
Email= banerjee.srabani@rediffmail
Telphone= 254111111
5)Dr. Palash Bhattacharya MBBS(Cal), MD(PGT) Pathology
Dept. Of Pathology
Institute of Post Graduate Medical Education& Research (IPGMER)
244A AJC Bose Road, K0lkata-700020, West Bengal , India
Email=
Phone no=91-3212243308
6) Dr. Pradip Dutta MD(cal) Medicine
Associate Professor, Dept of Medicine
Institute of Post Graduate Medical Education& Research (IPGMER)
244A AJC Bose Road, K0lkata-700020, West Bengal , India
7) Dr. Arnab Roy MBBS(cal)
Medical officer, HIV Clinic, Dept of Medicine
Institute of Post Graduate Medical Education& Research (IPGMER)
244A AJC Bose Road, K0lkata-700020, West Bengal , India
Email- arnray@yahoo.com
Telephone=9433133621
Competing interests:
None declared
Competing interests: No competing interests