Rapid Responses to:

PRIMARY CARE: Alba DiCenso, Gordon Guyatt, A Willan, and L Griffith Interventions to reduce unintended pregnancies among adolescents: systematic review of randomised controlled trials BMJ 2002; 324: 1426 [Abstract] [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

"Teenage Pregnancy Prevention Strategies: Why Should They Be "Effective"?

Donald B. Langille   (18 June 2002)

Expanding the Scope of Adolescent Health Education: Response to DiCenso et al.

Seth M. Noar   (26 June 2002)

Psychological Risk Reduction

Raul E Alessandri, Long Beach, NY 11561, USA   (27 June 2002)

De-mystifying Sex Education: Too little too late

Caroline Scherf   (18 July 2002)

Making sex education work.

Rebecca J Court, Siobhan H.M.Brown, Edward J Hannon and Alison L Wright.   (15 August 2002)

"Teenage Pregnancy Prevention Strategies: Why Should They Be "Effective"? 18 June 2002
Donald B. Langille, Associate Professor Department of Community Health & Epidemiology, Dalhousie University, Halifax, NS, Canada. B3H 4H7 Send response to journal: Re: "Teenage Pregnancy Prevention Strategies: Why Should They Be "Effective"?	In concluding that sexual health interventions are inadequate for bringing about delayed onset of intercourse and increased use of contraception in adolescents, DiCenso and colleagues(1) add to our understanding of the complexity of promoting sexual health in youth. However, they fail to discuss a fundamental issue their well conducted work should bring to mind. Why should we expect such interventions, which as described in Table 1 are largely educational in nature and of limited duration, (only 12 of the 26 interventions lasted a year or more), to be "effective" with respect to behaviour? Such interventions are very limited in their capacities to address the complex, multiple, and ongoing influences of parents, peers, health services providers, schools, socioeconomic conditions, religion and the media which shape the values, beliefs, and attitudes determining youth sexual risk-taking(2,3). How can we reasonably expect such limited interventions to overcome these influences? Though the authors identify the need to examine "social determinants of unintended adolescent pregnancy" as a future research area, discussion of what these determinants might be, and how difficult they are for largely educational interventions alone to address, is lacking. Further, the concept of "dose" in public health and educational interventions,(4) and the well recognized need to address multiple influences on youth sexual risk-taking,(5) are not mentioned. The danger in not discussing these limitations of the potential for intervention programs to impact behaviours lies in the possibility of providing ammunition to those who wish to discourage provision of sexual health information to young people altogether, based on a lack of program effectiveness at the levels examined. Though it is clear to me that this is not the authors' intent, it is also certain that some will interpret their study in this way. 1. DiCenso A, Guyatt G, Willan A, Griffith L. Interventions to reduce unintended pregnancies among adolescents: systematic review of randomised controlled trials. BMJ 2002;342:1429-1435. 2. Kirby D. Emerging answers: research findings on programs to reduce teen pregnancy (Summary). Washington, DC:National Campaign to Prevent Teen Pregnancy;2001. 3. Blum RW, Beuhring T, Shew ML, Bearinger LH, Steving RE, Resnick M. The effects of race/ethnicity, income and family structure on adolescent risk behaviors. Am J Pub Health 2000;90:1879-1884. 4. Fiore MC. US public health service clinical practice guideline:treating tobacco use and dependence. Respir Care 2000;45:1200-1262. 5. Department of Health. Social Exclusion Unit. Teenage pregnancy. London: The Department;1999.
Expanding the Scope of Adolescent Health Education: Response to DiCenso et al. 26 June 2002
Seth M. Noar, Senior Research Associate HIV Prevention Research, University of Kentucky, Lexington, Kentucky, 40506, USA Send response to journal: Re: Expanding the Scope of Adolescent Health Education: Response to DiCenso et al.	Dear Editor, Dicenso and colleagues provided a rigorous, systematic review of randomised, controlled trials to reduce unintended pregnancies in adolescents (1). However, what they did not provide, in my opinion, was a balanced interpretation of their findings. Namely, their assertion that sex education to date essentially has not worked is an inappropriate conclusion to make from their study findings. There are several reasons why I believe this to be the case. First, the criteria they used to select studies included in the review were extremely stringent. By including only randomised trials with certain characteristics, they have focused their study on a select number and type of trial. In so doing, they have excluded a number of well- designed, rigorously evaluated quasi-experimental trials (2). While the authors make the case that their selection criteria make the study better, one could argue that several successful, quality studies included in other reviews have been left out of this review. In fact, a similar review conducted in 2001 involving both randomised and quasi-experimental trials found evidence of behaviour change in several of the trials (2). It is important to point out that in school settings it is often difficult to conduct true randomised trials. And, when implemented and evaluated rigorously, quasi-experimental designs can be very powerful in nature (3), contrary to the authors’ suggestions. In addition, while the authors’ choose randomised trials as the highest standard of quality in terms of evaluation, from their data (Table 2) it is clear that even these trials had some flaws. And, what is not known is the true quality of the education provided. While many of these trials may have been well designed from a research design standpoint, it is not clear how many of them incorporated principles previously identified as crucial for successful sex education programmes (2). However, the crux of the issue here is not Dicenso and colleagues’ selection of studies or even their findings, but rather the interpretation of those findings. There is a clear inconsistency between the main research question in the study and the conclusions that were drawn from it. Since within 21 of 26 trials reviewed, or 81%, the ‘control’ condition was actually conventional sex education, this was not a study about whether sex education works or not. Rather, this was a study that compared the efficacy of theory-driven sex education with conventional sex education. What the authors found, contrary to many previous findings (2), is that theory-driven sex education did not outperform conventional sex education. Thus, the strongest conclusion that can be drawn from the study is that within this select group of studies, the two conditions come out equal in terms of behavioural outcomes. One could only have concluded that sex education does not work if it was compared to true control groups in which students received no sex education. It is troubling that the authors only point this out late in their paper as essentially an afterthought, and never clarify this issue completely. In addition, given the study’s findings, it would seem important for the authors to help answer an important question for which it appears they set out to answer: Why are some theory-driven sex education programmes successful while others are not? If theory-based education is no better than conventional, then what are the characteristics of successful sex education that results in positive behavioural changes? As the authors note, there was considerable heterogeneity among studies, and in some cases they could not explain this heterogeneity in relation to behavioural outcomes. Despite 10 hypotheses tested, they were not able to discern what separates a ‘good’ sex education programme from one that is weak. While we cannot hold the authors’ responsible for the findings of their study, they certainly could have better addressed these crucial issues. They did not, and it seems as a result that the field is no further ahead in understanding what the important features of a successful sex education programme are after this study as compared to before. Other such reviews have identified crucial characteristics of successful sex education programmes (2). Finally, here in the United States, and I suspect in other counties as well, sex education is generally not well integrated into classroom curricula. Despite the fact that schools are a promising place to deliver quality sex education (4), there are often political and social pressures on programmes that limit how much time is devoted to the topic, the quality of such education, and whether or not controversial topics including birth control, condom use, abortion, and the pleasurable aspects of sex can be discussed. This is the case both in conventional sex education and in theory-based trials, where various aspects of the curriculum must be negotiated with key stakeholders within school systems. What has been proposed by experts in the field (5), and what is clearly needed, are comprehensive programmes that address not only unintended pregnancies but also sexually transmitted infections including HIV/AIDS. Until we are able to move in the direction of comprehensive programmes, where sex education is not simply an add-on but rather a central component of health education, it will be difficult to have a large-scale impact on the sexual health of adolescents. In sum, my objection is certainly not to the rigorous, well-carried out study by DiCenso and colleagues (1). Such reviews can and often do help move the field forward. Rather, my objection is to these researchers drawing conclusions about sex education that were clearly not asked within the context of their study. Since nearly all of the adolescents in these trials received some type of sex education, we cannot know what the natural trajectory of behaviour would have been without such education. Thus, in my view, the strongest interpretation of these data is that theory-based programmes did not have more impact than conventional programmes, within the set of studies reviewed. While this runs counter too much literature and what many experts in this field believe and have found previously (2,4), it is the result of their review and one that demands both further interpretation and exploration. In addition, we do know from previous reviews (2) what the characteristics of good sex education programmes are, and at some level what separates ineffective from effective programmes, though we certainly have much to learn. This knowledge, coupled with policy changes that include comprehensive health programmes, can and should result in decreases in both unintended pregnancies and sexually transmitted infections. Seth M. Noar, Ph.D., Senior Research Associate HIV Prevention Research, Department of Communication, University of Kentucky, Lexington, Kentucky, 40506, USA 1. DiCenso A, Guyatt G, Willan A, Griffith L. Interventions to reduce unintended pregnancies among adolescents: systematic review of randomised controlled trials. BMJ 2002;324:1426-1434. 2. Kirby D. Emerging answers: research findings on programs to reduce teen pregnancy. Washington, DC: National Campaign to Prevent Teen Pregnancy; 2001. 3. Cook, TD, Campbell, DT, & Peracchio, L. Quasi experimentation. In: M D Dunnette & L M Hough, editors. Handbook of industrial and organizational psychology. 2nd ed. Palo Alto, CA: Consulting Psychologists Press; 1989. 4. Office of the Surgeon General. The Surgeon General’s call to action to promote sexual health and responsible sexual behavior. Rockville, MD: Office of the Surgeon General, 2001. 5. Rotherum-Borus, MJ, O’Keefe, Z, Kracker, R, Foo, H. Prevention of HIV among adolescents. Prev Science 2000;1:15-30.
Psychological Risk Reduction 27 June 2002
Raul E Alessandri, Attending Pathologist Long Beach Medical Center, Long Beach, Long Beach, NY 11561, USA Send response to journal: Re: Psychological Risk Reduction	DiCenso et al., by restricting their analysis to randomised control trials, exclude a large number of abstinence programs that do not use randomised controls out of ethical considerations, and because, in many cases, the efficacy depends on the involvement of the whole school, the community, the media and the families, in the attempt to influence the lifestyle of the children. This would preclude the use of control groups that would satisfy the critics. On the other side, the authors report of a successful program to reduce the use of drugs and alcohol, but seem to consider pregnancy and disease the only risks of early sexual activity. Interestingly, a recent report on a program to reduce drugs and alcohol, noted an unexpected result: a drop in sexual activity. The following abstinence programs, easily accessible through the Internet, have been reported to be highly successful: "Best Friends", "Project Reality", "Teen Aid", "Not me,Not now", "True Love Waits" and others. These programs have reported variously a 90% abstinence, coupled with a 1% pregnancy rate, compared to a rate of 80 to 90% for similar youths, and a national rate of 21.9%; a drop of 21% of sexually active teens; a drop in pregnancies of 12 to 13%; a drop of 19% in sexual activity (from 47 to 32%); and a drop in pregnancies from 63.4 to 49.5 per thousand girls. This experiences, not matched by any program of contraceptive sex education, cannot be dismissed by setting up barriers that practically preclude their proper consideration. During the period covered by these reports, MMWR and the surveys of the CDC on youth risky behaviour, as has been acknowledged in an "Occassional Report" of the A. Guttmacher Institute (1999), there has been a reversal in the curve of increasing pregnancies, in spite of the decrease in oral contraceptives and a minimal, non statistically significant increase in condoms. For more detailed information and references, there is a Interim Report submitted to the US Department of Health and Human Services, available on the Internet at: http://aspe.hhs.gov/hsp/abstinence02, and our article in the Linacre Quarterly 69:48, 2002 Raul Alessandri MD
De-mystifying Sex Education: Too little too late 18 July 2002
Caroline Scherf, Specialist Registrar in O&G Chlsea & Westminster Hospital, London SW10 9NH Send response to journal: Re: De-mystifying Sex Education: Too little too late	DiCenso et al’s report1 contains no useful information but instead supports calls for less sex-education, particularly if BMJ headline and summary are translated (as they are on contentious subjects like this) into non-medical news headlines. Disappointingly little critical comment was provided from the BMJ editors, implying their approval. This and the other article in the same issue about teacher delivered sex education in Scotland2 demonstrate the ineffectiveness of sex-education directed at young people aged 13 and above. The conclusion should be that the search for appropriate methods of sex education must be intensified rather than stopped – as implied by these two contributions. The US have a much higher teenage pregnancy rate than all other industrialised nations (52.1/1000 teenage women3). DiCenso et al may have concluded that perhaps one important reason for this sad record is related to the age at which children learn about sex as part of the secondary school curriculum in the US. Teenagers are wise and thoughtful and have a wealth of exposure to books, films, magazines and most importantly their own peers. These are the main sources of sex-education in the absence of this subject in the primary school curriculum. They have developed a skewed yet cemented view about sex and relationships which a teacher is very unlikely to be able to change. Clearly the hard work about teaching the facts of life needs to be transported to primary and even Kindergarten education and NOT stopped or substituted by ‘Abstinence campaigns’. When a child learns to read this new ability will provide him/her with tools to investigate and invariably encounter information he/she should be prepared for in order to critically deal with – some of which might be related to sex and relationships. This is not a new concept and in mainland Europe teenage pregnancy rates have been much more successfully reduced by starting sex-education at the Kindergarten level. Views of children as young as two about reproduction are being studied in great detail and incorporated in sex education strategies4. Here in the UK (highest teen pregnancy rate in Europe at 30.8/10003) new approaches have been sought and are being started, mainly targeting primary school children5,6 . Perhaps policy makers in the US realise that they urgently require a new approach to reducing teenage pregnancies and high-risk behaviour? Might that be withdrawal from the special treatment that sex education currently enjoys and instead consider firm introduction of this subject into the PRIMARY SCHOOL curriculum. Only when this subject is taught along with the ‘3 R’s’ will it loose its mysticism and children will be able to grow up into teenagers and adults capable of making properly informed, independent decisions about themselves. 1. DiCenso A, Guyatt G, Willan A, Griffith L. Interventions to reduce unintended pregnancies among adolescents: systematic review of randomised controlled trials. BMJ 2002;324:1426-1430 2. Wight D, Raab G, Henderson M, et al. Limits of teacher delivered sex education: interim behavioural outcomes from randomised trial. BMJ 2002; 324:1430-3 3. www.unicef-idcd.org, Innocenti Report Card No 3, July 2001 4. Forum Sexualaufklärung KINDER (1998), ed. H Lauer; Bundeszentrale für Gesundheitliche Aufklärung 51109 Köln 5. National Healthy Schools standard (2001) ‘Getting started’ Dept for Education and Employment 6. Spring Newsletter (2002) Healthier School Partnership, London SE1 9RY
Making sex education work. 15 August 2002
Rebecca J Court, Medical student University of Leeds, LS29JT, Siobhan H.M.Brown, Edward J Hannon and Alison L Wright. Send response to journal: Re: Making sex education work.	It was interesting to read the two articles published on the 15th June 2002 (vol. 324), which assessed the efficacy of sex education (Wight et al, DiCenso et al, 2002). We are a group of medical students at Leeds University, currently setting up an education programme in local schools which will involve medical students assisting with the Personal, Social and Health Education (PSHE) lessons. Wight et al and DiCenso et al both concluded that current sex education interventions were not effective in terms of reducing teenage pregnancy, delaying intercourse or increasing contraceptive use. Despite their conclusions we found their discussions optimistic. Their focus was not on dispensing with sex education, but on how interventions need to be refined and monitored in order that they can be more effective. This from our experience is certainly true. There were also certain limitations noted in both studies, with regards to the interventions and education programmes used, which may have significantly effected the results achieved. The papers reviewed by DiCenso et al and Wight et al simply evaluated the effect of a single intervention during a pupil's schooling. With the strength of societies pressures to be 'sexual' on young people today, it is naïve to believe that a single course of sex and relationship education, however well planned and delivered, will lead to changes in behaviour. Interventions need to be delivered over longer periods, starting earlier, and reflect the increasing amounts of influence society has on young people with age. It is interesting to note that Holland has the lowest teenage pregnancy rate in the World/Europe and they start assertiveness training at four years old (Doppenburg, 1993). Teaching style is of paramount importance to the delivery of information, especially on such a sensitive subject. Teachers have said that they find teaching sex education difficult and even embarrassing and seem more than happy to allow other people to fill this role (Jobanputra et al, 1999). There is a need for input from adolescents and a focus on communication and exploration of sexual relationships rather than scare tactics. Interactive rather than didactic methods help pupils personalise information. Small group work with pupils tends to be more effective than larger group work. Unfortunately teachers do not have the time or resources to make this a possibility, therefore outside influences would be of benefit to both pupil and teacher. The power of the peer group has long been recognised and during the past decade there has been a growing interest in involving young people as peer educators in sexual health education. In the UK programmes have been implemented involving medical students and older peers (e.g. sixth form pupils) as sex educators. Peer led sex education is effective in terms of increasing knowledge among students and pupil satisfaction (Jobanputra et al, 1999). Peer educators can have a huge influence on sexual behaviour. They can offer an opportunity to communicate on a meaningful level using language that is common to their age group, and they can be seen as 'trend setters' (wearing the right clothes etc). When these trend setters are giving out messages such as 'being a virgin is cool', teens are more likely to listen and adopt these behaviours (Smith et al, 2000). We are currently in the process of setting up a Leeds branch of 'Sexpression', an organisation set up by medical students across the country to deliver training and sex education (www.medsin.org/medsin.php?loc=~sex&msurl=sexpression). Sex education needs input from varied sources in order to be effective. While teachers and health care workers are successful at delivering factual information, peers can offer the chance for pupils to talk in small groups and address their worries with someone they feel at ease with. Sex education shouldn't just be about sex, it should teach pupils about the value of relationships and how to develop into a strong person with informed values and beliefs. · DiCenso A, Guyatt G, Griffith L and Willan A. Interventions to reduce unintended pregnancies among adolescents: systematic review of randomised controlled trials. British Medical Journal 2002; 324:1426. · Doppenberg H. Contraception and sexually transmitted diseases: what can be done? Experiences and thoughts from the Netherlands. British Journal of Family Planning 1993; 18:123-125. · Jobanputra J, Clack AM, Cheeseman GJ, Glazier A and Riley SC. Adolescent sex education: Medical students as peer educators in Edinburgh schools - A feasibility study. British Journal of Obstetrics and Gynaecology 1999; 106:887-891. · Smith MU and DiClemente RJ. STAND: A Peer Educator Training curriculum for Sexual Risk Reduction in the Rural South. Preventive Medicine 2000; 30:441-449. · Wight D, Raab GM, Henderson M, Abraham C, Buston K, Hart G and Scott S. Limits of teacher delivered sex education: interim behavioural outcomes from randomised trial. British Medical Journal 2002; 324:1430.