EFFECTS OF SEX RISK REDUCTION EDUCTION AMONG YOUTHS: A REVIEW
Sex education: An approach to reducing risky behaviours among youths
This literature review is based on studies that were conducted in Sub Saharan Africa between 1996-2004 which assessed the effectiveness of education in reducing sex related risks and sexually transmitted diseases including HIV/AIDS. The review is divided into two main themes focusing on interventions, which are a single component educational model and a multi component educational model(subdivided into social learning theory and behavioural learning theory), but also comprise an extra subtheme focusing on ethics.
Sexually Transmitted Infections (STIs) including HIV/AIDS have a destructive impact in developing countries which poses a real danger to the African population (Shakata2000). HIV/AIDS is the primary cause of diseases and has more dreadful impacts on the youths who are more vulnerable (Shakata 2000). Some of the factors that expose them include; unhealthy seeking behaviour, lack of knowledge due to cultural taboos, experimental sex, lack of communication skills and sometimes abuse of alcohol (Cowan, 2002). This calls for population based tailored preventative interventions that focus on reduction of risky behaviour and community development. Health education is one such intervention and is based on the assumption that health status of individuals or communities maybe influenced purposefully, although opinions differ about how such influence can be achieved (Kiger, 2004). Kiger (2004) stated that the present dominance of preventable diseases demands intervention i.e. health education. Hence the following studies implemented different approaches of health education.
Single Component Educational Model
Katz et al (2000) informs us that health education is a form of communication which aims to give people knowledge and skills needed to make choices about their health (which includes lifestyle and behaviour). Aderibige and Araoye in a cross sectional quasi-experimental study (n =512) among 10-19 years measured the impact of health education sessions that was based on lectures. During the course of these lectures participants were shown a film and educational material. Mbizvo et al conducted a study in rural and urban Zimbabwean communities within a 5 month period to assess the effect of a health intervention programme. The sample size was fairly large involving a random sample of 1689 secondary school children and there was low drop-out rate.
Although Mbizvo et al implied that an educational programme using the media model is effective in changing sexual risk behaviours, in his intervention group; there was no mention on how the education was applied, in which setting and whether it was done by experienced facilitators and what the communication materials in form of leaflets, posters and pamphlets contained to cover the programme details
The two study approaches imitate the media approach, which Kiger (2004) states that the media approach assumes that the health educator knows best about health and has a duty to persuade others to adopt values and practises which are known to be beneficial in combating disease. The study by Aderibige and Araoye did not mention in which setting (either rural or urban) it was done in. It is vital to know a setting in which a study was conducted as it informs aspects of applicability in other or similar settings. Moreover not all participants were accounted for. There is no mention if this is because some dropped out or there were other factors. Therefore this study limits to inform about the effectiveness or ineffectiveness of the intervention
Effect in improvement of reproductive health knowledge was noted with confidence intervals (OR=1.6, 95%CI=1.1-2.7) on pregnancy knowledge, (OR=1.4, 95%CI= 1.1-1.7) on contraception knowledge in the intervention group within the 5 months. Mbizvo et al also focused on assessing behaviour changes in relation to the intervention. But the results only measure of behaviour was the aspect of having a boyfriend or girlfriend, unlike Aderibige and Araoye who measured risky behaviour by number of sex partners one had, condom use and gifts for sex which gives a clearer picture of what risky behaviour is. There was also an increase in sexual activity among participants and also an increase in those who had a boyfriend or girlfriend, but Mbizvo et al do not reflect on these results (although small) to ascertain if the intervention had been effective in behavioural change. Data collection for these studies was done via questionnaires. This form of self reporting, although valuable in research, can introduce bias (Pilot and Beck 2008). Students can report to suit the objectives and this invalidates results.
Multiple Component Educational Models
a. Social Learning Theory
One of methods effective in changing sexual behaviour have been educational programmes that encompassed Social Learning Theory (SLT) ( Mellany at al 1995). The SLT was introduced by Bandura 1977, who concluded that an individual knowledge and skills acquisition is influenced by observing others through social interventions, experiences and media influences ( Ewles and Simnet). Klepp et al in a 2-3 month RCT conducted in Tanzania in 1992 on 249 primary school students assessed the effects of SLT education in reducing HIV risks and improve the tolerance and care of those suffering from HIV. The mean age of their population of the study was 13.4 years, who attended 20hours of the intervention.
In another study Bell et al in rural South Africa conducted a stratified RCT to assess the effectiveness of an intervention based on SLT. Their study was focused on black youths (n=557) aged 9-13, targeting risky behaviours but also strengthening family relationship processes as well as targeting peer influences. Although no confidence intervals were reported their study showed an increase in knowledge on HIV transmission(p=0.004) and less stigma towards HIV suffers (p=0.0005), however these results are not quite precise.
In the above studies these researchers in their intervention used lectures, role playing via plays/dramas and meetings with parents, community leaders and stakeholders. They hoped that people will learn through observing others' behaviours, attitudes, and outcomes of those behaviours (Keiger 2000). Most human behaviour is learned observationally from observing others, one forms an idea of how new behaviours are performed, and on later occasions this coded information serves as a guide for action.” (Bandura1977 cited by Naidoo and Wills). Observing alone although useful needs to be complemented by a sign of ones knowledge. Klepp et al also gave students to do their own research into the dangers of HIV/AIDS, protective behaviours, risk reduction and perceptions of society. These students wrote plays and songs and developed posters, and this gave them an opportunity to read more and show their understanding of what they had learnt.
The effective application of SLT in HIV/AIDS education leads to increased self efficacy. Self efficacy is fundamental as it leads to the development of knowledge and skill to make behavioural changes (Naidoo and Wills). Self efficacy contributes to the development of ones self esteem which contributes to one being more motivated towards adopting healthier ways of living (Ewles and Simnet). It was hoped that by adopting SLT , it would achieve what sex education should entail. Thus it should assist in boosting self esteem by equipping young people with relevant self efficacy skills (Naidoo and Wills).
Klept et al adopted a standardised tool to measure the outcome of the intervention using the World Health Organization's knowledge, attitudes, beliefs and practices instrument for adolescents, hence the results have a significant contribution. Although the attrition rate is 23.4%, its ability to influence the results are minimal, hence the results of this study have a significance. Despite no confidence intervals being reported ,this study reported knowledge from baseline had shown statistically significant increases on 12 month follow up with AIDS information (p=0.005) and AIDS knowledge (p=0.004). The limitation of this study is that they only measured having sexual intercourse as risky behaviour, which in doubt there is little effect. Nevertheless they accept this limitation by noting that they should have included condom use or number of sexual partners. There was also a high decrease in behavioural intention to engage in sexual intercourse (p=0.013) and initiation of sexual activity (p=0.006), which proves there was a behavioural effect of the intervention.
Bell et al used a standardised 10 component questionnaire ( eg The General health Questionaire, Global Indicator of Wellbeing,Child Problem Behaviour Checklist etc), to assess the effect of intervention, hence we can rely on the data.
Bell et al sample population was drawn from the majority black ethnic group, and hence there is a chance of replicating this study and intervention in other black communities, but not so sure how it can be replicated in a white community. Their study is also culturally relevant and builds on community strength for continuation post intervention. The mean age of 13.6 years might be young to be sexually active and hence assumed risky behaviour might be non existent. They do however accept the limitations of their tool in measuring sexual behaviour, and recommend that future educational interventions should address this issue.
The two studies by Bell et al and Klepp et al acknowledged that sex education needs to improve social networks ,and through this involved parents/guardians as well. This was meant to facilitate dialogue so that the platform to continuously address sex education post intervention was solid and
b.Behaviour Change Theory
Jewkes et al conducted a randomised control trial to determine the effectiveness of promoting safer sexual behaviour, by adopting a behavioural change theory approach in rural South Africa. This trial was conducted over a 2 year period with a total of 2776 participants with HIV sero- status and HSV 2 being established. The intervention was a 50 hour block running for six to eight weeks, facilitated by trained facilitators of the same sex and approximate same age. An important aspect was the pre testing of HIV and HSV 2 to use biological testing to assess the effectiveness of the intervention. Although they state it was a randomised control, participants were handpicked and thus it reveals potential bias and might limit genaralisability (Daly et al 1997). Annabel et al quasi experimental study of young people ( n=1544) conducted in Kenya over a 4 year period sought to assess the effectiveness of a culturally centred behavioural change intervention in reducing risky behaviour. The study comprised of 4-8weeks , interactive learning 90-120 minute sessions facilitated by a trained counsellor. In Zambia Agha and Rossem 2004 conducted a quasi experimental longitudinal study to assess the impact of a behavioural based intervention by peer educators in randomly selected secondary schools. Their intervention was a 1hour 45 minute single session to 481 students, with a reported 20% attrition rate.
Engaging in risky sexual behaviour has been documented as the major contributor of disease and unwanted pregnancy (Keiger 2000). It is thus vital that any form of sex/reproductive education should focus on ways to either change risky behaviour or solidify current positive behavioural practices. One way of doing that is implementing a behavioural changing model because it is known to encourage participants to change aspects of risky behaviour that is detrimental to their health (Naidoo and Wills). Kamali et al in after their RCT study recommended that effective behavioural interventions are needed to combat HIV/AIDS in mature epidemics like Sub-Saharan Africa.
The interventions by the above researchers encompassed several behavioural theories including education and reflection model. Affective (emotional) and cognitive (intellectual) are the two main components that constitute this model (Kiger 2000). Most of the interventions employed a variety of methods directed at skill development in decision making and clarification of health values, conducted in group work . This approach is essential in motivating individuals and assist in the changing of attitudes and behaviour ( Kiger 2000).
Reflection is an affective and intellectual activity whereby individuals engage to explore their experiences in order to lead to new understandings and appreciations, allowing them to make reasoned decisions or course of action ( Bulman an Shultz). This was an important element of Jewes et al study, as it diluted well with other aspects of behavioural change in the hope that in conjuction they will be all effective in changing behaviour. Though there is reported effectiveness, there was no evidence that their intervention lowered the incidence of HIV, but there was an associated reduction of 33% ( incidence ratio 0.67, 95% confidence interval 0.46-097, P=0.036) in HSV-2. Their results show no evidence of behaviour change in women ( although they attribute this to poor self reporting), but some changes in men( e.g. reduction of sexual partners and transactional sex). However there is an approximately 3-4% attrition rate to those who were tested for HIV in the study period, but because these values are relatively low there is a limited chance that it could affect the results, hence these results can be relied on. .The limitation to this biological assessment is that they did not include the testing of other sexually transmitted disease to evaluate other significant changes in behaviour. Jewes et al study was based on Welbourn programme, but they did not evaluate it the way it was intended, this could lead to limitations in generalising this study. Sample size used could have also influenced the generalisability due to the researchers being too optimistic in their calculations.
Annabel et al study had a response rate of 87% baseline and 90% post intervention, this shows a very low attrition rate and these results can be relied upon. In their intervention they adopted the behavioural theory, but delivered it in a culturally based way to reflect the local community. The results show an improvement in sexual behaviour, 10% and 5% decline in young men who had initiated sex and those that had 3 partners or more respectively in the last 3 years. Although the decline was different than mens, women also showed a decline of 3% and 9% initiating sex and having more than 3 partners respectively. However women reported a10% increase compared to men with 6% increase in condom use. Nevertheless there was a noted change in behaviour of the intervention group, but the limitation is the aspect of self reporting that might have introduced bias.
Results from Agha and Rossem 2004 showed that there was an increase in behaviour after six months and on second follow up as well. There was also an increase in knowledge with the following results; Heard of abstinence via peer educators OAR=2.33,95% CI 0.89-6.09, approval to abstaining OAR=1.60, 95% CI 0.99-260 and abstinence is effective in preventing STIs OAR=184, 95% CI 0.85-3.97 only to highlight some of the factors measured. Limitations of this study is the lack of randomising students, hence multiple factors might have influenced these results. The students also self reported and the two factors might introduce bias, and hence the results can not be generalised.
Research programmes should aim to make ethical considerations during their design and implementation (Crosby et al). Ethics thus is a fundamental component of research practice. Although many developed nations tend to have ethical approval boards, there is a limited scope of knowledge about their existence and standards in the developing countries i.e. Sub-Saharan Africa. Ethics in research is broad and covers an assortment of issues, but Marzcyk et al point out that prior to research participants competence to consent to research should be assessed based on several factors including age. Research conducted by Klept et al, Mbizvo et al and had a focus on participants aged 13-16 years, but in their reports there was no mention of any consent being gained. 13-16 year olds are usually taken as children, hence this age group has diminished ability to give informed consent (Marzyk et al). Before consenting participants needs to be aware of possible benefits of participation and dangers in delayed participation (Sim et al). Such information should have been given to participants prior to the research from an ethical point. Given that most of the researchers knowing their participants age sought consent from parents/guardians and the participants themselves. Not only did they seek consent, they also sought ethical approval from recognised ethics boards within the respective countries.
Sims et al highlight the need to shun coercing participants for them to be part of the research. They state that participants should voluntarily agree to research. Understanding that economic hardships are an issue in Sub Saharan Africa, is the use of incentives ethically justified? Jewes et al and Bell et al in their research in rural South Africa offered monetary incentives for those who participated. This is an element of coercion and its ethical justification needs further reviewing.
Although an assortment of educational models has been used, it is vital to note that all researchers recommend further review of long-term effects of interventions. The question now is of all these interventions which one is most effective. Although the media model used in the above studies have been criticized by some health educators on grounds that its short lived and motivation springing from emotional responses may not be maintained once the stimulus has been removed (Kiger 2004), future researchers need do a thorough evaluation of evidence that proves the efficiency, applicability and long term impact of the other models. The use of piloting can be starting point to give researchers a better idea on expected impact and feasibility, hence its use in further researches is recommended.