OUP user menu

The effect of the national demonstration project Healthy Respect on teenage sexual health behaviour

Janet S. Tucker, Ann E. Fitzmaurice, Mari Imamura, Suzanne Penfold, Gillian C. Penney, Edwin van Teijlingen, Janet Shucksmith, Kate L. Philip
DOI: http://dx.doi.org/10.1093/eurpub/ckl044 33-41 First published online: 6 April 2006


Background: As part of the independent evaluation of Healthy Respect (a national demonstration project to improve teenage sexual health in Scotland) this study examined the effect of the school-based sexual health education intervention comprising multiprofessional classroom delivery and alongside drop-in clinics on teenage sexual behaviour outcomes. Methods: Before-and-after cross-sectional surveys of secondary school pupils (average age 14 years and 6 months) were used in 10 Healthy Respect intervention schools in Lothian region and 5 comparison schools without intervention in Grampian region (2001 and 2003). Results: By 2003, the proportion of pupils in Lothian feeling confident about getting condoms and using condoms properly significantly increased, more Lothian pupils (particularly boys) showed improved knowledge about condoms being protective against sexually transmitted infections. No further evidence of improved knowledge, attitudes, or intentions was evident after the intervention. Pupils in Lothian remained more likely to think using a condom would be embarrassing (especially girls), would reduce sexual enjoyment (especially boys), and intentions about condom use (as closer predictors of actual behaviour change) showed no significant improvement. More Lothian (∼24%) than Grampian (∼19%) pupils report having had sexual intercourse at age <16 years, both before and after the intervention, with no evidence of a significant reduction in Lothian by 2003. Overall differences in attitudes to condom use by gender were noted. Findings remain consistent in both unadjusted and adjusted comparisons. Conclusion: These findings demonstrate limited impact on sexual health behaviour outcomes, and raise questions about the likely and achievable sexual health gains for teenagers from school-based interventions.

  • health promotion
  • sexual behaviour
  • teenage

Towards a Healthier Scotland, a national policy document, highlighted teenage sexual health as one priority area for population health improvement in 1999.1 This priority arose from evidence that pregnancy rates for 13- to 15-year-old people in Scotland had increased between 1983 and 1997,2 and that increasing proportions of young people reported having sexual intercourse aged <16 years.3,4 International comparisons, particularly with other western European countries, continued to show higher Scottish and UK rates of conceptions in older teenage groups (at 7–8% for 16- to 19-year-old people) and increasing rates of diagnosed sexually transmitted infections (STIs).57 Healthy Respect was established as a national health demonstration programme to work towards improving the sexual health of young people in Scotland by promoting a more informed and responsible approach to sexual matters, including STIs.1,8 The first phase of the demonstration project (2001–2004) was a partnership initiative bringing together health, education, and voluntary sector agencies' projects under the brand of Healthy Respect in Lothian Health Board region. The independent evaluation of Healthy Respect examined the success from the following three perspectives: process of implementation; comparative regional mapping of service development and inter-agency partnership working; and comparative impact using quasi-experimental design and adjusted, population-based behavioural and sexual health outcomes.9

This paper focuses on the independent evaluation of outcomes for pupils from Healthy Respect's Sexual Health And Relationships Education (SHARE) project as implemented in 10 volunteer schools in the Lothian region. Whilst recognising the continued need to ensure that young people are well informed, sexual health interventions should more closely focus on not just knowledge or attitudinal change, but also intentions and behavioural change.10 The study evaluates Healthy Respect's implementation of the theory-based SHARE programme,1113 previously tested in a cluster randomised controlled trial with little effect on primary outcomes of behaviour and no evidence of averting unplanned pregnancy to date.14,15 The re-launch of SHARE for Healthy Respect included revised teaching materials. It was envisaged that effectiveness would be further improved in a new implementation that required multidisciplinary staff training, planned multidisciplinary classroom delivery by teachers and nurses, and alongside access to sexual health services at drop-in centres for pupils.

The effect of the new Lothian Healthy Respect schools project was tested in a quasi-experimental study using before-and-after surveys of Lothian pupils compared with pupils from Grampian region who received standard non-SHARE sexual health education programmes. We aimed to test for improved outcomes for Lothian pupils by 2003, after the intervention. Outcome behaviours included knowledge, attitudes, and intentions related to STIs and condom use, and proportions of Lothian pupils reporting sexual intercourse at age <16 years.


Cross-sectional before-and-after questionnaire surveys took place in the autumn terms of 2001 and 2003 and included secondary school year 3 and year 4 pupils (median age 14 years and 6 months) in all 10 Healthy Respect SHARE schools in Lothian region, and in 5 comparison schools in the Grampian region with standard sexual health education programmes. The SHARE questionnaire was used in both survey rounds.16 A pilot study of 120 pupils in one school resulted in only minor revisions to layout.

Behavioural outcomes

Pre-specified outcomes of self-reported sexual intercourse at age <16 years, and knowledge, attitudes, and intentions about STIs and condom use were tested using selected items from the self-complete questionnaire (table 1).

View this table:
Table 1

Questionnaire items used in self-report of sexual health behaviour outcomes

Analysis and statistical power

Assuming that 25% of 15-year-old people had experienced sexual intercourse,3 the sample size required to detect an effect of 4% between regional samples and within Lothian region through time in a 2:1 sample ratio and at 80% power and 95% confidence is 2700 (Lothian) and 1350 (Grampian).

The impact and significance of the Healthy Respect SHARE intervention on improving outcomes for Lothian pupils from 2001 baseline is presented as the change in per cent responses to questionnaire items in Lothian compared to change in Grampian by 2003; i.e. the net regional percent difference for Lothian by 2003 (and 95% CI). Univariate logistic regression models were developed to explore gender effects controlling for region. To test for regional differences, multivariate models were also developed to adjust for any potentially confounding differences in socio-demographic characteristics between the regional samples. They were developed by entering gender and any socio-demographic variables that showed differences between the regional samples and that had significant one-way association with sexual health-dependent variables. Variables used in the model include gender, religion, parental education attainment, parental employment, housing type, and family structure (table 2). Results for regional differences are presented as odds ratios (OR) and 95% CI for 2001 and 2003, and the index category throughout is Grampian. To test for potential school effects, further post hoc multilevel modelling explored socio-demographic ‘pupil-level’ effects and ‘school-level’ effects1719 using the item on reported sexual intercourse at age <16 years. Analysis was undertaken using SPSS 12.0 and STATA 7.0.20,21

View this table:
Table 2

Socio-demographic characteristics of pupils by region [number (%)]

Sample schools

Lothian and Grampian are both relatively affluent east coast regions of Scotland, although both have variation in small area deprivation indicators. Grampian region is more rural than Lothian and has lower levels of deprivation overall.22

Ten Lothian schools that had agreed to undertake staff training and planned to implement the new Healthy Respect SHARE programme from January 2002 were selected for the evaluation. A comparison sample of Grampian (non-SHARE) schools were selected to participate seeking to match with individual Lothian SHARE schools using routine data about school size, rurality, and proportion of pupils with free school meals. All participating schools in both Lothian and Grampian were non-denominational state secondary schools, representing city and town settings (including smaller towns with rural catchments), and with varying levels of deprivation and school size. Approval for the study was sought from Directors of Local Education Authorities and head teachers. Before the survey pupils and parents were informed and offered the option to withdraw. The questionnaire was administered by trained survey staff under exam conditions in schools. The questionnaires were anonymous and confidential with allocated serial numbers to identify school and region only. The relevant Local Research Ethics Committees indicated that no ethical approval was required for this study.


Response rates

As part of the funded national demonstration project, all 10 selected Lothian Healthy Respect SHARE schools agreed to participate in the independent evaluation. Five of 17 invited Grampian schools agreed to take part. Reasons for non-participation of Grampian schools included possible closure of one school, or existing educational and other research survey commitments, but further key reasons given by school management were that the questionnaire topic of teenage sexual health was too sensitive and controversial. Of 5237 eligible pupils in 2001, 191 (3.6%) did not consent, and a further 722 (13.7%) were not in class during the survey, giving a response rate of 83% (4324) [80% (2760/3431) for Lothian and 87% (1564/1806) for Grampian]. In 2003, of 5193 pupils, the respective figures were 181 (3.4%) non-consent and 632 (11.9%) not in class and a response rate of 84% (4381) [83% (2798/3353) for Lothian and 86% (1583/1840) for Grampian].

Socio-demographic characteristics of samples

Regional samples were similar in terms of the gender, age composition, number of siblings, family type, and ethnicity in both time periods (table 2). There were also some significant differences between the regions. Compared to Grampian, Lothian pupils were less likely to live in an owner occupied house; more likely to have a Christian religious affiliation; less likely to have parents with higher educational attainment levels; less likely to have a mother in employment; and less likely to have a father in full-time employment (table 2). Although the absolute size of these differences between the regional samples could be small, these variables were included in developing multivariate models for adjustment.

Behavioural results

Table 3 presents descriptive statistics of pupil responses to the questionnaire items by gender, region, and time period.

View this table:
Table 3

Number (%) reporting intercourse at age <16 years, and positive responses in behavioural items related to condom use and STIs by region, gender, and round

Impact on reported sexual intercourse at age <16 years for Lothian pupils

In 2001, 952 (22%) of respondents reported having had sexual intercourse, and in 2003, 909 (21%) reported having had sex (table 3). More Lothian (24%) than Grampian pupils (19%) reported previous sexual intercourse with little change by 2003 for boys or girls. The net difference for Lothian by 2003 [−0.7% (95% CI −4.2 to 2.9)] was not significant (ns). After adjusting for socio-demographic characteristics, table 4 shows the higher OR for Lothian remain similar with no evidence of the regional gap narrowing [OR for Lothian in 2001 = 1.29 (1.10, 1.52), P = 0.002; OR for Lothian in 2003 = 1.35 (1.15, 1.60), P < 0.001].

View this table:
Table 4

Impact assessment on reports of sexual intercourse at age <16 years and behaviours related to condom use and STIs

Multilevel analysis was undertaken to explore school effects in this item. Reported previous sexual intercourse varied widely between all schools (from 14.1 to 40.6% in 2001, and 14.1 to 36.1% in 2003) and notably also between schools within regions. Of the total variation, ∼3% on the log odds scale was attributable to school-level effects in 2001 and 3.6% in 2003. After adjusting for all the pupil socio-demographic characteristics and then for variation between schools, 4.5% of the variation was explained by socio-demographic characteristics of pupils and a further 0.7% (ns) of the residual variation by school-level variance in 2001. In 2003, 3.3% was explained by socio-demographic characteristics and a further 2.1% by school-level variance. After taking account of socio-demographic pupil-level characteristics and school-level effects in the Grampian and Lothian samples, the odds of pupils reporting previous sexual intercourse remains higher in Lothian compared to Grampian in 2001 [1.27 (1.00–1.60) P = 0.049] and 2003 [1.31 (0.94–1.82), P = 0.11], although the effect in 2003 is no longer significant.

Impact on behaviour related to STIs and condom use for Lothian pupils

In 2001, Lothian pupils consistently demonstrated less knowledge, less positive attitudes, and intentions related to condom use compared with Grampian pupils. By 2003, there were significant increases for Lothian compared with Grampian in 3 of the 11 items related to condom use (table 3). The majority of pupils (∼70%) replied to two items indicating that they thought obtaining and using condoms was easy/very easy. In 2001, Lothian pupils were significantly less likely to consider it easy to obtain or use condoms properly compared to Grampian pupils. However, by 2003 Lothian pupils were more likely to be confident about getting and using condoms. The net difference in the regional changes in these attitudinal, self-efficacy items by 2003 in table 4 show a significant increase for Lothian [8.2% (95% CI 4.2–12.2)] and [9.7% (95% CI 5.6–13.9)], respectively. There is a significant increase for Lothian in one further knowledge/belief item. Although, overall only ∼40% of pupils agreed that they were more likely to contract an STI unless they used condoms (table 3), by 2003 the proportion of Lothian pupils agreeing that condom use reduces the chance of contracting STIs increased significantly compared with Grampian [4.7% (95% CI 0.4–9.1)] (table 4).

There were no significant differences in the remaining eight items about attitudes or intentions related to condom use and risk of STIs by 2003 for Lothian. Overall few (∼10%) pupils reported negative attitudes (of embarrassment or that use of condoms might reduce sexual satisfaction or were too expensive) that might be barriers to actual use. Indeed, by 2003 reported negative attitudes appeared to fall slightly in both regions. Overall around two-thirds of respondents intended to obtain their own condoms and discuss using condoms with their partners before having sex in both regions and at both time periods (tables 3 and 4).

Gender differences in behavioural responses related to STIs and condom use

From univariate analysis there are notable and significant differences by gender in the responses to the behavioural items about condom use and STIs, but no significant association between gender and rates of reported sexual intercourse (table 3). These gender differences and direction of effect are summarised in table 5.

View this table:
Table 5

Summary of gender differences and direction of effect in behavioural items related to condom use and STIs


Findings indicate that by 2003 Healthy Respect's SHARE intervention achieved an impact on some behavioural outcomes for Lothian pupils. There were increased proportions of pupils in Lothian feeling confident (self-efficacy) about getting condoms, and using condoms properly. Also more Lothian pupils (particularly boys) agreed by 2003 that STIs are likely to be contracted unless condoms are used. However, no further improvement on other attitudes and intentions were detected. Pupils in Lothian remained more likely to feel embarrassed (especially girls) and more likely to think condom use would reduce sexual enjoyment (especially boys). Lothian pupils' intentions about condom use (as closer predictors of actual behaviour change) showed no sign of improvement despite the above noted improvements in self-efficacy and knowledge. Similarly, there was no reduction by 2003 in the proportion of Lothian pupils who reported having had sexual intercourse.

One interpretation of these findings may be that, while SHARE was successful in providing information on the availability and the practical use of condoms, it appeared less successful in changing attitudes and intentions as precursors to changing behaviour, and in convincing pupils of the benefit of condom use as protection from sexual infections. These findings are in line with other reports that most school-based health promotion interventions tend to demonstrate greater impacts on practical knowledge, but have a limited effect on changing attitudes and behaviour.15,17,23 The age-specific levels of reported sexual intercourse at age <16 years are very much in line with previously reported national surveys in Scotland and the UK, given the age of respondents at 14 years and 6 months.4,5,24

It is worth noting that over half of the pupils in both regions were unsure or did not believe they were likely to contract an STI over the next 10 years if condoms are not used. Notably, around one-third were unsure or did not agree that condoms were effective in preventing HIV/AIDS in both regions and both time periods. The clear differences in knowledge, attitudes, and intentions about STIs and condom use by gender highlight some attitudes that may act as barriers to condom use and are reported here to better inform interventions and practice. Similar gender differences are previously reported by Thomson et al.25 Healthy Respect's intervention appeared to have had little impact on these gender gaps.

These results arise from a pragmatic, quasi-experimental study to test for effect of Lothian's Healthy Respect's SHARE intervention using predefined research questions to avoid type 1 errors. Limitations of the study include lack of classroom observation to explore the actual implementation of the new programme26 and possible selection bias arising from both volunteer schools in Lothian and low recruitment of schools in Grampian where only 5 of 17 comparison schools agreed to participate. Our response rates show low consent rate of schools and high consent rate of pupils, typical of previous reports of school-based sexual health education studies including trials.14,15,23 Reasons given by schools for not participating included competing demands of core-curriculum school activities, but notably some schools refused because the content of the questionnaire was perceived as too explicit and unacceptable to school management or to their communities. Reluctance of schools to participate in research and development of new programmes highlights remaining controversy and resistance to sexual health education and its underpinning values. We cannot exclude the possibility that selection bias in school recruitment might have contributed to the relative lack of effect observed. For example, if only schools that are confident (and possibly already more effective) in sexual health education volunteer to participate, demonstrating comparative effect of a new intervention in these settings may be more difficult. Similarly, with no classroom observation we cannot exclude the possibility that limited effect was due to limited implementation of the new Healthy Respect SHARE programme.

Finally, although the detailed socio-demographic characteristics of our Lothian pupil samples suggest less affluent backgrounds compared to those of the Grampian pupil samples, the adjusted results from multivariate analysis of differences by 2003 between the regions remain and are consistent with the measures of the net regional per cent change for Lothian by 2003. We conclude that these data provide evidence of limited impact attributable to Healthy Respect's SHARE intervention to date, and that is not explained by differences in the regional pupil-level socio-demographic characteristics. We note too that in subsequent sensitivity analyses, using multilevel models for school effects the size of attributable cluster school- effect also appears small.18,27

For policy-makers and practitioners these findings of one model of partnership working in a schools-based project raise questions about whether schools are the most appropriate venues for interventions aimed at improving young people's sexual health, and explicit valuation of the likely and achievable health behaviour gains from evidence. Although schools offer an opportunity to influence young people in a mass service delivery system, children move through in age cohorts and vary in emotional and physical maturity. Also it is clear from our findings that there are gender differences in behaviour relating to STIs and condoms. Finally, from the literature there are particular and vulnerable subgroups of the pupil population (e.g. those excluded or absent due to negative attitudes to school) where strong associations with negative sexual health outcomes have been demonstrated.28 Perhaps alternative approaches that do not treat young people as one homogeneous group but recognise different needs by gender, maturity, age, ethnicity, and social class might have greater success. Further focus and robust evaluation are required of alternative preventive approaches underway in the UK. These include wider multi-agency working in communities between health, social work, education, and voluntary organisations to provide innovative approaches for socially excluded young people.

Key points

  • A revised sexual health educational intervention for multidisciplinary classroom delivery by trained teachers and nurses plus alongside access to health services at drop-in centres was developed and implemented as part of a national health demonstration project.

  • Comparative results showed some attributable impact on improved knowledge and self-efficacy about obtaining and using condoms properly.

  • No further impact on attitudes or intentions to use condoms or in reduction of reported sexual intercourse at age <16 years were evident.

  • Findings demonstrate limited impact on sexual health behaviour outcomes, and raise questions about the likely and achievable sexual health gains for teenagers from school-based interventions.


This study was funded by the Scottish Executive Health Department, grant number CZH/4/11. We acknowledge the generous co-operation of schools, pupils, and members of the Healthy Respect project. Danny Wight and Marion Henderson gave permission to use the SHARE questionnaires and we thank them for their advice.


View Abstract