Effectiveness of a School-Based AIDS Education Program among Rural Students in HIV High Epidemic Area of China

Original article

Effectiveness of a School-Based AIDS Education Program among

Rural Students in HIV High Epidemic Area of China

Yan Cheng, Ph.D.

, Chao-Hua Lou, M.D.

, Lisa M. Mueller, M.A.

,

Shuang-Ling Zhao, Ph.D.

, Jian-Hua Yang, Ph.D.

, Xiao-Wen Tu, Ph.D.

,

and Er-Sheng Gao, M.D.

*

a_{The Department of Epidemiology and Social Science on Reproductive Health, Shanghai Institute of Planned Parenthood Research, Fudan University,} Shanghai, People’s Republic of China

b_{The Department of Epidemiology and Social Science on Reproductive Health, Shanghai Institute of Planned Parenthood Research, Shanghai,} People’s Republic of China

c_{PATH, Washington, DC}

Manuscript received November 15, 2006; manuscript accepted July 19, 2007

Abstract Purpose: To evaluate the feasibility and effectiveness of a life-planning skills training program using participatory methods among rural senior high school students in Shangcai County, Henan Province, China.

Methods: The study was a quasi-experimental study conducted in three Shangcai County senior high schools with comparable socioculture– economic and demographic characteristics (two interventions and one control). The intervention, a life-planning skills program that uses participatory training methods, combining information education with effective skills building, was provided to all first-grade students (14 –18 years old; 87% of them are between 15 and 17 years old) in the intervention group from October 2003 to December 2003. In total, 717 students from the intervention group, and 457 from the control enrolled at baseline, and over 91% of these were followed up at posttest.

Results: Group ⫻ time interaction effects in ordinal logistic regression analysis were found on HIV/AIDS-related knowledge (p ⬍ .0001), attitudes toward daily contact with HIV-positive persons (p ⬍ .0001), and subjects’ protection self-efficacy (p ⬍ .0001), suggesting the intervention increased subjects’ knowledge significantly, changed their attitudes positively, and improved their protection self-efficacy. The intervention also significantly improved subjects’ communication with teachers and peers on HIV/AIDS issues (p ⬍ .0001). However, no significant change was observed on respondents’ attitudes toward premarital sex or their communication with parents between the two surveys (p ⬎ .05).

Conclusions: Three months of short-term HIV/AIDS education through life-planning skills train-ing was welcomed by students and positively influenced HIV/AIDS-related knowledge, attitudes, protection self-efficacy, and communication among senior high school students in a rural area with high HIV prevalence. © 2008 Society for Adolescent Medicine. All rights reserved.

Keywords: Adolescents; HIV/AIDS education; Life-planning skills training; Participatory methods; China; HIV prevention; School-based programs

In the mid-1990s, people living in Shangcai County, Henan Province, which is located in the middle of China, witnessed increases in HIV and AIDS. HIV was spread among poor farmers through unhygienic blood plasma col-lection, and later spread to families through sexual inter-course and transmission from mothers to babies. By 2002, over 10,000 of the county’s population of 1,300,000 people

*Address correspondence to: Er-Sheng Gao, M.D., Department of Epidemiology and Social Science on Reproductive Health, Shanghai In-stitute of Planned Parenthood Research, 2140 Xie Tu Road, Shanghai 200032, P.R. China.

E-mail address: [email protected]

1054-139X/08/$ – see front matter © 2008 Society for Adolescent Medicine. All rights reserved. doi:10.1016/j.jadohealth.2007.07.016

had been diagnosed as HIV positive, making it known as an area of high HIV prevalence in China[1,2]. As the county began to see a large outbreak of HIV and AIDS cases, rumor and stigma became widespread, promoting fear, indiffer-ence, and an unstable social atmosphere. In addition, insuf-ficient knowledge about HIV prevention also exposed local residents, especially vulnerable youth, to risk behaviors. The government recognized this situation and included HIV/AIDS propaganda and education, as well as reducing unsafe behavior among high-risk populations (sex workers and injecting drug users) as key components of its HIV/ AIDS prevention strategy.

Teenagers are an important part of society, and they will be an important force against HIV transmission in the near future. In addition, their opinions, attitudes, and behaviors play a critical role in constructing a compassionate social environment free from discrimination for people living with HIV and AIDS[3]. Evaluations of school-based HIV pre-vention programs have shown them to be one of the most important sources of education interventions [4 –7]. Thus, school-based education was considered an important effort for HIV/AIDS prevention activities. In Shangcai County, adolescents aged 10 –24 constitute about one-fourth of the population. Because most of them join the workforce after graduating from senior high school, it is even more impor-tant for them to receive HIV prevention education while still in school. Before the program, a few school-based HIV prevention activities in China improved participants’ AIDS knowledge and their attitudes toward HIV-infected persons; however, they were mostly conducted in traditional, passive methods (lecture and materials dissemination) that do not hold students’ attention, and lacked skills development [8,9]. Emerging computer- and video-assisted education mod-els are rare due to limited resources. Thus, the introduction of low-cost, practical, participatory methods for sexual reproduc-tive health (SRH) education is vital, especially in China, where rural adolescents account for about two-thirds of the total population.

Participatory life-planning skills (LPS) training, in-formed by the classic cognitive/social learning theoretical framework, has been widely adopted and achieved success in the United States, Africa, England, and Australia on decreasing adolescents’ risk-taking behaviors [10 –14]. In this program, it was introduced in the field of sexual edu-cation through the curriculum “Path to Growth,” developed by the China Family Planning Association and PATH. The training manual used materials fromLife Planning Skills: A Curriculum for Young People, developed by PATH, ado-lescent training materials by the Hong Kong Family Plan-ning Association, the Youth Peer Education Manual for HIV/AIDS Prevention, developed by the Yunnan/Austra-lian Red Cross HIV/AIDS Care and Prevention Project, and the HIV/AIDS Prevention Education Teacher’s Guide, de-veloped by the United Nations Children’s Fund. The cur-riculum combines sexual education with effective skills

building (communication, decision making, and self-protection) while preparing youth for the future. A variety of participatory methods are used to create an interactive atmosphere[15–17]. Similar curriculum also achieved suc-cess in some African countries[18].

In China, some pilot studies adopting LPS training pro-grams to improve adolescents’ reproductive health knowl-edge and attitudes have proved successful[5,17]. Based on previous research, the curriculum was further edited to meet the situation of China and was introduced in more sites including Shangcai County. This study looked at an LPS program in a rural area with a high rate of HIV prevalence and evaluated its effectiveness on students’ HIV/AIDS knowledge, attitudes, protection self-efficacy, and commu-nication behaviors, which is different from previous studies in two aspects. First, it is the first time to evaluate the impact of LPS training among students from an HIV high epidemic rural area. Second, the study examined the change of pro-tection self-efficacy and communication behaviors, which was rarely studied in previous evaluations in China. In this paper we only consider the short-term impact of the 3-month intervention, that is, HIV-related knowledge and attitudes, because there was not enough time to observe behavior change.

Methods

Study design

It is a quasi-experimental study. Three senior high schools in Shangcai County, Henan Province, China, were selected as research sites, with two as the intervention group and one as the control. The selection criteria of the inter-vention schools included: low student drop-out rates and support from school authorities for the intervention pro-gram. Two out of 10 local senior high schools met these criteria, so both of them were included in the study as the intervention sites. The control school was chosen on the basis of its similarity to the intervention schools on the educa-tional level, academic performance, and family background. Also, it was far enough to avoid cross-contamination. All the first-grade students (14 –18 years old; 87% of them between 15 and 17 years old) in selected schools were recruited as study participants. Youth and their parents were informed of the study purpose before the study. Parental consent was directly sought in the parent–teacher meeting that was held before the baseline survey, and no parents disagreed. Youth consent was sought in written form in the classroom before the survey, and all of them agreed to participate. School supplies were given to those who com-pleted the questionnaire in gratitude for their participation. PATH’s human subject’s protection committee approved the research protocol.

The baseline survey was conducted in September 2003 before the intervention began, and a similar survey was

conducted with both groups after 3 months’ intervention. One thousand one hundred seventy-four students (96.6% of the first-grade students) completed the baseline survey. A total of 93.8% were successfully followed up at posttest, including 91.9% from the intervention group, and 96.5% from the control group. Efforts were made to ensure ano-nymity and privacy while surveys were conducted via anonymous self-administered questionnaires. Trained inter-viewers assisted respondents, when necessary, to understand questions. All completed questionnaires were reviewed by research staff for completeness and consistency. To better understand the effects of intervention and find the influenc-ing factors, focus group discussions (FGDs) was conducted among teachers, school leaders, and parents in the interven-tion school.

Intervention

The LPS training curriculum covered nine topics: ado-lescent development, sexually transmitted infection (STI) prevention, HIV prevention, relationships and communica-tion, values clarification and decision making, reproduction and contraception, sexuality, planning for the future, and drug abuse prevention. Among them, “HIV prevention,” “values clarification,” and “decision making” focused on HIV transmission and prevention, facts and myths sur-rounding HIV, and how to act as the force against HIV/ AIDS, whereas the benefits of waiting until marriage to have sex, how to resist peer pressure, and make decision about love and sex were taught in “decision making” and “sexuality.” Each topic lasted 90 minutes, and all of LPS sessions were completed in 3 months (October 2003– December 2003). A lot of participatory activities were used, such as group discussion (to facilitate the communi-cation and information sharing), role playing (to create hypothetical scenario for participants to practice skills), brain storm (to encourage prompt responses from partici-pants and create an active atmosphere), case study (list actual cases and discuss), and games (make students learn-ing from playlearn-ing). All of these allow young people to learn by active doing rather than passive listening. During the process, skills including decision making, communication, and self-protection were conveyed to participants. The LPS training activities were conducted by 28 facilitators. All of them were teachers chosen because they had superior com-munication skills and were welcomed by students. Most of them had AIDS education background. Before the interven-tion, all of the facilitators received 5-day facilitator training on both the contents of the curriculum and how to conduct the participatory methods. To make sure the training courses were effectively carried out, school cluster workshops were organized by the county-based family planning association (FPA) staff in each intervention school. Facilitators were asked to practice LPS sessions in the workshop and the team(s) adjusted course contents to reflect local

circum-stances as appropriate. For those who lacked experience in handling such curriculum, retraining of facilitators was adopted to improve the intervention quality. No sexual health education or training was given to control group students.

Prior to the implementation of the LPS program in schools, local FPA staff worked with the local education sector to build a supportive environment for the program. Staff met with key school leaders, held discussions with parents, and publicized the need for SRH education for adolescents via meetings and mass media campaigns. The selection of control group and administering the question-naire was also conducted with the cooperation of local education sector.

Measures

There were five main measures (i.e., 1, knowledge score; 2, attitude score; 3, protection self-efficacy score; 4, proportion having open attitude toward premarital sex; and 5, proportion of communication with people on HIV/AIDS issues):

1. To evaluate participant’s knowledge level, there were 33 questions related to HIV/AIDS knowledge in the questionnaire. Correct answers were credited with a score of one and incorrect answers with a score of zero. Then the sum of each question’s score was converted into a new score, with the maximum of 100. The higher the score, the greater the knowledge. 2. The questionnaire included five statements regarding whether or not respondent was willing to have daily contact with people living with HIV or AIDS (eat together, study together, receive food/hair service, go to their home, and share telephone). For each state-ment, one point was given if answered willing and zero to unwilling, then points were added up to get the sum. The sum was also converted into a new score, with the maximum of 100. The higher the sum, the more friendly the attitude towards people living with HIV or AIDS.

3. There were three statements related to protection self-efficacy in the questionnaire. Answers were scored according to the score sheet below (Table 1) then added up to get the sum. The sum was also converted Table 1

Score sheet for self-efficacy issues

Statement Points

Agree Do not know Disagree a. If I do not want to have sex at this

moment, I am able to refuse it 3 2 1 b. If I were sexually harassed, I

know how to protect myself 3 2 1

c. If I have sex, I will take some protective measures to prevent

into a new score, with the maximum of 100. The higher the sum, the higher the level of protection self-efficacy.

4. The statement “sex before marriage is OK if they are really in love” was used to evaluate students’ attitudes toward premarital sex. Answers included agree, dis-agree, or don’t know.

5. The questionnaire included three questions on com-munication about HIV/AIDS issues: whether re-spondents had discussed HIV/AIDS issues with parents, peers, and teachers. Those who responded “yes” were considered to have had the behavior of communication.

Statistical analysis

The difference between the intervention and control groups in numerical variables (nonnormal distribution) and in categorical variables was examined with nonparameter test (Wilcoxon–Mann–Whitney tests) and chi-square test, respectively. The effects of the intervention on knowledge, attitudes, and protection self-efficacy were analyzed using ordinal logistic regression models (the knowledge, attitudes, and protection self-efficacy were ordered as four categories by quartiles) controlling for other possible factors. The effect of the intervention on attitudes toward premarital sex was analyzed using multinomial logistic regression model (agree⫽1, disagree⫽2, don’t know⫽3). The effect of the intervention on communication of HIV/AIDS issues with parents, peers, and teachers was analyzed using binary lo-gistic regression models (whether communicated with these people as dependent variables; yes⫽1, no⫽0). Data were entered twice with EpiInfo 6.04 software, and then validated until two inputs were completely the same. They were analyzed with SAS 8.01.

Results

Profile of study participants

At baseline, respondent’s average age was (16.22 ⫾ 0.89) years old (age varied between 14 and 18). The ma-jority of students came from middle-income rural families. Most of the students’ parents were farmers with educational levels of middle school and below. About 20% of respon-dents reported knowing that people living with HIV lived nearby.Table 2shows the sociodemographic characteristics of the two groups. No significant differences were observed in residence registration, family type, family economic sta-tus, parents’ occupation, and mothers’ educational level between groups. However, students from the control group tended to be slightly older, and their father was more likely to have a higher educational level than those in the inter-vention group. Gender disparities were also observed. These factors were then controlled for in the analysis.

Exposure and evaluation to intervention activities

Among 771 intervention group students who participated at posttest, 95% reported they had received the LPS pro-gram, the average number of training sessions students received was 8. 76% reported that the sessions were very useful, and 67% evaluated the facilitators as qualified. Of those who had participated in LPS sessions (95%), 44% reported sharing what they had learned with their parents, relatives, and neighbors. At posttest, 93% of the interven-tion group students suggested that LPS training should be carried out among all middle school students in Shangcai County. LPS education also contributed to the change of teachers’ and parents’ attitudes toward sex education. In the focus group discussion, they praised this type of method as “interesting, new, and meaningful,” “making sex education less embarrassing and facilitating relationships between stu-dents and parents and teachers,” and as “The course can tell students how to be more self-assured, how to plan their future, which must affect their future lives.”

Effectiveness of intervention Knowledge

Table 3shows that the median scores for HIV/AIDS knowl-edge in the intervention and control groups at baseline were about 50.0 points with no significant differences between two groups (p⬎.05). However, at posttest, the intervention group scores increased to 78.8, which is significantly higher than the control group scores (57.6). The increase of the score from baseline to posttest in the intervention group (27.3) was nine times that in the control group (3.1). Those who answered “no” to the common misconception about whether insect bites trans-mit HIV was significantly higher in the intervention group (85.5%) than that the control (27.6%) at posttest, whereas no significant differences were observed between groups at base-line (about 23%).

A group ⫻ time interaction effect was found on HIV/ AIDS knowledge scores in ordinal logistic regression anal-ysis, providing evidence that the difference between the intervention and control groups varied with the implemen-tation of the intervention. Table 4 reveals the increase in knowledge scores is attributable to the intervention because the scores were comparable at baseline while different at posttest. We can therefore state that the intervention activ-ities increased youth HIV/AIDS knowledge significantly (Table 4, Model I).

Attitude toward people living with HIV/AIDS

No significant differences of respondents’ attitudes toward people living with HIV and AIDS were found between the two groups at baseline (p ⬎ .05). Approximately 42% reported they were not willing to do any daily activities with people infected with HIV such as eating together, studying together, receiving services from them, going to their homes, or using the same telephone, indicating high stigma levels. The

propor-tion decreased in both the control and intervenpropor-tion groups after the intervention period, but the decrease in the intervention group (29.1%) was significantly higher than that in the control group (10.2%) (p⬍.0001) (Table 3).

Based on respondents’ attitudes score and using the same ordering method as with knowledge, ordinal logistic regres-sion analysis was used to evaluate the effectiveness of the intervention on respondents’ attitudes toward daily contact with people living with HIV and AIDS. A group ⫻ time

interaction effect on attitudes was found (p⬍.0001), which indicated that the intervention might change subjects’ atti-tudes and made their attiatti-tudes toward people living with HIV and AIDS more friendly (Table 4, Model II). Protection self-efficacy

At baseline, there were 69.1%, 69.0%, and 47.8% sub-jects who reported they could refuse unwanted sex, protect themselves from sexual harassment, and have safe sex, Table 2

Sociodemographic profile of participants at the baseline: intervention and control groups (%)

Characteristics Control group Intervention group pa

(inter.-contr.) Total (n⫽457) Male (n⫽257) Female (n⫽200) pa _Total (n⫽717) Male (n⫽425) Female (n⫽292) pa Age (years) Under 16 34.1 28.8 41.0 .0015 44.4 40.2 50.3 .0035 .0009 16–17 46.2 46.3 46.0 41.6 42.6 40.1 Over 17 19.7 24.9 13.0 14.0 17.2 9.6 Residence registration Urban 11.8 11.7 12.0 .9919 9.8 10.2 9.3 .7083 .3931 Township 9.6 9.7 9.5 8.4 9.0 7.5 Rural 78.6 78.6 78.5 81.8 80.8 83.2 Type of familyb Large family 37.4 35.8 39.5 .4301 30.9 31.8 29.8 .8544 .0607 Core family 61.5 62.6 60.0 67.4 66.6 68.5 Single parent 1.1 1.6 0.5 1.7 1.6 1.7 Economic statusc Well off 22.7 21.6 24.0 .3907 23.3 25.8 19.7 .4575 .2051 About average 54.2 53.9 54.7 58.2 54.7 63.1 Below average 23.1 24.5 21.3 18.5 19.5 17.2 Father’s occupation Farmer 40.6 40.5 40.9 .8954 40.9 40.6 41.3 .2909 .8795 Worker/self-employed 23.0 22.3 23.9 24.5 25.1 23.7 Businessperson 19.5 20.0 18.9 18.8 17.5 20.6 Teacher/governmental/professional 14.4 14.9 13.4 13.6 14.9 11.8 Othersd _{2.5 2.3 2.9 2.2 1.9 2.6} Mother’s occupation Farmer 59.4 59.1 60.0 .8921 60.2 59.5 61.2 .1915 .7888 Worker/self-employed 11.3 11.2 11.4 12.1 13.8 9.7 Businessperson 15.7 17.0 13.8 15.1 14.1 16.7 Teacher/governmental/professional 8.5 7.9 9.3 8.5 8.6 8.2 Othersd _{5.1 4.8 5.5 4.1 4.0 4.2}

Father’s education statusc

Illiterate/primary 9.7 10.7 8.5 .8511 10.8 9.8 12.2 .6061 .0340

Junior secondary 42.8 41.9 43.9 46.7 49.0 43.4

Senior secondary 41.5 40.2 43.1 38.2 37.1 39.9

College or above 5.3 6.8 3.4 3.4 3.5 3.2

Unknown 0.7 0.4 1.1 0.9 0.6 1.3

Mother’s education statusc

Illiterate/primary 35.6 40.5 29.1 .4984 37.5 41.0 32.4 .0985 .1597

Junior secondary 37.8 32.9 44.4 37.2 34.1 41.7

Senior secondary 23.3 22.0 24.9 23.0 22.6 23.7

College or above 1.2 1.7 0.5 1.1 0.9 1.3

Unknown 2.1 2.9 1.1 1.2 1.4 0.9

a_␹2_{test between males and females or between the intervention and control groups; nonsequential variables tested by Pearson chi-square; sequential} variables tested by CMH chi-square.

b_{Large family means a big family including children, parents, grandfather and/or grandmother, and all the family members live together. Core family} means a family only including children and parents. Single parent means a family only including children and mother or father.

c_{Sequential variables.}

respectively, and no significant differences were observed between the two groups (p ⬎ .05). The proportions in-creased in the intervention group, whereas they dein-creased in the control group at posttest, to 74.1%, 75.4%, and 56.3% in the intervention group, respectively, while being 61.6%, 59.6%, and 34.9% in the control, respectively. The differ-ences between the two groups were statistically significant at posttest (p⬍ .0001) (Table 3).

Based on subjects’ protection self-efficacy score and using

the same ordering method as with knowledge, a group⫻time interaction effect was found on subjects’ protection self-efficacy (odds ratio [OR]⫽2.371, 95% confidence interval [CI]: 1.711–3.285, p⬍ .0001) (Table 4, Model III). The result indicated that the intervention can improve subjects’ protection self-efficacy.

Attitudes toward premarital sex

At baseline, 32% of respondents approved of premarital sex. At posttest, the proportion increased 2.4% in the inter-vention group compared to 2.0% in the control, but those who opposed premarital sex accounted for a larger propor-tion (44.5%) than in the control group (35.5%). No signif-icant difference was observed between two groups in either survey.

Multinomial Logistic regression analysis was done to evaluate the effect of intervention; results showed that the intervention had no effect on changing respondents’ atti-tudes toward premarital sex (p⬎ .05).

Communication behavior

At baseline, there were 52.6%, 65.7%, and 22.8% of subjects who reported having communicated with parents, peers, and teachers on HIV/AIDS issues, respectively, and no significant differences were observed between the two groups (p ⬎ .05). The proportions increased in both the control and intervention group at posttest, but the increase in the intervention group was higher than the increase in the control group (with 7.1%, 19.2%, and 38.6% compared to 4.7%, 1.3%, and⫺1.0%, respectively) (Table 3).

We used the binary logistic regression analysis to eval-uate the effectiveness of the intervention on respondents’ communication regarding HIV/AIDS issues. It was shown Table 3

Main indicators at baseline and end line, intervention and control groups (median/proportion)

Indicators Intervention group Control group

Total Male Female Total Male Female

HIV/AIDS knowledge score (median)

Baseline 51.5 54.6 45.5 54.6 60.6 45.5

End line 78.8a _81.8a _75.8a _{57.6 63.6 48.5}

Attitude score (median)

Baseline 20.0 20.0 20.0 20.0 20.0 20.0

End line 80.0a _80.0a _80.0a _{40.0 40.0 40.0}

Self-efficacy score (median)

Baseline 81.3 80.7 82.2 80.7 81.2 80.1

End line 85.8a _86.3a _85.1a _{79.0 79.6 78.4}

Having communicated with parents on HIV/AIDS issues (proportion %)

Baseline 53.6 50.5 58.1 51.0 51.8 50.0

End line 60.7 57.4 65.1 55.7 53.7 58.0

Having communicated with peers on HIV/AIDS issues (proportion %)

Baseline 65.9 69.6 60.6 65.4 71.9 57.0

End line 85.1a _86.4a _83.3a _{66.7 72.7 59.1}

Having communicated with teachers on HIV/AIDS issues (proportion %)

Baseline 21.1 22.5 19.0 25.4 28.8 21.0

End line 59.7a _59.6a _60.0a _{24.4 27.1 20.7}

a_{p⬍.0001, Wilcoxon test (for median) and chi-square test (for proportion) between intervention and control groups.}

Table 4

Ordinal logistic regression analysis with HIV/AIDS knowledge as outcome variable (Model I), with attitude towards daily contact with HIV positive people as outcome variable (Model II) and with self-efficacy as outcome variable (Model III)

Variables OR 95% CI p

Model I (HIV/AIDS knowledge)

Group (intervention vs. control) 0.858 0.677–1.088 .2063 Time (end line vs. baseline) 1.190 0.910–1.558 .2041 Group⫻time 11.405 8.043–16.173 ⬍.0001 Model II (attitude towards daily

contact with HIV-positive people)

Group (intervention vs. control) 1.006 0.773–1.308 .9651 Time (end line vs. baseline) 1.447 1.072–1.952 .0157

Group⫻time 2.842 1.942–4.159 ⬍.0001

Model III (self-efficacy)

Group (intervention vs. control) 1.120 0.889–1.411 .3367 Time (end line vs. baseline) 0.747 0.574–0.972 .0298 Group⫻time 2.371 1.711–3.285 ⬍.0001 Note:For all models, controlled for gender, age, residence registration, parents’ education level, parents’ occupation, family economic status, per-ceived care from parents, type of family, having browsed pornographic web-sites or not, having discussed sex-related issues with parents or friends or not.

that the intervention increased respondents’ communication behavior with peers and teachers significantly. Communica-tion with teachers on HIV/AIDS issues increased most signif-icantly, with 5.876 of OR (95% CI: 3.828 –9.020,p⬍.0001) for the group⫻time interaction effect (Table 5, Model III), followed by the communication with peers (OR⫽2.638, 95% CI: 1.734 – 4.012,p⬍.0001) (Table 5, Model II). However, a group⫻time interaction effect was not found on respondents’ communication with parents (OR⫽ 0.913, 95% CI: 0.623– 1.338) (Table 5, Model I).

Discussion

As health education and life skills have evolved during the past decade, there is growing recognition of and evi-dence for the role of psychosocial and interpersonal skills in the development of young people [15,16]. Student partici-pation in active learning can strengthen student–teacher relationships, improve the classroom climate, accommodate a variety of learning styles, and provide alternative methods for teaching. In the study, acceptance and participation in LPS training suggested that such interactive education ac-tivities were welcomed by the adolescents, and they per-ceived the intervention as valuable and enjoyable. After 3 months of LPS training, students were more aware of how to protect themselves from HIV as well as more likely to care for people living with HIV. LPS training strengthened students’ sex protection self-efficacy, such as how to resist pressure to have sex. The intervention also influenced re-spondents’ communication with teachers and peers regard-ing HIV and AIDS, which was helpful for mitigatregard-ing

re-spondents’ fear of AIDS and building their HIV and AIDS awareness in a friendly atmosphere. Results are consistent with previous studies suggesting that information provision coupled with skills building can positively impact youth reproductive health[6,19]. Compared to previous sex edu-cation projects conducted in rural areas of China[8,9], the interactive methods used in this study seem more effective, which not only increased respondents’ knowledge, promot-ing their friendly attitudes to HIV-infected persons, but also resulted in some positive changes on resisting peers’ sexual pressure self-efficacy and communication behaviors.

In general, in many Chinese rural areas like Shangcai County, a lot of young people leave home to find a job in the city after graduation from middle school. Without basic risk perception and self-protection skills, these youth may be-come sexually active and practice unsafe sexual behaviors. In addition, due to engagement customs in rural areas, premarital sex is prevalent, which also increases the high risk of induced abortion, STIs, and HIV among youth. Furthermore, youth who live in areas with high HIV prev-alence, such as Henan, Yunnan, and Xinjiang provinces, also face the anxiety of living with HIV-positive people. They need information to dispel their fears. Thus, effective youth-friendly programs with information dissemination and skills building are crucial to rural youth. LPS training is such an appropriate program.

Young people have been known to become “a powerful force for change in their own households, in the lives of their peers, and in the wider community too” [20]. Our analysis also indicated that LPS training increased students’ communication skills, which facilitated nearly half the stu-dents to convey what they learnt to others in their commu-nity. It is worthy to note that students’ communication with teachers was significantly increased by the intervention. Facilitators also mentioned that the intervention improved the relationship between students and teachers. In the LPS training, the role of teachers changed from instructors to friends of adolescents, which facilitated information sharing among them.

However, the intervention did not significantly increase youths’ communication with parents. A probable reason is that most students lived in a dormitory and only went home for 2 days monthly; thus, they had little chance to talk with their parents about HIV/AIDS issues. In addition, there was no significant difference between the intervention and con-trol groups related to attitudes toward premarital sex. The increase of HIV-related knowledge did not lead to corre-sponding changes in students’ attitudes to premarital sex because these attitudes are also related to values, sexual mores, ethics, and environment. Also, because it is the first introduction of such active sex education in the traditional conservative rural surrounding, more experiences need to be learned based on the intervention. It is likely that only a repeated and long-term intervention could significantly in-fluence these deeply entrenched attitudes. On the other Table 5

Binary logistic regression analysis with all respondent’s communication with parents, peers, and teachers on HIV/AIDS issues as outcome variable (Model I, Model II, and Model III)

Variables OR 95% CI p

Model I (communication with parents on HIV/AIDS issues)

Group (intervention vs. control) 1.224 0.932–1.606 .1458 Time (end line vs. baseline) 1.404 1.030–1.914 .0318

Group⫻time 0.913 0.623–1.338 .6402

Model II (communication with peers on HIV/AIDS issues)

Group (intervention vs. control) 1.104 0.833–1.464 .4910 Time (end line vs. baseline) 1.167 0.846–1.608 .3474 Group⫻time 2.638 1.734–4.012 ⬍.0001 Model III (communication with teachers

on HIV/AIDS issues)

Group (intervention vs. control) 0.850 0.620–1.165 .3128 Time (end line vs. baseline) 0.992 0.698–1.408 .9620 Group⫻time 5.876 3.828–9.020 ⬍.0001 Note:For all models, controlled for gender, age, residence registration, parents’ education level, parents’ occupation, family economic status, perceived care from parents, type of family, having browsed pornographic websites or not, having discussed sex-related issues with parents or friends or not.

hand, the program should consider including more related contents such as family communication skills and sexual values and life planning in the program. The results also proved that such education does not increase students’ open attitude to premarital sex.

LPS training can be widely applied to rural settings because it only requires simple tools that are common in teaching activities such as paper, pencil, blackboard, and so on. In addition, it can be designed within the context of a regular classroom environment where students can practice skills. Thus, scaling up the LPS training program may be a practical way to improve the reproductive health of adoles-cents and defend against the spread of HIV. However, LPS does require high-quality trained facilitators. It is facilitators that stimulate students’ capacity to observe, think, commu-nicate, and practice. Because this study was done in an area with high HIV prevalence, many students had heard of or knew people living with HIV, which gave them a lot to discuss in the classroom. With the facilitators’ guidance, students built scientific awareness by communication with each other. Thus, the quality of the facilitators is crucial to the success of the LPS training program. In the study, the introduction of school cluster workshops and retraining of facilitators helped to strengthen the training quality. In ad-dition, advocating and obtaining support from the local educational sector, school personnel, and parents is also important to the success of the intervention.

Several study limitations should be noted. First, the cho-sen schools were not randomly assigned to intervention and control groups. We chose the intervention schools first, then found a comparable school to use as a control. Although we tried to match some confounding factors such as social– economic status and demographic characters, there may be some unknown factors influencing the effect of interven-tion, which might increase or decrease the real effect of intervention. Second, the measurement of protection self-efficacy score constitutes three statements and attitude to-ward premarital sex is measured by one statement, which seems limited and might decrease the accuracy of measure-ment. Also, self-reported self-efficacy scores is only a sim-ilar measurement for self-protection skills change. Third, the intervention period was short (only 3 months). Assess-ing the medium- to long-term impact of this and similar programs is essential to determine whether the observed patterns are sustained and to measure whether the observed changes lead to behavior change and safe behaviors.

Acknowledgments

The authors thank PATH and the China Family Plan-ning Association (CFPA), who provided generous sup-port for the evaluation study through the China Youth

Reproductive Health Project, funded by the Bill & Melinda Gates Foundation.

References

[1] Piao BY. A survey of AIDS villages in Henan Province: we don’t want society’s concern, we want economic growth [Online]. Avail-able at:http://www.China-aids.org/chinese/News/News045.htm (Ac-cessed June 15, 2005).

[2] Wanyanhai. AIDS, law and human right in Henan Province, China [Online]. Available at:www.aizhi.net(Accessed May 1, 2005). [3] Ma YH, Hu PJ, Cheng YC. The study of Aids related knowledge and

attitudes and behavior and needs of health education among high school students and teachers. China Public Health 1999;15(6):145–9. [4] Gallanta M, Maticka-Tyndaleb E. School-based HIV prevention

pro-grammes for African youth. Soc Sci Med 2004;58:1337–51. [5] Chen J, Wang J, Sun J. Research of school AIDS prevention

educa-tion on knowledge of secondary school students in Beijing. China J School Health 2001;22(1):46 –7.

[6] O’Donoghue J. Zimbabwe’s AIDS action programme for schools. Eval Program Plann 2002;25:387–96.

[7] Kirby D, Short L, Collins J, et al. School-based programs to reduce sexual risk behaviors: a review of effectiveness. Public Health Rep 1994;109(3):339 – 60.

[8] Lou CH, Ding YG, Gao ES. Study on effectiveness of sex and reproductive health education among high school students. China Public Health 2002;18(6):645– 6.

[9] Zheng QQ, Li YQ, Lu P. Evaluation of the effect of a short-term sex health education among rural high school females. Matern Child Health Care China 2001;16(8):487–9.

[10] Botvin GJ, Eng A, Willians CL. Preventing the onset of cigarette smoking through life skills training. Prev Med 1980;9:135– 43. [11] Hamburg BA, ed. Life-Skills Training: Preventive Interventions for

Young Adolescents. Washington, DC: Carnegie Council on Adoles-cent Development, 1990.

[12] Perry C. Results of prevention programs with adolescents. Drug Alcohol Depend 1987;20:13–9.

[13] LaFromboise TD, Howard Pitney B. The Zuni life skills development curriculum: a collaborative approach to curriculum development. Am Indian Alaska Native Mental Health Res Monogr Series 1994;4:98 –121. [14] Kreutter KL, Gewirtz H, Davenny JE, et al. Drug and alcohol pre-vention project for sixth graders: first-year findings. Adolescence 1991;26(102):287–93.

[15] WHO. Skills for health. Information series on school health document 9 [Online]. Available at: http://www.who.int/school_youth_health/ media/en/sch_skills4health_03.pdf(Accessed May, 2005). [16] PATH. Reaching out-of-school youth with life-planning skills

edu-cation: the African Youth Alliance’s behaviour change communica-tion efforts in Arusha, Tanzania [Online]. Available at:http://www. path.org/files/AH_aya_chawakua.pdf(Accessed Sep. 10, 2006). [17] Chen J, Sun J, Rong W, et al. Assessment of school AIDS prevention

education of secondary school students in Shenyang. China Public Health 2001;17(1):84 – 6.

[18] Magnani R, Maclntyre K, Mehyrar Karim A, et al. The impact of life skills education on adolescent sexual risk behaviors in KwaZulu-Natal, South Africa. J Adoles Health 2005;36:289 –304.

[19] Best Practices in School AIDS Education. The Zimbabwe Case Study. New York: United Nations Joint Programme on HIV/AIDS, 2000. [20] Nwokocha ARC, Nwakoby BAN. Knowledge, attitude, and behavior

of secondary (high) school students concerning HIV/AIDS in Enugu, Nigeria, in the year 2000. J Pediatr Adolesc Gynecol 2002;15:93– 6.