Two-Year Follow-Up of a Sexual Health Promotion Program for Spanish Adolescents

Abstract

The aim is to evaluate the effects of the Competencias para adolescentes con una sexualidad saludable (COMPAS) program and compare them with an evidence-based program (¡Cuídate!) and a control group (CG). Eighteen public high schools were randomly assigned to one of the three experimental conditions. Initially, 1,563 Spanish adolescents between 14 and 16 years of age participated, and 24 months after their implementation, 635 of them completed a survey. Self-report measures collected data on sexual behavior, knowledge, attitudes, intention, sexual risk perception, and perceived norm. Compared to the CG, COMPAS increased the level of knowledge about sexually transmitted infections and improved the attitudes toward people living with human immunodeficiency virus at the 2-year follow-up. Neither intervention had a long-term impact on behavioral variables. Results suggest that COMPAS has a comparable impact to the other intervention on the variables predicting consistent condom use. Reinforcing the messages and skills that have the greatest impact on condom use and adding booster sessions following program completion as strategies to maintain long-term effects are necessary.

Keywords

adolescence HIV sexual health STIs program evaluation preventive programs

According to the latest report by the Spanish National Epidemiology Center (Centro Nacional de Epidemiología, 2014), approximately 25.4% of new acquired immune deficiency syndrome (AIDS) cases in Spain were diagnosed between 20 and 39 years of age, which suggests that a large majority were infected during adolescence and early youth due to the disease’s long incubation period and the high proportion of late diagnoses. In 2013, 2,973 Spanish teenagers between 14 and 16 years of age voluntarily terminated their pregnancies. In most of these cases, the pregnancies were unintended, and they accounted for 2.74% of all voluntary terminations in Spain (Ministerio de Sanidad, Servicios Sociales e Igualdad, 2015). These data show the vulnerability of adolescents to sexually transmitted infections (STIs) and unplanned pregnancies due to their involvement in risky sexual behavior. Inconsistent condom use and an early onset age for sexual relations have been identified as predictors of human immunodeficiency virus (HIV) infection and other STIs in multiple studies (Beadnell, Morrison, & Wilsdon, 2005; Pettifor, Van der Straten, Dunban, Shi-Boski, & Padian, 2004).

Interventions to promote healthy sexual habits are aimed at reducing sexual risk behavior, including activities aimed at increasing consistent condom use by influencing the precursors according to the main theoretical variables based on health behavior models such as the information–motivation–behavioral skills (IMB) model (Fisher & Fisher, 1992) and the theory of planned behavior (Ajzen, 1991). Schools are a key environment for prevention and health promotion due to being highly accessible within the community and for the important role they play in the education and development of young people (Ministerio de Sanidad y Consumo, 2008; Monsalve, 2012). However, the proportion of Spanish schools that carry out prevention interventions is low. Exceptions to these include the autonomous communities of Catalonia and Cantabria with coverage of 90%; in some, such as Madrid and Navarre (Ministerio de Sanidad, Servicios Sociales e Igualdad, 2009), only short talks are held. Although some schools implement school programs for this purpose, the impact of these actions is unknown since their effectiveness is rarely evaluated in terms of outcome (Espada, Morales, Orgilés, Piqueras, & Carballo, 2013).

In Spain, there is very little evidence about the effects of programs to promote sexual health and prevent HIV because they are rarely evaluated, and when they are, they are evaluated in the short term. According to the systematic review by Espada, Morales, Orgilés, Piqueras, and Carballo (2013), most short-term evaluations present relevant methodological issues. In this review, only 2 of the 14 identified studies evaluated the effect of the intervention at three time points: pretest evaluation, posttest, and follow-up. Hernández-Martínez et al. (2009) evaluate an HIV-prevention intervention 6 months after its application in a sample of 481 adolescents aged 16 and 17. Espada (2007) presents the results of an intervention that promotes healthy sexual habits, Competencias para adolescentes con una sexualidad saludable (COMPAS; Competencies for adolescents with a healthy sexuality), and includes a follow-up assessment at 12 months in 832 adolescents from different geographic areas of Spain. In the scientific literature, we have yet to find controlled studies in Spain, evaluating the effectiveness of programs to promote sexual health that extend beyond a 12-month follow-up.

The COMPAS program promotes healthy sexual practices implemented in schools based on the development of skills in addition to the transmission of information. COMPAS is aimed at teenagers between 14 and 18 years, with the aim of preventing the transmission of HIV/AIDS and other STIs along with unplanned pregnancies. Several controlled studies evaluate the immediate effects of COMPAS (Espada, Morales, Orgilés, Jemmott, & Jemmott, 2015; Espada, Orgilés, Morales, Ballester, & Huedo-Medina, 2012; Morales, Espada, Orgilés, Secades-Villa, & Remor, 2014) and 12-month outcomes (Morales, Espada, & Orgilés, 2015). At the posttest, COMPAS has shown to be effective in increasing the level of knowledge about HIV and other STIs, sexual risk perception, and the intent to engage in safe sex, and it promotes a favorable attitude toward protection methods and issues related to prevention. At 12 months, adolescents who received the COMPAS program showed a higher level of knowledge about HIV and STIs, a more favorable attitude toward condom use and prevention aspects, perceived high rates of peer condom use, and they initiated in vaginal sex later than the control group (CG; Morales et al., 2015). However, the impact of outcomes at the 24-month follow-up is unknown.

The aim of this study is to evaluate the effectiveness of COMPAS through psychosocial (knowledge, attitude, risk perception, and perceived norm) and behavioral constructs (age at first intercourse with vaginal penetration, oral sex and anal sex, consistent condom use, and number of sexual partners) in monitoring the effects 2 years after the program’s implementation. The effects of COMPAS are compared with those of ¡Cuídate!, an evidence-based intervention for Latino adolescents in the United States, and a CG who did not receive any intervention. Based on previous studies (Espada et al., 2015; Morales et al., 2015), it is hypothesized that COMPAS will be at least as effective as ¡Cuídate! and more effective than no intervention to promote healthy sexuality in Spanish adolescents 2 years after its implementation.

Method

Study Design and Participants

In this cluster-randomized trial, 1,563 adolescents from different geographic areas in Spain (north, south, and east) participated. In 2012, students aged 14–16 in 9th and 10th grades of high school or the equivalent were invited to participate in this study. Of the eligible participants, 97% obtained parental written consent. Figure 1 shows the number and proportions of adolescents throughout the study by condition and time. All the adolescents who answered the baseline and 24-month follow-up assessments were included in the analyses. Twenty-four months after the programs’ implementation, 635 of the participants (40.62% retention) completed the survey (between January and April 2014).

Figure 1.

Flowchart for group-randomized, controlled design. Participants who were not followed up were absent at the time of the follow-up session at school and did not answer the evaluation online for unknown reasons.

Assessment of Attrition Bias at Follow-Up

Regarding the analysis of external invalidity, differences (p < .05) were only found in age (p = .01). Students who dropped out were older (M = 15; SD = 1.10) than those surveyed at 2-year follow-up (M = 14.83; SD = 1.01). No differences were found on sex (p = .06), be sexual experienced or not (p = .94) and consistent condom use (p = .68) between those who dropped out and those who did not. The attrition analysis comparing experimental conditions (internal invalidity) revealed statistically significant difference in the retention of participants at the 2-year follow-up between the experimental conditions (p = .03), where the groups receiving the COMPAS (75%) and ¡Cuídate! (65%) interventions had higher retention percentages than the CG (42%; Figure 1). However, the data confirm that the differences in the dropout rate between groups were not related to consistent condom use (p = .23), which is the main study outcome.

Interventions

COMPAS program

COMPAS is a school-based sexual health education intervention developed and tested in Spain. The theoretical models underlying COMPAS are the social learning theory (Bandura, 1986) and the IMB model (Fisher & Fisher, 1992). COMPAS consists of five sessions, each lasting 50 min. The first two sessions are dedicated to addressing knowledge about identifying sexual risk, HIV/AIDS and other STIs, methods of protection, and mistaken beliefs. The third session addresses contents such as making decisions related to condom use. The last two modules are about communication skills and ability to negotiate in several sex situations. The complete program description is detailed in Espada, Morales, Orgilés, Jemmott, and Jemmott (2015). The goals of the intervention are: (a) increase knowledge about HIV and other STIs, (b) obtain more favorable attitudes toward condom use and normative beliefs of safe sex behaviors, (c) increase the risk perception of unprotected sex, (d) increase the intention to not engage in sexual risky behaviors, (e) improve problem-solving skills and condom use negotiation skills, (f) increase consistent condom use, and (g) reduce the number of sexual partners. More detailed information about the intervention components and the effectiveness of COMPAS are published elsewhere (Espada et al., 2012, 2015; Morales et al., 2014).

Adapting the original ¡Cuídate! program to Spanish adolescents

¡Cuídate! is considered the only high-quality program implemented for the Latino population by the Centers for Disease Control and Prevention (2014). It was initially developed to reduce sexual risk among Latino adolescents (13–18 years) living in the United States (Villarruel, Jemmott, & Jemmott, 2006). It was culturally adapted from the Be Proud! Be Responsible! curriculum that has been effective in reducing sexual risk among African American youth (Jemmott, Jemmott, & Fong, 1992, 1999, 2010). The principal goals of the program are to influence attitudes, behavioral and normative beliefs, self-efficacy, and negotiation skills regarding HIV risk-reduction behaviors, specifically abstinence and condom use.

The adaptation of ¡Cuídate! to adolescents from Spain is detailed in a previous article (Espada et al., 2015), thus a brief description of the process is provided. A pilot study was conducted to adapt ¡Cuídate! to adolescents living in Spain. As a first step, focus groups with adolescents were conducted to evaluate the appropriateness of the curriculum to Spain. Specific cultural aspects of ¡Cuídate! for Latinos from the Unites States were revised to be more relevant for Spanish adolescents. The latest version of the curriculum was approved by psychologists who are experts in sexual risk reduction among adolescents. The Spanish version of ¡Cuídate! maintains the principal core elements of the original ¡Cuídate! and the key theories: social cognitive theory and the theory of planned action (Ajzen, 1991; Bandura, 1986; Fishbein & Ajzen, 1975). Its curriculum consists of six 45-min sessions. The contents of the first three sessions are dedicated to teaching about HIV/AIDS and concepts associated with its prevention. The fourth and fifth sessions address contents about condom use (barriers, skills for correct use). The last session’s contents are about communication skills. The complete program description is detailed in Espada et al. (2015).

Differences between both curriculums

COMPAS and ¡Cuídate! were tested with Spanish-speaking adolescents targeted to reduce sexual risk behaviors. Similar theoretical models underlie both group-based interventions and they include the same components to promote sexual health. The methodology used consists of high participation and interaction by both facilitator and participants. However, some differences between the interventions are noticed. Unlike COMPAS, ¡Cuídate! promotes sexual abstinence, and includes condom demonstrations and a safer sex negotiation strategy, a talking circle to begin and end each session, HIV/STI videos, and provides condoms to the participants. Unlike ¡Cuídate!, COMPAS includes a description of how HIV affects the body, a self-talk activity, a guided visual imagery activity, and styles of communication applied to dealing with sexual risky situations (e.g., assertive, passive, and argumentative). ¡Cuídate! and COMPAS can be considered similar in duration, and they employ theoretical models and content that are similar; these are the main reasons why ¡Cuídate!, as an evidence-based program, was selected to evaluate the effectiveness of COMPAS (Espada et al., 2015; Morales et al., 2015). The characteristics and components of both interventions are detailed elsewhere (Espada et al., 2015).

Procedure

The research ethics committee at the responsible institution revised and approved the study. A protocol to standardize the phases of recruitment, implementation, and evaluation of the interventions was designed. Using a computerized random number generator, three high schools from each of the five participating provinces were randomly selected and invited to participate in the study—15 centers in all. One of them was ineligible due to other sexual risk activities that had been implemented the preceding year. In order to meet the sample sizes per area—a minimum of 300 participants—four additional high schools were randomly selected. Ultimately, three schools in two provinces participated while in the other three provinces, four centers participated. The authorities at the participating schools committed to not carry out other similar interventions during the project’s duration. The project coordinator randomly assigned the 18 high schools to one of the experimental conditions: COMPAS, ¡Cuídate!, or as a no-intervention CG.

Both interventions implemented 1-hr sessions each week during school hours. There was one facilitator per province. The facilitators were 2 men and 3 women possessing experience in health promotion with adolescents. Their median age was 28.8 (range: 25–29). All of them received 6 hr of training for each intervention. Additional videos demonstrating the implementation of both programs to teenagers were provided to the facilitators. Both programs are highly structured, and facilitators implemented it following intervention manuals. Moreover, online tutorials were also provided to resolve questions about the procedure. These variables are considered as facilitators of the fidelity of implementation (Carroll et al., 2007). The fidelity of implementation was high according to previous studies that evaluated the dimensions of dose and adherence in ¡Cuídate! (Escribano, Morales, Orgilés, & Espada, 2015), in addition to those of dose, adherence, and acceptance in COMPAS (Escribano et al., Manuscript in preparation). The effects of both interventions were evaluated before the intervention and again 24 months later using the Google Forms online survey. The evaluations were conducted in computer classrooms at the high schools. A member of the team conducted phone interviews to collect data from participants who were no longer enrolled in school at the 24-month follow-up. Confidentiality was ensured by providing a personal code for each adolescent. The participants voluntary accepted to participate in the study and group incentives were provided. More detailed information about the procedure is published elsewhere (Espada et al., 2015; Morales et al., 2015).

Measures

Sociodemographic variables

The following variables were evaluated: gender, age, city of residence, school year, family structure (married, separated and/or divorced, single mother/father, and lost one or both parents), and socioeconomic status by the Family Affluence Scale (FAS; Boyce, Torsheim, Currie, & Zambon, 2006). This questionnaire assesses the number of cars, computers, individual bedrooms, and vacations during the last 12 months. FAS validity with respect to current economic indicators, such as the gross national product, is .87.

Behavioral measures

Five questions relating to sexual practices were raised: (a) sexual orientation: heterosexual, bisexual, and homosexual; (b) if they have had sex, including petting, masturbation, oral sex, vaginal sex, or anal sex; (c) age of first intercourse, distinguishing vaginal, anal, and oral penetration; (d) number of sexual partners in the last 6 months; and (e) percentage of condom use in the last 6 months (scale from 0 to 100), which is categorized in consistent condom use with a dichotomous variable response—100% was considered consistent use and the rest was considered inconsistent use.

Psychosocial measures

Knowledge about HIV and other STIs

The Scale of Knowledge about HIV and other STIs (Espada, Guillén-Riquelme, Morales, Orgilés, & Sierra, 2014) was used. This consists of 24 items with three possible answers: true, false, or do not know, and information is collected about five factors: general, about condoms, routes of transmission, prevention, and other STIs. The sum of the five factors provides a total score ranging from 0 to 24, with higher scores indicating higher levels of knowledge about HIV and other STIs. The questionnaire has high internal consistency (α = 0.88).

Attitudes toward aspects related to HIV. Attitudes toward aspects related to HIV were assessed by means of the HIV Attitudes Scale for Adolescents (Espada, Ballester, et al., 2013). It measures four dimensions about attitude toward condom use when obstacles exist, HIV testing, condom use, and people living with HIV/AIDS. The total score ranged from 12 to 48, and higher scores represent more favorable aspects toward HIV/AIDS. These items explained 65% of the variance, and its Cronbach’s α was 0.77.

Intention to engage in safe sex

This evaluated the intention to engage in safer sexual behavior over 12 months with 5 items, which make up two factors: (1) the intention to find, use, and negotiate condom use with sexual partners (α = 0.80) and (2) the intention of condom use under the influence of alcohol and other drugs using a 5-point Likert-type scale response, ranging from 1 = definitely not to 5 = sure (α = 0.75).

Sexual risk perception

This was assessed by two questions with 4-point scale responses, with 1 being no risk and 4 great risk: “How much risk is there in the following activities? (a) having oral sex without a condom or (b) engaging in sexual intercourse/penetration without a condom.” These questions are asked for three different situations related to the possibility of having HIV, other STIs, or an unplanned pregnancy. Cronbach’s α value for the instrument was 0.87.

Perceived norm was evaluated by asking the perception of the frequency of condom use in their peers in sexual intercourse with a scale of 1 being never and 4 always.

Statistical Analysis

The analyses were conducted using SPSS, Version 22. Logistic regression was used to analyze the influence on the dropout rate of age, sex, being sexually active, and consistent condom use. Descriptives of the three conditions were calculated by analysis of variance with repeated measures Generalized Linear Model (GLM). Generalized estimating equations (GEE) were used to evaluate the effect of the program after a 24-month follow-up, adjusting the values for baseline differences in sex, age, and type of sexual experience.

GEE models are commonly used to evaluate trials that involve clusters (e.g., schools), since they control the correlations among responses within clusters (Liang & Zeger, 1986). Effects of the two interventions were tested with planned comparisons of prespecified hypotheses (Rosenthal & Rosnow, 1985). To test the 24-month follow-up effects of COMPAS, one contrast compared COMPAS with the CG and another contrast compared COMPAS with ¡Cuídate! To test the 24-month follow-up effects of ¡Cuídate!, a contrast compared ¡Cuídate! with the CG. Based on previous studies (Espada et al., 2015; Jemmott, Jemmott, O’Leary, et al., 2010; Morales et al., 2015), the school (i.e., the cluster) was the unit of randomization, and the individual was the a priori unit of analysis; therefore, the school was controlled in all the analyses. All of the adolescents who answered the baseline and 24-month follow-up assessments were included in the analyses.

Results

Participants’ Characteristics

The characteristics of the participants are shown in Table 1. Statistically significant differences in the baseline, such as their age (p < .05), were controlled in all subsequent analyses.

Table 1.

Baseline Comparability of Self-Reported Behaviors of 24-Month Follow-Up Participating Students by Experimental Condition.

Characteristics		COMPAS (n = 300)	¡CUÍDATE! (n = 200)	CONTROL (n = 135)	Total (N = 635)	p Valueª
Male, n (%)		133 (44.3)	95 (47.5)	56 (41.5)	284 (44.7)	.544
Age, M (SD) years		16.70 (0.71)	16.85 (0.74)	16.65 (0.67)	16.73 (0.71)	.020
16 Years old		138 (46)	72 (36)	61 (45.2)
17 Years old		117 (39)	86 (43)	59 (43.7)
18 Years old		45 (15)	42 (21)	15 (11.1)
People who have married parents, n (%)		234 (78)	163 (81.5)	101 (74.8)	498 (78.4)	.335
Family income, n (%)	Low	86 (28.7)	61 (30.5)	49 (36.3)	196 (30.9)	.310
	Middle	190 (63.3)	123 (61.5)	71 (52.6)	384 (60.5)
	High	24 (8)	16 (8)	15 (11.1)	55 (8.7)
Heterosexual, n (%)		281 (94.6)	186 (93.9)	123 (92.5)	590 (93.9)	.564
Bisexual, n (%)		9 (3)	10 (5.1)	7 (5.3)	26 (4.1)
Homosexual, n (%)		7 (2.4)	2 (1)	3 (2.3)	12 (1.9)
Sexually experienced,^b n (%)		208 (69.3)	153 (76.9)	101 (74.8)	462 (72.9)	.151
Experienced in sexual risk practices,^c n (%)		158 (53)	121 (61.1)	83 (61.5)	362 (57.4)	.112
Multiple partners in the past 6 months, n (%)		72 (53.7)	45 (52.9)	28 (47.5)	145 (52.2)	.713
Condom use (0–100), % (SD)		89 (19.83)	86 (21.92)	86 (21.58)	87.63 (20.86)	.560
Consistent condom use, n (%)		64 (21.3)	41 (20.5)	25 (18.5)	130 (20.5)	.807

^aSignificance tested using F test for continuous variables and χ² test for categorical variables. ^bPetting, vaginal sex, oral sex, anal sex, or mutual masturbation. ^cVaginal sex, anal sex, and oral sex.

Effects of the Interventions

Behavioral measures

Twenty-four months after application, neither COMPAS nor the comparison program, had a significant impact on behavioral variables compared to the CG. Adolescents who received ¡Cuídate! significantly delayed their age of first oral sex (M = 15.73; SD = 1.12) compared to COMPAS (M = 15.54; SD = 1.14) (Table 2). No statistically significant differences between both interventions in other behavioral variables were found (Table 3).

Table 2.

Adjusted Marginal Means for Self-Report Sexual Risk Behavior and Psychological Variables by Experimental Condition and Time.

Outcomes	Baseline			24-Months Follow-Up
	COMPAS	¡CUÍDATE!	CONTROL	COMPAS	¡CUÍDATE!	CONTROL
STI and HIV knowledge, M (SE)
General HIV	5.43 (0.56)	5.35 (0.56)	5.45 (0.57)	5.89 (0.56)	5.85 (0.56)	5.50 (0.57)
Condom use	1.45 (0.24)	1.38 (0.24)	1.48 (0.24)	1.50 (0.25)	1.52 (0.25)	1.46 (0.25)
Routes of transmission	3.30 (0.54)	3.15 (0.53)	3.33 (0.53)	4.47 (0.54)	4.34 (0.54)	4.11 (0.54)
Prevention	1.53 (0.38)	1.45 (0.38)	1.49 (0.38)	2.30 (0.38)	2.34 (0.34)	2.29 (0.39)
Other sexual infections	1.52 (0.53)	1.52 (0.53)	1.77 (0.53)	1.48 (0.48)	1.42 (0.53)	1.28 (0.54)
Total	14.00 (1.49)	13.62 (1.48)	14.30 (1.50)	15.98 (1.50)	15.84 (1.50)	14.85 (1.52)
HIV-related attitudes, M (SE)
Obstacles	2.41 (0.21)	2.35 (0.21)	2.37 (0.21)	2.39 (0.21)	2.50 (0.21)	2.44 (0.22)
HIV test	1.70 (0.14)	1.71 (0.14)	1.73 (0.14)	2.64 (0.14)	2.73 (0.14)	2.68 (0.14)
Condom	3.28 (0.22)	3.25 (0.21)	3.25 (0.21)	2.54 (0.22)	2.64 (0.21)	2.56 (0.22)
People living with AIDS	2.21 (0.20)	2.12 (0.20)	2.12 (0.20)	2.57 (0.20)	2.61 (0.20)	2.53 (0.20)
Total	9.61 (0.52)	9.43 (0.51)	9.49 (0.51)	9.15 (0.52)	9.46 (0.51)	9.28 (0.53)
Intention to engage in safe sex, M (SE)	4.22 (0.29)	4.19 (0.29)	4.20 (0.28)	4.18 (0.29)	4.29 (0.29)	4.22 (0.29)
Sexual risk perception, M (SE)	2.65 (0.20)	2.60 (0.20)	2.66 (0.20)	2.76 (0.20)	2.73 (0.20)	2.75 (0.20)
Perception of frequency of peer condom use, n (%)^a
Always	108 (27.1)	83 (28.8)	93 (29.1)	69 (23.2)	40 (20.1)	23 (17)
Almost always	188 (47.2)	113 (39.2)	141 (44.1)	163 (54.9)	100 (50.3)	77 (57)
Sometimes	82 (20.6)	75 (26)	72 (22.5)	58 (19.5)	58 (29.1)	35 (25.9)
Never	20 (5)	17 (5.9)	14 (4.4)	7 (2.4)	1 (0.5)	0 (0)
Age at the first vaginal sex, M (SE)	14.38 (1.08)	14.51 (1.09)	14.51 (1.09)	15.33 (1.12)	15.61 (1.09)	15.48 (1.12)
Age at the first anal sex, M (SE)	15.23 (3.45)	14.36 (3.48)	14.61 (3.45)	16.28 (3.62)	16.86 (3.50)	15.90 (3.59)
Age at the first oral sex, M (SE)	15.27 (1.14)	14.51 (1.13)	14.77 (1.10)	15.54 (1.14)	15.73 (1.12)	15.67 (1.13)
Consistent condom use, % (SE)	30 (56.6)	35 (53)	37 (52.1)	64 (21.3)	41 (20.5)	25 (18.5)
Multiple partners, % (SE)	30 (50.8)	45 (50.6)	43 (51.8)	72 (53.7)	45 (52.9)	28 (47.5)

Note. STI = sexually transmitted infection; HIV = human immunodeficiency virus.

ªThe numerator represents the number of participants who selected the option and the denominator indicates the number of participants who responded to the question. Analysis was adjusted for gender-, age-, and school-level sexual experience at baseline.

Table 3.

Generalized Estimating Equations Model-Based Significance Tests and Effect Size Estimates for the Intervention Effect on Self-Reported Sexual Behaviors Over the 24-Month Follow-Up Period.

	COMPAS—CONTROL		¡CUÍDATE!—CONTROL		COMPAS—¡CUÍDATE!
Outcome	AORª (95% CI)	p Value	AOR^b (95% CI)	p Value	AOR^c (95% CI)	p Value
Age at the first vaginal sex	0.87 [0.56, 1.37]	.562	1.00 [0.65, 1.52]	.993	1.14 [0.73, 1.79]	.553
Age at the first anal sex	1.89 [0.55, 6.17]	.318	0.77 [0.35, 1.66]	.514	0.43 [0.13, 1.41]	.168
Age at the first oral sex	1.64 [0.93, 2.87]	.060	0.77 [0.46, 1.27]	.307	0.47 [0.28, 0.85]	.013*
Consistent condom use	0.85 [0.42, 1.74]	.661	0.98 [0.50, 1.91]	.960	1.17 [0.56, 2.43]	.659
Total sexual partners	0.965 [0.43, 2.15]	.931	1.27 [0.60, 2.72]	.526	1.31 [0.57, 2.96]	.517
Multiple partners	0.99 [0.83, 1.16]	.904	1.00 [0.87, 1.16]	.911	1.02 [0.86, 1.19]	.811

Note. Each analysis was adjusted for gender-, age-, and school-level sexual experience at baseline. AOR = adjusted odds ratio; CI = confidence interval.

ªAn odds ratio > 1 indicates that the score is higher in COMPAS compared to the control group. ^bAn odds ratio > 1 indicates that the score is higher in ¡CUÍDATE! compared to the control group. ^cAn odds ratio < 1 indicates that the score is higher in COMPAS compared to ¡CUÍDATE! *p < 0.05.

Psychosocial outcomes

Table 4 shows the differences between the conditions in the outcome variables. At the 24-month follow-up evaluation, the adolescents who received COMPAS reported significantly higher scores for knowledge about other STIs (p < .05). The attitudes toward people living with HIV were significantly more favorable (p < .05) compared to the CG.

Table 4.

Generalized Estimating Equations Model-Based Significance Tests and Effect Size Estimates for the Intervention Effect on Self-Reported Sexual Behavior and Psychological Variables Over the 24-Month Follow-Up Period.

		COMPAS—CONTROL		¡CUÍDATE!—CONTROL		COMPAS—¡CUÍDATE!
Outcome		AORª (95% CI)	p Value	AOR^b (95% CI)	p Value	AOR^c (95% CI)	p Value
STI and HIV knowledge
General HIV		0.98 [0.78, 1.21]	.860	0.90 [0.70, 1.15]	.418	0.92 [0.72, 1.18]	.553
Condom use		0.97 [0.89, 1.06]	.574	0.90 [0.81, 0.99]	.042*	0.92 [0.83, 1.02]	.124
Routes of transmission		1.00 [0.81, 1.25]	.942	0.86 [0.69, 1.08]	.201	0.86 [0.69, 1.07]	.193
Prevention		1.04 [0.90, 1.20]	.580	0.96 [0.82, 1.11]	.616	0.93 [0.80, 1.07]	.336
Other sexual infections		0.77 [0.61, 0.97]	.031*	0.77 [0.60, 0.98]	.039*	1.00 [0.80, 1.25]	.969
Total	633	0.74 [0.41, 1.33]	.318	0.50 [0.26, 0.96]	.039*	0.70 [0.37, 1.32]	.277
Attitudes related to HIV
Obstacles		1.04 [0.97, 1.19]	.252	0.98 [0.90, 1.06]	.694	0.94 [0.87, 1.02]	.152
HIV test		0.97 [0.93, 1.02]	.333	0.98 [0.93, 1.04]	.651	1.01 [0.96, 1.06]	.576
Condom		1.03 [0.94, 1.12]	.474	0.99 [0.90, 1.10]	.959	0.97 [0.88, 1.05]	.496
People living with AIDS		1.09 [1.02, 1.17]	.023*	1.00 [0.92, 1.09]	.973	0.92 [0.85, 0.99]	.041*
Total	609	1.13 [0.93, 1.37]	.193	0.94 [0.76, 1.18]	.637	0.84 [0.68, 1.03]	.104
Intention to engage in safe sex	527	1.02 [0.93, 1.13]	.567	0.99 [0.89, 1.10]	.877	0.96 [0.87, 1.06]	.46
Sexual risk perception	625	0.99 [0.92, 1.06]	.741	0.95 [0.87, 1.02]	.166	0.96 [0.88, 1.04]	.321
Perception of frequency of peer’s condom use	629	1.03 [0.87, 1.22]	.695	1.07 [0.89, 1.30]	.437	1.03 [0.86, 1.24]	.697

Note. Each analysis was adjusted for gender-, age-, and school-level sexual experience at baseline. AOR = adjusted odds ratio; CI = confidence interval; STI = sexually transmitted infection; HIV = human immunodeficiency virus; AIDS = acquired immune deficiency syndrome.

The results show that the group receiving ¡Cuídate! had a higher level of knowledge about HIV, specifically about condom use (p < .05), other STIs (p < .05), and the total scale (p < .05) compared to the CG at the 24-month follow-up. No significant differences in other psychosocial constructs were observed between ¡Cuídate! and the CG.

Comparing the effectiveness of both interventions, differences were found for the dimension attitude toward people with HIV/AIDS. The group receiving ¡Cuídate! shows a more favorable attitude toward people with HIV/AIDS (p < .05) with respect to COMPAS. For other psychosocial variables, no statistically significant differences between groups are evident between COMPAS and ¡Cuídate!.

Discussion

At the 2-year follow-up evaluation, the COMPAS program did increase the level of knowledge about STIs and promoted a more favorable attitude toward people living with HIV as compared to the CG. The positive effects of the intervention in both constructs were maintained over time, according to the results observed in the evaluation of the program in the posttest (Espada et al., 2015) and at the 12-month follow-up (Morales et al., 2015). However, some of the effects of COMPAS that were evident over the short term and at the 12-month follow-up were not observed at the 24-month assessment. Specifically, there were no differences between the COMPAS group and the CG in the level of knowledge about HIV, condom use, HIV transmission routes, and the attitude for condom use when barriers exist. Other short-term effects observed in COMPAS, such as the increase in the level of knowledge about prevention methods, the perception of risk associated with unprotected sex, the intention to engage in safer sex, and a positive attitude toward HIV testing and condom use, were not maintained at the 24-month follow-up. One year after implementation, COMPAS had a positive effect on the perception of peer condom use and delayed the age of first vaginal intercourse. However, these effects were not observed at the 24-month follow-up.

In a recent study, analyzing the mediating effects of the effectiveness of COMPAS to increase consistent long-term condom use, Escribano, Espada, Morales, and Orgilés (2015) conclude that the attitude of adolescents about condom use when there are barriers and the intention to use condoms are key to increasing the effectiveness of intervention to promote consistent condom use in Spanish adolescents. These results suggest that additional and periodic educational efforts to promote safe sex should be addressed at schools, for example, activities focused on overcoming perceived barriers for condom use and promoting a more favorable attitude toward condom when there are obstacles. This is an opportunity to consider the advantages and disadvantages of using condoms, discuss them with adolescents, and empower them with negotiating skills to have safe sex. The main aim of these activities should be to increase condom use intention, which increases the odds of using condoms when they have sex (Ajzen, 1991).

The results at 24 months showed that ¡Cuídate! increased the level of knowledge about condom use and HIV/STIs in general compared to the CG. The program also had a positive impact on the knowledge about other STIs, coinciding with the results of COMPAS. ¡Cuídate! had no significant impact on the remaining psychosocial variables, such as attitudes toward HIV, risk perception, perceived norm, and intent to engage in safer sexual behaviors nor did it impact any behavioral variables. A study evaluating the effectiveness of ¡Cuídate! at 48 months after implementation in Mexican adolescents via sexual behavior outcomes (Villarruel, Zhou, Gallegos, & Ronis, 2010) indicates that no impact on consistent condom use is evident, nor does it manage to decrease the number of sexual partners, results that are in line with the present study. However, in this study (Villarruel et al., 2010), ¡Cuídate! managed to delay the age of first vaginal intercourse. In this study, ¡Cuídate! also managed to delay the age of onset of oral sex like COMPAS but not compared to the CG. From the results of this study, one can conclude that 24 months after implementation, the effects of the ¡Cuídate! program did not last in many of the psychosocial variables, namely, the attitude toward HIV, intent to engage in safe sex, nor did it delay the age of first intercourse with oral sex, observed in the effects after 12 months (Morales et al., 2015).

The effects of the COMPAS and ¡Cuídate! interventions at the 24-month follow-up on psychosocial variables were equivalent—coinciding with those found in previous studies that compared the efficacy of both interventions in the short term (Espada et al., 2015) and at a 12-month follow-up (Morales et al., 2015). One exception to this was the attitude toward people living with HIV, which was more favorable in adolescents who received the ¡Cuídate! program. Regarding behavioral variables, ¡Cuídate! delayed the onset age of oral sex compared to COMPAS. This difference in the effectiveness of both interventions is not surprising since ¡Cuídate! promotes sexual abstinence as the primary method of protection in sexual relationships, and COMPAS only promotes correct and consistent condom use if sex occurs.

The effect of both interventions on psychosocial variables (risk perception, intention of condom use, etc.) immediately after application (Espada et al., 2012, 2015; Morales et al., 2015) decreased over time, while effects were found in behavioral variables at the 12-month follow-up (age of sexual debut; Morales et al., 2015). This trend is observed in previous studies, such as Pergallo, Gonzalez-Guarda, McCabe, and Cianelli (2012). At the 24-month follow-up, improvements in the level of knowledge about STIs and the attitude toward people infected with HIV were found in both interventions. It is remarkable that in the literature, there are few studies reporting a significant and positive impact on reducing sexual risk in adolescents and young people in the long term (Chen et al., 2010; Gong et al., 2009). The Focus on Youth in the Caribbean intervention, included by the Centers for Disease Control and Prevention as a Best Evidence Program, has been extensively evaluated in longitudinal studies by several authors (Chen et al., 2010; Gong et al., 2009; Wang et al., 2013), showing a high long-term impact in precursors—knowledge, skills, intention, and self-efficacy—and sexual behavior variables. Gong et al. (2009) found an improvement in the level of knowledge, perceptions of their ability of use condoms, behavioral intention, and increased condom use at 24 months after the implementation of an HIV-prevention intervention in a controlled study. Chen et al. (2010) and Wang et al. (2013) show a significant increase in the level of knowledge about HIV/AIDS, self-efficacy, and intention and ability to use condoms at 36 months postevaluation. However, it must be noted that booster sessions were performed after completion of the procedure. Based on the results obtained from these studies and the comparison with the outcomes of the present study, the application of periodic booster sessions seems to be an effective strategy in maintaining the effects of interventions on sexual behavior over time (Chen et al., 2010; Malow, Kershaw, Sipsma, Rosenberg, & Dévieux, 2007; Peragallo et al., 2012).

This study presents several limitations that should be taken into account when interpreting the results. First, the high rate of lost participants at 24 months postintervention can be explained by the high school dropout rate in Spain (26%), which is one of the highest in Europe (Felgueroso, Gutiérrez-Domenech, & Jiménez, 2013), and that a high proportion of the participants had graduated after 2 years. This rate is consistent with other studies evaluating programs, such as Malow, Kershaw, Sipsma, Rosenberg, and Dévieux (2007), who reported a dropout rate of approximately 50% of the participants at the 2- and 5-year follow-ups. The authors indicate that the high dropout rate in longitudinal studies is one of the main reasons why researchers avoid this type of design despite the advantages offered. The low retention rate may have reduced the statistical power of tests of the efficacy of the intervention; therefore, the results should be interpreted with caution. However, the sample across three conditions was equivalent in sociodemographic and sexual experience variables. Also, the CG had higher dropped out percentages than the both interventions groups. We do not have a firm explanation for this result, but it could be explained by the low motivation of students and less implication of schools of CG to get involved in the project, since they did not received any intervention. Second, the sole use of a self-report assessment method is another limitation. Biological measures provide objective measures of program effectiveness (i.e., HIV status, chlamydia, syphilis, and/or gonorrhea). The study sample cannot be considered representative of Spanish adolescents since the participants were recruited from only 5 of the 50 provinces in Spain.

This is the first controlled study conducted in Spain evaluating the long-term efficacy of a school program promoting sexual health, and it also compares its effects to an evidence-based intervention and a CG. With these results, we can conclude that COMPAS maintains the level of knowledge about HIV and other STIs in addition to a favorable attitude toward people living with HIV in the long term. Other effects that were observed in the short term, but faded later, included knowledge about HIV, condom use, HIV transmission routes and prevention methods; attitude toward condom use in general; and when barriers exist, sexual risk perception, and the intention to engage in safe sex. Twelve months after application, the effects remained on knowledge about HIV, condom use and HIV transmission routes, and attitude about condom use, despite barriers for their use, and it is effective for increasing the perceived norm and delaying the age of the first vaginal sexual intercourse. Compared to the CG, COMPAS has a comparable impact to the other evidence-based intervention on variables predicting consistent condom use. When comparing the effectiveness of both interventions, ¡Cuídate! only showed better results than COMPAS in attitude toward people living with HIV, and it also delayed the onset age of oral sex. However, none of the interventions had a significant impact on consistent condom use.

Future research should aim at promoting strategies to resolve problems associated with condom use and the intention to use condoms and to identify others that facilitate long-term consistent condom use. Finally, periodical booster sessions are recommended after the program completion as a possible solution to maintain the effects on sexual behaviors.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Support for this research was provided by the Foundation for Research and Prevention of Aids in Spain, (Ref. FIPSE 360971/10) and by the Program Valid for Research Staff training of the Council of Culture, Education and Science of the Valencian Autonomous Government (Ref. ACIF/2012/132).

References

Ajzen

(1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211.

Bandura

(1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall.

Beadnell

Morrison

D. M.

Wilsdon

Wells

E. A.

Murowchick

Hoppe

… Nahom

(2005). Condom use, frequency of sex, and number of partners: Multidimensional characterization of adolescent sexual risk-taking. Journal of Sex Research, 42, 192–202. doi:10.1080/00224490509552274

Boyce

Torsheim

Currie

Zambon

(2006). The family affluence scale as a measure of national wealth: Validation of an adolescent self-report measure. Social Indicators Research, 78, 473–487.

Carroll

Patterson

Wood

Booth

Rick

Booth

(2007). A conceptual framework for implementation fidelity. Implementation Science [Electronic version], 2. doi:10.1186/1748-5908-2-40

Centers for Disease Control and Prevention. (2014). Compendium of evidence-based HIV behavioral interventions. Risk reduction chapter: CUÍDATE! Retrieved from http://www.cdc.gov

Centro Nacional de Epidemiología. (2014). Vigilancia Epidemiológica del VIH/sida en España: Sistema de Información sobre Nuevos Diagnósticos de VIH y Registro Nacional de Casos de Sida. Actualización 30 Junio de 2014. Retrieved from http://www.msssi.gob.es

Chen

Stanton

Gomez

Lunn

Deveaux

Brathwaite

… Harris

(2010). Effects on condom use of an HIV prevention programme 36 months post-intervention: A cluster randomized controlled trial among Bahamian youth. International Journal of STD & AIDS, 21, 622–630. doi:10.1258/ijsa.2010.010039

Escribano

Espada

J. P.

Morales

Orgilés

(2015). Mediation analysis of an effective sexual health promotion intervention for Spanish adolescents. AIDS and Behavior, 19, 1850–1859. doi:10.1007/s10461-015-1163-2

10.

Escribano

Espada

J. P.

Orgilés

Morales

Influence of fidelity of implementation in the effectiveness of a program sexual health in adolescents. Manuscript in preparation.

11.

Escribano

Morales

Orgilés

Espada

J. P.

(2015). Influencia de la fidelidad de la implementación en la eficacia de un programa de promoción de la salud sexual con adolecentes. Health and Addictions, 15, 103–114.

12.

Espada

J. P.

(2007). Programa COMPAS. Programa educativo sobre el sida. Retrieved from http://aitanainvestigacion.umh.es

13.

Espada

J. P.

Ballester

Huedo-Medina

T. B.

Secades-Villa

Orgilés

Martínez-Lorca

(2013). Development of a new instrument to assess AIDS-related attitudes among Spanish youngsters. Anales de Psicología, 29, 83–89. Retrieved from http://www.redalyc.org/articulo.oa?id=16725574010

14.

Espada

J. P.

Guillén-Riquelme

Morales

Orgilés

Sierra

J. C.

(2014). Validación de una escala de conocimiento sobre el VIH y otras ITS en población adolescente. Atención Primaria, 46, 558–564. doi:10.1016/j.aprim.2014.03.007

15.

Espada

J. P.

Morales

Orgilés

Jemmott

J. B.

III Jemmott

L. S.

(2015). Short-term evaluation of a skill-development sexual education program for Spanish adolescents compared with a well-established program. Journal of Adolescent Health, 56, 30–37. doi:10.1016/j.jadohealth.2014.08.018

16.

Espada

J. P.

Morales

Orgilés

Piqueras

J. A.

Carballo

J. L.

(2013). A review of school HIV/AIDS prevention programs in Spain. International Journal of Hispanic Psychology, 5, 99–112.

17.

Espada

J. P.

Orgilés

Morales

Ballester

Huedo-Medina

T. B.

(2012). Effectiveness of a school HIV/AIDS prevention program for Spanish adolescents. AIDS Education and Prevention, 24, 500–513. doi:10.1521/aeap.2012.24.6.500

18.

Felgueroso

Gutiérrez-Domenech

Jiménez

M. S

. (2013) ¿Por qué el abandono escolar se ha mantenido tan elevado en España en las últimas dos décadas? El papel de la ley de educación (LOGSE). FEDEA. Retrieved from http://www.fedea.net

19.

Fishbein

Ajzen

(1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Wesley-Addison.

20.

Fisher

J. D.

Fisher

W. A.

(1992). Changing AIDS-risk behavior. Psychological Bulletin, 111, 455–474.

21.

Gong

Stanton

Lunn

Deveaux

Marshall

… Chen

(2009). Effects through 24 months of an HIV/AIDS prevention intervention program based on protection motivation theory among preadolescents in the Bahamas. Pediatrics, 123, e917–e918. doi:10.1542/peds.2008-2363

22.

Hernández-Martínez

García-Serrano

Simón-Hernández

Coy

García-Fernández

A. M.

Liante

… Mateos

(2009). Efectividad de un programa de educación sexual en adolescentes acerca de la adquisición de conocimientos y cambios de actitud ante el empleo de métodos anticonceptivos. Enfermería Clínica, 19, 121–128.

23.

Jemmott

J. B.

III Jemmott

L. S.

Fong

G. T.

(1992). Reductions in HIV risk-associated sexual behaviors among black male adolescents: Effects of an AIDS prevention intervention. American Journal of Public Health, 82, 372–377.

24.

Jemmott

J. B.

III. Jemmott

L. S.

Fong

G. T.

McCafree

(1999). Reducing HIV risk-associated sexual behavior among African American adolescents: Testing the generality of intervention effects. American Journal of Community Psychology, 27(2), 161–187.

25.

Jemmott

J. B.

III Jemmott

L. S.

Fong

G. T.

(2010). Efficacy of a theory-based abstinence-only intervention over 24 months: A randomized controlled trial with young adolescents. Archives of Pediatrics and Adolescent Medicine, 164, 152–159.

26.

Jemmott

J. B.

Jemmott

L. S.

O’Leary

Ngwane

Icard

L. D.

Bellamy

S. L.

… Makiwane

M. B.

(2010). School-based randomized controlled trial of an HIV/STD risk-reduction intervention for South African adolescents. Archives of Pediatrics & Adolescent Medicine, 164, 923–929. doi:10.1001/archpediatrics.2010.176

27.

Liang

K. Y.

Zeger

S. L.

(1986). Longitudinal data analysis using generalized linear models. Biometrika, 73, 13–22.

28.

Malow

R. M.

Kershaw

Sipsma

Rosenberg

Dévieux

J. G.

(2007). HIV preventive interventions for adolescents: A look back and ahead. Current HIV/AIDS Reports, 4, 173–180.

29.

Ministerio de sanidad, Servicios sociales e Igualdad. (2015). Interrupción voluntaria de embarazo. Datos definitivos correspondientes al año 2013. Retrieved from http://www.msssi.gob.es

30.

Ministerio de Sanidad, Servicios sociales e Igualdad. (2009). ICAP 2009. Informe del cuestionario de actividades de prevención del VIH en las comunidades autónomas. Retrieved from http://www.msssi.gob.es+

31.

Ministerio de Sanidad y Consumo. (2008). Criterios de calidad para el desarrollo de proyectos y actuaciones de promoción y educación para la salud en el sistema educativo: Guía para las administraciones educativas y sanitarias. Retrieved from http://www.msssi.gob.es

32.

Monsalve

(2012). La Promoción y Educación para la Salud desarrollada desde los centros educativos como herramienta clave de las intervenciones de salud. Revista Internacional de Humanidades, 2. Retrieved from http://www.academia.edu

33.

Morales

Espada

J. P.

Orgilés

(2015). A 1-year follow-up evaluation of a sexual-health education program for Spanish adolescents. European Journal of Public Health. doi:10.1093/eurpub/ckv074

34.

Morales

Espada

J. P.

Orgilés

Secades-Villa

Remor

(2014). The short-term impact of peers as co-facilitators of an HIV prevention programme for adolescents: A cluster randomised controlled trial. The European Journal of Contraception and Reproductive Health Care, 19, 379–391. doi:10.3109/13625187.2014.919445

35.

Pergallo

Gonzalez-Guarda

R. M.

McCabe

B. E.

Cianelli

(2012). The efficacy of an HIV risk reduction intervention for Hispanic women. AIDS and Behavior, 16, 1316–1326. doi:10.1007/s10461-011-0052-6

36.

Pettifor

A. E.

Van der Straten

Dunbar

M. S.

Shi-Boski

S. C.

Padian

N. S.

(2004). Early age of first sex: A risk factor for HIV infection among women in Zimbabwe. AIDS, 18, 1435–1442.

37.

Rosenthal

Rosnow

R. L.

(1985). Contrast analysis: Focused comparisons in the analysis of variance. Cambridge, MA: Cambridge University Press.

38.

Villarruel

A. M.

Jemmott

J. B.

III Jemmott

L. S.

(2006). A randomized controlled trial testing an HIV prevention intervention for Latino youth. Archives of Pediatric and Adolescents Medicine, 160, 772–777. doi:10.1001/archpedi.160.8.772

39.

Villarruel

A. M.

Zhou

Gallegos

E. C.

Ronis

D. L.

(2010). Examining long-term effects of Cuídate —A sexual risk reduction program in Mexican youth. Revista Panamericana de Salud Pública, 27, 345–351.

40.

Wang

Stanton

Chen

Dinaj-Koci

Braithwaite

… Lunn

(2013). Predictor of responsiveness among early adolescents to a school-based risk reduction intervention over 3 years. AIDS and Behavior, 17, 1096–1104. doi:10.1007/s10461-012-0144-y