Quality improvement as a framework for behavior change interventions in HIV-predisposed communities: a case of adolescent girls and young women in northern Uganda

This paper highlights the problem of risky sexual behavior that present a greater risk to HIV infection and transmission of HIV among adolescent girls and young women (AGYW). We use the terms ‘adolescent girls and young women’ to refer to ages 10–24 being the target group of an evidence based behavioral intervention. The effect of a quality improvement for behavior change (QBC) model on reducing risky sexual behavior among a high risk population of AGYW in northern Uganda is explained.

AGYW in the post conflict northern Uganda experience at least three or more HIV risk factors every day. They are victims of gender-based violence including rape, experience cross generational sex, engage in transactional sex, have multiple sexual partnerships, live in absolute poverty-stricken households, are grossly illiterate, and abuse drugs. The risk is even higher with changing individual-level behaviors that characterize AGYW as they transition from childhood to adulthood. The risk of exposure to various risky behaviors even increases further in societies with diminished stability [1].

According to the AIDS indicator survey [2], northern Uganda has HIV prevalence of 10.1% among women aged 15–49 and HIV prevalence among young people is markedly higher among young women than among young men and women become infected at younger ages than men [2]. Sexual transmission accounts for 76% of new HIV infections and is partly attributed to inconsistent and incorrect use of condoms, casual sex, multiple sex partners and extramarital sex [2]. The risk of acquiring HIV applies to AGYW with multiple partnerships, as well as to those with male partners who have (or have had) multiple partners [2]. Vulnerability to HIV infection is especially higher during forced or coercive sex due to the increased chance of vaginal tearing, especially in younger women whose vaginas are not mature. In the extreme circumstances, family members and partners are the perpetuators of violence on the victims including rape [3]. The World Health Organization (WHO) reports that globally, over one-third of women have experienced some form of physical or sexual violence [3].

A critical constraint experienced by young people wishing to practice safer heterosexual intercourse consists of the degree of control they have within sexual encounters. In general, young women have far less control over their sexual encounters than do young men. One obvious way power is manifest in heterosexual relationships is through the presence or threat of violence against young women by young men. Over the past decade strong evidence has emerged on the relationship between partner violence and HIV. There is equally strong evidence for and recognition of successful community strategies to prevent intimate partner violence and vulnerability to HIV [4]. In northern Uganda, for example, the legacy of violence due to political conflict embedded violence as a normal part of gendered relation including rape.

The exploration and formation of identity is considered by many to be the primary developmental goal of adolescence. Socially, adolescents are searching for a sense of belonging from peers, who influence behavior. These milestones occur during a period of decreased adult supervision when young people still have limited knowledge, self-confidence and life skills, which can lead to engagement in behaviors that heighten HIV risk [5]. In the context of HIV prevention, it is essential for parents to be aware of the unique biological, social and emotional challenges of adolescence in order to skillfully help their child navigate their way through adolescence to adulthood, in addition to meeting their financial needs. Optimal parenting requires flexibility and adaptability to meet the particular needs of an individual child or adolescent in a developmentally-appropriate manner. Several studies have demonstrated a link between young people’s sexual behavior and levels of parental monitoring and parent–child communication in Western countries. However, little is known about this association in African settings, especially among young people living in high poverty settings [6].

Adolescent behavior is influenced by family, peers, school, neighborhood, and the broader socio-cultural context [7]. Parents are positioned to assume a central role in protecting against adolescent involvement in risk behaviors [8]. Parents may exert their influence on adolescent risk behaviors through different pathways including parent-adolescent communication and parental monitoring [9]. Studies have shown that parent-adolescent communication about sexual behavior especially communication before the onset of youth sexual activity promotes healthy sexual decision-making and decreases adolescent involvement in sexual risk behaviors [10].

Many interventions focus on changing individual-level behaviors rather than addressing the larger contextual and structural landscape within which young people live. In response to this concern, USAID ASSIST Project rolled out the QBC model using the community-based quality improvement (QI) teams under the “Determined Resilient Empowered AIDS-free Mentored and Safe” (DREAMS) initiative in the post-conflict northern Uganda districts of Gulu and Omoro (Acholi sub region) and Oyam and Lira (Lango sub region).

The regions was a battleground for the Lord’s Resistance Army (LRA), also known as the Lord’s Resistance Movement, a rebel group which operated in northern Uganda for more than 2 decades and later in South Sudan, the Central African Republic, and the Democratic Republic of the Congo. It was originally known as the United Holy Salvation Army and largely functioned as a personality cult of its leader Joseph Kony a self-declared prophet. The rebel activities were characterized by human rights violations, including murder, abduction, mutilation, child-sex slavery, and forcing children to participate in hostilities [11]. As a result, the population in the region lived in concentrated camps and inhumane conditions that facilitated HIV infection and transmission.

DREAMS is a multi-country initiative supported by the United States President’s Emergency Plan For AIDS Relief (PEPFAR) that brings together a combination of interventions already known to prevent HIV among AGYW [12]. DREAMS targets AGYW in ten Sub-Saharan African countries where nearly half of all new HIV infections occurred among AGYW globally in 2014 [12]. However, the success of known interventions is affected by the model of delivery used to guarantee the right target group receives all components of the intervention [12].

The United States Agency for International Development (USAID) Applying Science to Strengthen and Improve Systems (ASSIST) Project tested the QBC model over a 2-year period among a high-risk population of AGYW that experience at least three or more HIV risk factors in life such as gender-based violence (GBV) including rape, cross-generational sex and early marriage, transactional sex, multiple sexual partnerships and drug abuse [13]. The model comprises three core interventions: (1) skills building and peer to peer support in topics such as saying no to sex and negotiating condom use to improve the ability of AGYW to stop risky behavior and to influence safer sex practices with their partners; (2) setting up and making functional community QI teams to mobilize community resources to support the young women and their partners to stop risky behavior; and (3) holding service delivery camps to provide HIV prevention services and commodities to AGYW, their partners and other community members mobilized by the community QI teams.

The QI team members were local community members from the same cultural and social context as the AGYW target population, including community elders, religious leaders, peer facilitators, local council members, Village Health Teams (VHTs), health workers, and teachers. Peer facilitators were AGYW from different communities that facilitated skills building and peer to peer support sessions and used their life experiences to relate to AGYW in similar circumstances. Most at risk AGYW were identified using ‘The Girls Roster toolkit.’ The toolkit is a practical tool to help increase girls especially poor adolescent girls’ access to vital resources, facilities, and services. It helps in understanding the full universe of girls in a program catchment area; breaking them into meaningful segments by age, by schooling, and by marital, childbearing, and living-arrangement status; and, through appropriate community engagement, increasing girls’ access to a fair share of community resources.

The USAID Strengthening Decentralization for Sustainability (SDS) project supported the processes of training community resource persons to use the tool and to map priority communities. The ASSIST project randomly selected 12 of the priority communities to demonstrate implementation of QI for HIV prevention using the QBC model. Within the 12 communities all the AGYW mapped and registered using the Girls Roster tool were engaged in the ASSIST supported activities. QI teams identified AGYW in their communities by their various risks such as gender-based violence (GBV) including known rape victims, cross-generational sex and early marriage, transactional sex, multiple sexual partnerships and drug abuse. Teams worked with AGYW to develop risk reduction plans consonant with the norms, culture, and aspirations of the target population and followed up with the AGYW to support behavioral change using a set of messages that were tailored to the specific needs of the AGYW. Participatory approaches to behavioral change communication were used, such as sharing of experiences, problem identification, and personalized risk reduction plans.

QI teams worked with facility teams to create demand for HIV prevention services through quarterly integrated health service camps for AGYW and partners of young women to improve access to a standard care package for HIV prevention including HIV counseling and testing (HCT), screening and treatment of sexually transmitted infections (STIs), family planning services of their choice, condom awareness and distribution, risk reduction counselling, and provision of tetanus toxoid to partners in preparation for voluntary medical male circumcision (VMMC). According to Ministry of Health [2], HCT is the entry point to HIV prevention. There is also an epidemiological relationship between STIs and HIV infection. A study conducted by Sensalire et al. [14] showed that awareness of HIV prevention through the use of condoms increased the likelihood of use by four times among adolescents. According to the World Health Organization [15], medical male circumcision reduces the risk of female-to-male sexual transmission of HIV by approximately 60%.

Quality improvement teams also targeted parents of all AGYW irrespective of whether they were minors or not. Several studies have demonstrated a link between young people’s sexual behavior and exposure to information and levels of parental monitoring and parent–child communication in Western countries [6]. In the context of a behavior change intervention, it was essential for parents to be aware of the unique biological, social, and emotional challenges of adolescence in order to skillfully help their child navigate their way through adolescence to adulthood, in addition to meeting their financial needs. The involvement of parents under the intervention was based on the hypothesis that good parenting reduces sexual risk behavior among adolescents. QI teams mobilized parents and caregivers of AGYW for parenting training sessions with the aim of empowering them with knowledge and skills to better engage and communicate with their children and support them reduce their risk of HIV sustainably.

In addition to AGYW and their parents, the QI teams also involved partners of young women. They were mobilized for partners’ meetings by the young women themselves and partly by the QI teams in instances where the partner was known to them by virtue of being community members in their close proximity. In some instances, partners are the perpetuators of violence on the victims, including rape and at baseline 10% of the AGYW had experienced sexual violence from partners. QI teams held meetings with partners to identify risk factors and mechanisms to address them. The partner meetings addressed the risk of multiple sexual partners, unprotected sex, domestic violence of all forms, unwanted pregnancies, social and financial empowerment of partners and repugnant norms that risk to HIV infection. Partners therefore, became change agents in the communities.

A critical constraint experienced by young people wishing to practice safer heterosexual intercourse consists of the degree of control they have within sexual encounters. In general, young women have far less control over their sexual encounters than do young men. One obvious way power is manifested in heterosexual relationships is through the presence or threat of violence against young women by male partners [16]. Partner violence increases the risk of HIV, and there is strong evidence for and recognition of successful community strategies to prevent intimate partner violence and vulnerability to HIV [16]. In northern Uganda, due to the prolonged political conflict, GBV, including rape, is common. This intervention recognized the connection between HIV infection and gender violence and therefore, involved partners and parents.

The aggregate effect of radical and sustained behavioral changes in a sufficient number of individuals potentially at risk is needed for successful reductions in HIV transmission. Second, the effect of behavioral strategies could be increased by aiming for many goals (e.g., delay in onset of first intercourse, reduction in number of sexual partners, increases in condom use, etc.) that are achieved by use of multilevel approaches (e.g., couples, families, social and sexual networks). Third, reductions in HIV transmission need a mix of communication channels to disseminate messages to motivate people to engage in a range of options to reduce risk [17].

The QBC intervention addressed multiple sexual partners, inconsistent use of condoms in high risk sex and transactional sex among AGYW through skills building and peer to peer support to abstinence, negotiating condom use; setting up QI teams and holding service delivery camps to provide HIV prevention services and commodities to AGYW, their partners and community mobilization. These multilevel approaches following the intervention in target population, resulted into a significant decline in risky sexual behavior, which suggests the appropriateness of the QBC model in addressing HIV risk in similar settings.

Having multiple sexual partners and transactional sex is one of the important behavioral HIV risk factors that were addressed using the QBC model intervention. Studies have shown that frequent change of sexual partners is associated with higher HIV risk and HIV incidence among young sex workers and sexually exploited adolescent girls remains high in many settings in Sub-Saharan Africa [18]. The study found that transactional sex and multiple sexual partnerships was at 23 and 17% at baseline respectively. In order to lower the risk of HIV by reducing the number of sexual partners, a better understanding of the factors which affect young women’s risk behaviors is needed [19]. Thus peer to peer approach was used to determine the driving factors for multiple sexual partnerships and transactional sex so as to deploy socially and contextually appropriate behavioral change interventions. This resulted into a reduction in the number of AGYW engaging in risky sexual behaviors as compared to the baseline, thus suggesting a need for mix of individual and contextual approaches to effectively tackle the risk of HIV among high risk AGYW.

We investigated exposure to violence by AGYW and found that 10% had been exposed to sexual violence by their partners. In northern Uganda, the legacy of violence due to political conflict embedded violence as a normal part of gendered relations, including rape. Studies have shown that a critical constraint experienced by young people wishing to practice safer heterosexual intercourse consists of the degree of control they have within sexual encounters. Young women have far less control over their sexual encounters than do young men and experience sexual violence [4]. The intervention therefore, targeted partners of the AGYW midway the intervention. The decline in sexual violence to 2.3% and other forms of violence following the intervention among male partners suggest, that male partners can be agents of behavioral change themselves. It also suggests that targeting AGYW alone without their partners may not eliminate partner-driven risk behavior. This suggests complementary interventions that focus both young girls and their partners are more effective than interventions that target only AGYW.

Parents have the potential to protect against adolescent sexual risk, including early sexual behavior, inconsistent condom use, and outcomes such as pregnancy and STIs. Identification of the specific parenting dimensions associated with sexual risk in adolescence and young adulthood is necessary to inform and focus prevention efforts [20]. We found that engaging parents/guardians of the AGYW resulted in positive behavioral changes, notably a reduction in transactional sex. There was a statistical association between parental support and the reduction in the proportion of AGYW in transactional sex. Parental support was mostly in the form of financial support to the AGYW. This underscores the value of family interventions in sexual risk prevention. Several studies have demonstrated a link between young people’s sexual behavior and levels of parental monitoring, parent–child communication, and parental discipline. The importance of parental involvement is often underscored in risk-reduction efforts targeted towards adolescents because parents/guardians are in regular contact with their children and, therefore, presumably in a good position to shape and influence young people’s behaviors [8].

According to Sensalire et al. adolescent girls are in a stage of social, psychological, economic, and biological changes [1]. The hardships surrounding the intervention communities such as poverty, GBV further activate adolescent risk lifestyles. The findings show that as interventions addressing specific risk factors unfolded, risk behavior declined. This contextualized and focused interventions have more effect on adolescent sexual behavior in as far as they focus on particular risks appropriate to the target population at a time. This means that interventions for prevention of HIV among adolescents need to focus on specific behaviors prevailing in that community rather than on generalized communication.

Results further indicated a decline in unwanted teenage pregnancy at follow up since all AGWY who were pregnant desired to have the pregnancy at the time they conceived compared to the baseline. The results give rise to two important considerations. First, AGYW who were pregnant are at high risk of HIV by virtue of unprotected sex. Secondly the findings verify low contraceptive use.

As the paper notes, everyone was a year older and a third of them were married at the end of the study. Perhaps the change in sexual behavior of AGYW following the roll out of the QBC model was reinforced with marriage or the more rational decision making that comes with maturity. However, this would only apply to a small proportion of those who were married implying that the majority of AGYW sexual behavior was due to interventions related with QBC model, other factors constant.

The QBC model utilized a participatory approach and involved various community resource persons including peers to address risks associated with HIV. These were known as QI teams. The decline in risky sexual behavior, namely, reduction in multiple sexual partnerships, and transactional sex among AGYW, within a relatively short period of time implies that behavioral interventions involving the community are effective, possibly because of the close proximity to the target group and the general understanding of the socio-cultural factors that drive infection. Broad-based community participation in a behavioral intervention facilitates the selection and use of culturally appropriate strategies to reach the target audience.

The study had some limitations. First, it is unclear whether similar processes exist for boys since the sample was comprised entirely of girls. Second, our study applied to a longitudinal sample but without a control group. Thus, it’s in some way unknown whether other factors could have contributed to changes in sexual behavior. Although this study provides some evidence of change of sexual behavior following the roll out of the quality improvement for behavior change model (QBC), the study did not test for sustainable change of the desired behavior in the long run given the short study duration. All the behaviors are self-reported. It could be that the participants learned quite clearly during the intervention what they should be doing and may have given the response they thought the interviewers wanted to hear. However this was minimized through an independent assessment of the cohort’s sexual behavior and the use of multiple checks in the data collection tools.

In light of these limitations, future research should aim to examine QBC models for girls and boys and with a control group to rule out existence of other factors that could influence short term changes in sexual behavior. The longitudinal study designs should provide reasonable time to determine sustainable behavior change.

Scale-up of the QBC intervention to reduce or prevent HIV infection among AGYW should be supported because it offers a feasible solution within the reach of the target population. Its success however depends on use of a participatory approach and involvement of the community. Therefore, interventions aimed at reducing or preventing HIV infection among AGYW should involve peers and respected members of the community working as a team. Quality improvement approach is a fundamental tool in as far as the change strategies and activities are developed and implemented in a way that is appropriate to the target groups and behaviors of interest.