Young people aged 15 to 24 years comprise half of all new sexually transmitted infections (STIs) in the United States, despite the fact that this age group accounts for just 27% of the sexually active population [1]. The Centers for Disease Control and Prevention recently reported that 25% of sexually active adolescent girls currently have an STI [1]. They also found that young women aged 15 to 24 years had the highest reported cases of chlamydia and gonorrhea, and that there was a 7.8% increase from 2016 in reported cases of primary and secondary syphilis. STIs are associated with significant morbidity and mortality, including pelvic inflammatory disease and infertility in women [2]. Adolescent girls are particularly vulnerable to STI acquisition compared with their male peers, as they face unique biological, social, and cultural vulnerabilities, such as increased cervical ectopy [1,3] and decreased power over sexual relationships [4,5]. There is a considerable need for a new paradigm in adolescent STI prevention efforts, particularly for girls, given the current data that suggest that prevention efforts do not seem to be effective.

The co-occurrence of alcohol use and sexual risk taking contribute significantly to STI acquisition, particularly in vulnerable populations, such as adolescent girls [6]. Alcohol continues to be the most widely abused substance by adolescents in the United States [7]. National survey data show that 42% of 10th graders and 62% of 12th graders reported lifetime use of alcohol [7], and 11% of females aged 12 to 17 years reported alcohol use in the last 30 days compared with 8.8% of males [8]. Adolescent girls who use alcohol and other substances are disproportionately more likely to engage in risky sexual behaviors that can result in STIs [6]. Owing to the interrelatedness of alcohol use and STI acquisition, and because both are associated with poor health outcomes, addressing alcohol use as a factor in sexual risk behavior may be particularly helpful for the development of sexual health interventions.

A number of evidence-based, sexual health programs delivered in school settings have demonstrated effectiveness in reducing risks during adolescence, including delaying sexual activity and increasing the consistent use of contraceptives [9]; however, there are structural and financial obstacles to the dissemination of these programs, and to maintain behavior changes. Furthermore, national surveys indicate that 83% of sexually experienced girls aged 15 to 17 years did not receive any formal sex education in school until after their first sexual experience [10].

Sexual health knowledge is lacking among adolescents. One study found that 65% of 14 to 21-year-olds were aware of hepatitis, 57% were aware of human papillomavirus, and only 6.5% were aware of trichomoniasis [11]. In a recent study, adolescent girls who had been diagnosed with an STI were more knowledgeable about that particular STI compared with girls who had not been; however, their knowledge of other STIs were similar to girls who had never been diagnosed with one, suggesting that many adolescents do not learn about STIs until after they are diagnosed with one [12]. Limited STI knowledge among adolescents is associated with inconsistent condom use [13,14], multiple sex partners, delaying disease treatment, and failing to return for STI testing results [12]. There is a lack of STI testing among adolescents and young adults aged 15 to 25 years, with 42% of sexually active adolescents reporting not testing because they said they were “not at risk for an STIs”—an alarming statistic for a population that accounts for half of all new STI diagnoses [1,15].

Mobile health (mHealth) interventions use mobile phones or other wireless technology to disseminate health information in an effort to alter health behaviors and outcomes [16]. Owing to widespread smartphone and internet use, mHealth interventions are ideally suited for adolescents. A recent (2018) national survey noted that 9 out 10 teens go online multiple times daily, with 45% of teens saying they are online almost constantly [17]. Smartphone access was also high, with 95% of teens reporting they had access to a smartphone [17]. Content for an mHealth app can be standardized, tailored, and presented through interactive features, giving it the potential to have a wide reach, potentially leading to a greater impact at a lower cost [18]. Overall, mHealth interventions have demonstrated increased sexual health knowledge, increased STI testing and communication with clinicians about STI risk behavior, higher rates of condom and contraception use, lower rates of unprotected sex, and reductions in alcohol use among adolescents [16,19-21].

Despite their promise, there is still a dearth of knowledge about efficacy, communication best practices for mHealth interventions, and a need for further research into effective content creation and health promotion messaging [16,22]. Furthermore, there is a lack of evidence-based preventive interventions that specifically target both sexual health and alcohol use among adolescent girls aged 15 to 19 years, a particularly vulnerable group [23,24]. This pilot study sought to fill these gaps by conducting in-depth individual interviews and anonymous surveys with female adolescents to (1) understand attitudes about sexual health and alcohol use, and (2) to determine preferences for content, messaging, and format for an mHealth app aimed at reducing risky sexual health behaviors and alcohol use.

This pilot study used both qualitative and quantitative methods to explore the views of adolescent girls on the desired qualities and content for an mHealth app on sexual health and alcohol use. The first phase of the study was qualitative, in which we conducted in-depth interviews (IDIs), and the second phase was quantitative, in which we conducted web-based surveys using a social media–recruited sample. We recruited adolescent girls aged 15 to 19 years who met study inclusion criteria to participate in a 1-hour focus group. Inclusion criteria included endorsing sexual risk behavior (eg, unprotected sex) and alcohol use within the last 30 days. Owing to recruitment challenges, including a lack of study screener completion, lack of girls reporting both risk factors, and difficulty scheduling participants for focus groups, we changed the study format to individual IDIs and modified our inclusion criteria to include adolescent girls aged 15 to 19 years who endorsed either a sexual behavior risk (eg, unprotected sex) or an alcohol behavior risk in the last 30 days. This study was approved by the Institutional Review Board of the University of Michigan.

Recruitment for the focus groups took place between August 2017 and December 2017, and IDI recruitment took place between January 2018 and April 2018. Recruitment for the web-based survey occurred in July 2018. For the qualitative phase of this study, we used both active and passive recruitment methods, including onsite recruitment at an urban family medicine clinic that is the primary care provider for an underserved population of children, adolescents, and adults. Approximately 55% of the clinic’s patients identify as Latina or African American. Participants were recruited in the waiting room before their scheduled appointment. Participants were approached, the study was described, and if they expressed interest, they were asked to take the screener. Interested participants provided verbal assent for the screener and received a small gift (eg, lip balm) for participating. Eligible girls were invited to participate in an individual IDI. Participants received US $30 cash for completing the IDI.

We included passive recruitment to the study design to enhance recruitment efforts. Passive recruitment included advertising on a large public university campus in the Midwest, at local clinics, and on college listservs in a broader population area compared with active recruitment. Interested participants reached out via email and received a link to the screener; those who met inclusion criteria were invited to participate in an IDI.

In the second phase of the study, we tested the feasibility of using social media (eg, Facebook) to recruit a national sample of at-risk adolescent girls for research studies, collect survey data on sexual health risk and substance use, and gauge interest in a potential mobile phone app that addresses these issues. We ran social media ads on Facebook and gathered survey data for a period of 2 weeks. We targeted the ad by age (15-19 years old), gender (female), location (the United States), demographics (eg, education level, life events, and relationship status), interests (eg, entertainment and pop culture), and behaviors (eg, shopping). The survey was conducted through Qualtrics software (Qualtrics, Provo, UT) with no confidential information such as names, addresses, or Internet Protocol addresses collected. Participants who completed the social media survey did not receive an incentive.

Background characteristics of the IDI participants are shown in Table 1. Recruitment occurred in two separate phases. During the first recruitment phase, we approached 61 girls through active (clinic-based) recruitment, screening a total of 50 girls over the course of 5 months. Of the 50 girls screened, only 7 (14%) were eligible, but no eligible girls participated in a focus group discussion.

Phase 2 of recruitment was conducted both via active (clinic-based) and passive recruitment strategies. A total of 15 girls were approached during active recruitment, of which 7 completed the screener, and 4 (57%) were eligible. Through our passive recruitment strategies, 19 girls contacted us, 18 completed the screener, and 50% (9/18) were eligible. All 9 of these participants completed an IDI. A CRAFFT score was determined using the screener data for each participant. The range of scores was from 0 to 4, with a score of 4 or more indicating possible substance dependence.

Participant characteristic results from the web-based survey are presented in Table 2. Respondents were asked questions on their relationship status and condom use, as well as their alcohol and drug use (see Table 3). The majority of respondents (86%; 314/367) reported that they would use a sexual health phone app for adolescent girls. Surveys were considered complete and included in the analysis if all questions were viewed by the participants. Participants could choose not to respond to questions that made them feel uncomfortable, and their surveys were still included in the analysis.

Data from the nine IDIs resulted in four major interrelated themes described in the sections below. These themes included alcohol and college life, alcohol and condomless sex, STIs and pregnancy, and technology use and app preferences.

Participants overwhelmingly said that they avoided messages in the don’t drink genre. They saw them as ineffective and were generally dismissed because they are seen as being unrealistic and out-of-touch:

Including practical messaging to reduce alcohol consumption was thought to be the most effective approach, such as highlighting the benefits of drinking less (ie you can still have fun but avoid hangovers). Health and safety came up frequently when discussing alcohol-related topics that participants would like to see in an app. Specifically, they said they want to learn about healthy levels of drinking versus problem drinking, and how to gage and manage their own personal alcohol limits. They recommended that our app contain a drink-tracking feature like the university’s app discussed above. They also said it would be helpful to include a centralized list of resources that provide substance use help.

Participants resoundingly said that they wanted an app that focuses on helping young women, not on shaming them. To them, this meant that the app would use positive language, and would validate young women’s desire to try alcohol as something that was normal and not as something that was bad or immoral. Validating that it was equally normal to not want to drink was something they asked to include as well as emphasizing that if they did choose not to drink, they would not be alone in their choice.

Similar to alcohol-related messaging, participants said that the most effective way to talk to young women about sexual health was to be factual in an empowering manner, not in a moralizing or judgmental one. One participant said that when issues of women’s sexual health were put in a moralizing manner, it was too “easy to cross into the slut shaming line,” which to her, made them ineffective. She further discussed how she disliked her high school sexual education class because it centered around scare tactics:

When we asked where young women went for information on STIs, most participants said that they generally did not search for information on STIs. They said that STIs were just not something people their age worried much about. Overall, participants desired detailed information on STIs, not just how to prevent them. They wanted to know how they can affect your body and overall health, which ones can be cured and which ones cannot, how the testing process works for different STIs, and what the implications are for their sexual partner(s) with a positive test.

Participants explained that an app should also include information on sexual consent, coercion, and issues of sexual assault and how to prevent it. As many women become sexually active during adolescence and into college, they thought it was very important to include the definition of consent and how to give it. They also wanted the app to have information on how to recognize and get out of potentially dangerous situations, and steps to mitigate and navigate a sexual assault situation. Finally, participants said they wanted empowering messages about their sexuality and their bodies. They said that young women want to be in control of their sex lives without experiencing stigmatizing labels. They want to be comfortable about their sexuality and comfortable using condoms. As one young woman put it:

This pilot study used both qualitative and quantitative methods to explore the views of adolescent girls on the desired qualities and content for an mHealth app on sexual health and alcohol use. Overall, the results of this pilot work were encouraging. We identified and recruited adolescent girls who endorsed recent alcohol use and sexual risk behavior to participate in the individual IDIs. Our study was feasible with respect to the study procedures, and passive recruitment methods were most effective in identifying girls who endorsed these risks. Moreover, we found that using social media to recruit and engage adolescent girls who endorsed high-risk behaviors was feasible and acceptable. Of the 382 individuals who started the survey, 367 (96.0%) completed it. Moreover, 86% (314/367) endorsed that they would use a sexual health app for girls.

Several themes emerged from the IDIs, with the major themes involving alcohol and college life, alcohol and condomless sex, STIs and pregnancy, and technology use and preferences. Participants offered, from their personal experiences as well as those of others, their perspectives supporting that alcohol use and sexual risk behavior are closely connected in this phase of life. The co-occurrence of alcohol use and sexual risk taking contribute significantly to STI acquisition, particularly in vulnerable populations, such as adolescent girls.

There was a general sense of inexperience with STIs and, subsequently, a lack of knowledge on prevention and vulnerability. Two very clear themes emerged during our interviews with respect to our questions on STIs and pregnancy. First, young women are largely ignorant of STIs and their health effects, and second, young women are relatively unconcerned about STIs, but rather, extremely concerned about pregnancy. The deep-seated concern about pregnancy was not only because of the significant life change it would bring but also because of the moral implications in a society where having a teenage pregnancy brings harsh judgment and shame. STIs were seen as something that could easily be hidden if you contracted one, but pregnancy was something out in the open. In our IDIs, we found participants to be open about recommendations for a sexual health app, for example, “if you want to talk to young people about STIs, focus on why young people should prevent them, not just how they can prevent them.”

The use of mHealth interventions to target sexual health and substance use among youth holds promise in reaching large numbers of at-risk adolescents remotely and to tailor content to individual user needs. In 2017, a systematic review of mHealth interventions for alcohol/drug use prevention among young adults found mixed results among the individual studies, but an overall support for the efficacy of mHealth interventions in reducing substance use. They included a variety of mHealth modalities such as text messaging, web-based apps, and smartphone apps. The limitations of these studies were largely because of small sample sizes, lack of data on long-term effects, lack of examination into potential gender differences in intervention effects, and lack of diversity in participant age and environment [27].

Our review of federal registries for evidence-based interventions (EBIs) on adolescent substance use prevention and adolescent sexual health found none that focused on both of these health risks simultaneously for nontreatment seeking or nonincarcerated adolescent girls [23,24]. While there were several EBIs focused on adolescent girls’ sexual health [28-33], there was a dearth of EBIs for substance use prevention in the 15 to 19 age range, and none that focused on young women. Furthermore, the sexual health EBIs were either specific to clinic-seeking, at-risk, or racial and ethnic minority adolescent girls. The sexual health EBIs with the most promising results still faced the limitation of reach and scalability; none of them were mHealth, and all relied on a clinic, classroom, or community-based setting with trained facilitators [28-33]. Not only would a mobile sexual health app be able to reach more adolescents but it would also be more cost-effective and provide greater impact.

This study included a number of strengths. We explored views on a topic of great public health significance and successfully engaged youth to share their attitudes and feelings on sensitive topics. We used social media as a tool to engage high-risk adolescents to collect informative data, demonstrating that this is a suitable platform for our future work. Owing to our study population comprising of college-enrolled female adolescents, there are still gaps in understanding the preferences and needs of female adolescents from disadvantaged backgrounds and the results cannot be generalized to the population at large. The inability of this study to successfully recruit girls to complete focus groups exemplifies the challenges that need to be understood among this population to best reach them effectively.

This pilot study demonstrates a clear need to reach, inform, and empower adolescent girls on the topics of sexual health and alcohol/drug use. Given the escalating rates of STI acquisition that continues to affect this vulnerable group, addressing the co-occurrence of alcohol/drug use and sexual risk taking is a research area of high priority, and also for which data are insufficient and scarce. Future work will integrate these areas to inform the development of a culturally sensitive mHealth app to empower adolescent girls and to reduce health risks and prevent STIs among them.