Sexual assault—defined as forced, coerced, or incapacitated sexual contact—is a common form of trauma among women. According to epidemiological data, 27% to 44% of women are sexually assaulted in their lifetime, and 2.2% to 4.7% are sexually assaulted in a given year [1,2].

Sexual assault is associated with increased risk for both posttraumatic stress (PTS) and alcohol misuse [3]. Approximately 1 in 4 survivors meets criteria for past-year PTS, and 1 in 7 meets criteria for past-year alcohol use disorder [4]. These conditions frequently co-occur in survivors [5], which is concerning given evidence that people with comorbid PTS and alcohol use disorder have poorer treatment engagement and outcomes than people with either disorder alone [6-8]. The co-occurrence of these conditions has been attributed to a functional relationship such that alcohol may be used to cope with PTS [9-11], and this avoidance coping might negatively reinforce further maladaptive alcohol use [10,12-14]. Thus, it is important to identify ways to address these interrelated conditions among sexual assault survivors.

Early intervention (ie, intervention in the months after assault to prevent the development of chronic conditions) could be a key strategy for addressing PTS and alcohol misuse after sexual assault [15,16]. After sexual assault, it is considered common and nonpathological for survivors to display elevated trauma-related symptoms [17]. These symptoms resolve naturally for many but not all survivors [18]. Early interventions are meant to be used at the initial emergence of symptoms—before the symptoms become entrenched and potentially less malleable—to reduce risk for longer-term problems. Early interventions have reduced PTS after sexual assault and other types of traumas and have been more effective among individuals with higher baseline symptoms [16,19]. Effective interventions have typically used cognitive behavioral interventions to target risk factors for PTS, such as avoidance coping and cognitive distortions. However, few sexual assault survivors seek mental health treatment related to their assault [20,21] for reasons that include privacy concerns, shame, and the lack of health insurance [22,23].

Mobile health (mHealth) interventions (ie, interventions delivered via a web platform or smartphone app) [24,25] present an opportunity to increase prompt access to early interventions [26]. Smartphone ownership is pervasive, especially among young adults [27]. Indeed, in a national study, approximately one-third of American adults had an mHealth app installed on their device [28]. Although technology-based early interventions have reduced PTS after other forms of trauma [29], none have reduced posttrauma alcohol misuse, and there is currently no evidence-based mHealth early intervention to address both PTS and alcohol misuse after sexual assault.

We conducted a pilot randomized clinical trial of THRIVE, a coached mHealth intervention intended to address PTS and alcohol misuse among female participants who have experienced sexual assault within the prior 10 weeks. This paper reports results related to the goal of assessing intervention feasibility and acceptability in the intervention condition. Regarding feasibility, we hypothesized that (H1) completion rates for daily activities would be similar to those of other web-based interventions (ie, 10 days of a 14-day alcohol use and emotion regulation intervention completed by sexual assault survivors in a study by Stappenbeck et al [30]) and that (H2) participants would show significant learning as evidenced by increases in correct responses to knowledge questions from baseline to after the intervention. Regarding acceptability, we hypothesized that (H3) participants would report above-average usability on a standardized measure and that (H4) most of the participants would respond positively on items assessing satisfaction with the intervention.

THRIVE is a coached multicomponent mHealth intervention for survivors of recent sexual assault. The app was created by the study team with input from survivors and survivor-serving professionals. The app includes daily exercises with a duration of approximately 5 minutes and 10- to 20-minute weekly coaching calls over a 21-day period, starting within 10 weeks of sexual assault. Total expected user time burden with 100% activity completion and coaching call attendance is 5.4 hours. Figure 1 displays the dashboard of the THRIVE app.

The THRIVE app includes 2 cognitive behavioral modules—cognitive restructuring (Thoughts) and activity scheduling (Self-Care)—intended to be completed daily and a relationships module (Relationships) intended to be completed on an as-needed basis.

The trial of THRIVE was preregistered on ClinicalTrials.gov (NCT03703258). Detailed information about the study procedures is available in the primary outcomes paper [42].

Recruitment started on January 13, 2021, and ended on August 24, 2021. Participants were recruited via mass emails sent to the enrolled student body of the campuses of the University of Washington, fliers posted in community locations, social media advertisements, and referrals from survivor-serving agencies.

Study staff conducted brief phone screenings with 128 individuals to preliminarily assess eligibility criteria. Ineligible participants were offered referrals by the screener. Preliminarily eligible participants (50/128, 39.1%) were guided in installing the study app (but were not able to log in), scheduled for a phone call with a study coach, and sent a consent form link. Participants who provided consent (49/50, 98%) were redirected to the self-report baseline survey to confirm eligibility. Participants who were found to be ineligible (8/49, 16%) were notified via an automated message, paid for survey completion, and offered referrals. Eligible participants (41/49, 84%) were randomized and sent a link to log in to the app. The app automatically displayed the version of the app to which they were randomized: either the full THRIVE app (containing intervention modules as well as self-monitoring and supportive elements) or the control app (containing self-monitoring and supportive elements).

Participants randomized to the intervention condition (20/49, 41%) are the focus of this analysis because they received the hypothesized active elements of the app (ie, the intervention modules). Participants were encouraged to use the app daily over a period of 21 days and attend weekly 10- to 20-minute coaching calls with the first author, a PhD-level licensed clinical psychologist. The 21-day period of use was initiated by logging in to the app for the first time; after 21 days had passed, participants could continue to access all app features other than the self-monitoring elements for 12 weeks but were no longer prompted to access the app.

On day 21 after the baseline assessment, participants were sent the postintervention assessment, which contained the knowledge assessment. Because of a survey programming logic error, participants did not receive the acceptability or usability scales in the postintervention assessment as intended. Participants were recontacted in October 2021 to complete these scales. Of the 20 participants, 16 (80%) completed the scales at this point.

The procedures received approval from the University of Washington institutional review board (STUDY00005025). Participants were provided with consent information verbally before screening and in a written information statement before study enrollment. The information provided contained all elements required for informed consent, including a description of the purpose of the research; the voluntary nature of participation, participants’ rights, and the risks of participation; the availability of referral options; data retention and storage information; protections for, and limits, to confidentiality; procedures for reporting complaints and adverse events; and an option to be contacted by the investigators to discuss questions before consenting if needed. Data were deidentified. Participants were paid US $20, US $40, and US $15 in digital gift cards for completing the baseline assessment, postintervention evaluation, and acceptability or usability scales, respectively.

The inclusion criteria were as follows: (1) informed consent, (2) self-identification as female, (3) experience of unwanted nonconsensual sexual contact within the past 10 weeks, (4) age ≥18 years, (5) fluency in English, (6) daily smartphone and internet access for 3 weeks and weekly access for 3 months, (7) ≥1 past-month alcoholic drink, (8) ≥1 past–6-month high-risk drinking episode (defined as either >3 drinks on 1 day or >7 drinks in 1 week), and (9) ≥3 symptom clusters endorsed on the PTS Disorder Checklist [43]. The exclusion criteria were active suicidality or psychosis.

The first author conducted all weekly coaching calls with participants and took detailed notes during the calls for clinical documentation purposes and to track feedback. As part of these calls, the first author asked participants about their experience of using the app that week. In the final call, the first author asked for more extensive feedback about participants’ overall experience with THRIVE, including aspects they liked and disliked as well as recommendations for changes to THRIVE. These call notes are used as qualitative data in this analysis.

We used qualitative data from the coaching calls to contextualize and elaborate on the quantitative results. To analyze these data, we used a grounded theory approach [49]. The first author organized the coaching call feedback into collections reflecting the component of the intervention being described (ie, general app, symptom self-monitoring, activity scheduling setup exercise, daily activity scheduling exercise, cognitive restructuring setup exercise, daily cognitive restructuring exercise, coaching calls, and relationships activities). Next, the first and second authors examined the feedback independently and engaged in an open coding process in which they noted potential themes within each section of the intervention. The first and second authors then met to identify and refine first-order themes (ie, themes reflecting specific pieces of feedback about each section of the intervention) and develop a codebook reflecting these themes. The first and second authors then grouped these themes into second-order themes, which represented patterns in first-order themes across intervention components (eg, factors increasing usability). Subsequently, the first and second authors independently coded the feedback using the codebook and then met to resolve discrepancies and update the codebook as needed.

The qualitative analysis of the coaching call notes is summarized in Multimedia Appendix 1. These results elaborate on quantitative findings regarding completion rates, usability, and satisfaction.

The goal of this study was to evaluate the feasibility and acceptability of THRIVE, an mHealth early intervention designed to reduce prospective risk for PTS and alcohol misuse after sexual assault. The results suggest that THRIVE is a feasible intervention that is generally seen positively by participants. These promising results suggest that further development and testing of THRIVE are warranted.

The study results provide evidence for the feasibility of engaging participants in THRIVE. All intervention condition participants completed at least some cognitive behavioral in-app exercises and attended at least 1 coaching call, with the majority (16/20, 80%) attending all 4 calls. The absolute number of daily activities completed (ie, 10) was similar to that observed in a prior web-based intervention with a similar (but nonacute) sample [30], in support of H1. However, the proportion of days completed was lower because the duration of THRIVE was 21 days compared with the duration of 14 days in the study by Stappenbeck et al [30]. This might indicate that a 14-day intervention period is more realistic than the 21-day period used in THRIVE; future research could explore whether a shorter duration of THRIVE would be effective. It is also possible that participants were less engaged in THRIVE because they were in the acute postassault period, unlike those in the study by Stappenbeck et al [30]. Indeed, in the qualitative results, several of the participants stated that they had outside responsibilities or stresses that took precedence. It is possible that the longer duration of THRIVE allowed participants in the acute phase to receive a similar dose of intervention even with lower frequency of engagement.

In addition, we found that the relationship-focused module was used infrequently. This module was intended to be used on an as-needed rather than daily basis; therefore, participants did not receive regular prompting to access it. The qualitative data revealed that participants did not notice this section owing to its visual design, but those who accessed it found it highly useful. This indicates that the visual design of the app could be improved to make this helpful section more noticeable; periodic prompts could also be added to draw participants’ attention to this module.

We found that THRIVE was successful in conveying its key concepts, providing further evidence of intervention feasibility. Specifically, in support of H2, participants demonstrated increases in comprehension from baseline to the postintervention assessment of the psychoeducational content introduced in the intervention. This indicates that participants were paying attention to the psychoeducational content and that it was presented clearly enough to be retained over the 3-week period. This demonstrates that even a light-touch intervention such as THRIVE can successfully convey important therapeutic concepts.

Participants generally reported high usability. In the quantitative data, participants gave the THRIVE app a usability score corresponding to a B+ usability grade (in support of H3), highlighting strong usability but room for improvement. This was consistent with the qualitative results: the coaching call notes documented that participants found the exercises to be generally clear and presented in a helpful way but also noted some instances in which specific activities were unclear or difficult. Consistent with prior models of the utility of coaching in mHealth interventions [38], the call notes documented that participants found the coaching calls helpful for clarifying concepts and troubleshooting activities. This feedback provides clear direction for revisions to THRIVE, while also suggesting that coaches may be able to address usability barriers.

Participants also reported high satisfaction with THRIVE. In the quantitative results, most of the participants rated the app (15/16, 94%) and coaching calls (14/16, 88%) as helpful, in support of H4. In the qualitative results, the coaching call notes documented many aspects of the intervention that participants had stated were appealing, including the intervention content and the visual design. Importantly, a key contributor to satisfaction was a perception that the intervention helped with participants’ recovery. It is notable that several of the participants (3/20, 15%) returned to use THRIVE after the intervention period was over, which provides further evidence that THRIVE was seen as useful to participants. Nevertheless, the satisfaction results also indicate areas for improvement, such as identifying ways to make emotionally difficult exercises easier to tolerate and less repetitive. These changes should be explored in future versions of THRIVE.

This study includes several limitations. First, owing to a programming error, participants did not complete acceptability measures immediately after the intervention and, as a result, may have had difficulty remembering their opinion of the THRIVE app. Although most of the participants (16/20, 80%) completed the satisfaction survey when it was resent to them, we were unable to obtain satisfaction data from 4 (20%) of the 20 participants, and it is possible that these participants had lower satisfaction with the intervention. Second, our use of an exclusively female, primarily college-student sample limits our ability to understand feasibility and acceptability among the broader community population that might benefit from using mHealth apps after sexual assault, including men and minoritized racial and ethnic groups. Third, this study was conducted during active COVID-19 restrictions (eg, remote learning for college students and restrictions on social gatherings). Although our ability to enroll survivors of recent sexual assault during this period highlights the feasibility of mHealth interventions for sexual assault survivors even during times of global crisis, the restrictions might have affected app use rates. Fourth, although our mixed methods design is a strength, the use of coaching call notes as qualitative data (rather than separate qualitative interviews) has several biases. In particular, although we obtained feedback from at least 1 coaching call from all participants, the coaching call data do not fully reflect the perspectives of the participants who did not complete all coaching calls (4/20, 20%). As notes were taken by the coach during calls, and the calls were not audio recorded, it is possible that feedback could have been missed or erroneously documented. In addition, because coaching calls were completed by the first author (the trial principal investigator and the creator of the intervention), participants might have felt pressure to report more positive opinions of the intervention. We reduced this likelihood by emphasizing that a study goal was to find ways to improve the intervention; nevertheless, future research should use qualitative interviews conducted by outside evaluators to understand participant opinions.

These results suggest that THRIVE may be a feasible and acceptable way to deliver an early intervention among sexual assault survivors. If efficacious, THRIVE would be a promising low-barrier strategy to increase the reach of early interventions among this population with high needs.