The Iona Mind app (Iona) for the management of mild to moderate low mood [28] or excessive worry [21] informed the intervention adopted for adaptation (see Figure 1 for user flow through Iona).

After onboarding, the user is asked to complete the Patient Health Questionnaire-8 [29] and Generalized Anxiety Disorder-7 Scale [30] patient-reported outcome measures to assess the severity of low mood and worry symptoms, respectively. They then engage with an interactive CBT Five Areas model [31] before progressing through the intervention-specific factor techniques presented as 6 educational modules. Where the user scores mild to moderate symptom severity for both low mood and worry symptoms, they are directed to the technique corresponding with the highest symptom severity. Where symptom severity is identical, however, users are recommended the technique for low mood, although choice still can be exercised. An “SOS” button is available on every screen for users who either score within the severe range during onboarding or show signs of deterioration during use. If the person is seeking immediate help, the “SOS” page contains NHS Emergency contact details.

Where treatment for low mood is recommended, the user is introduced to the behavioral activation [32] or cognitive restructuring [33] techniques derived from low-intensity CBT [34]. Alternatively, where symptoms associated with worry are prominent, the Worry Management technique is proposed [21]. To facilitate choice, a description of each technique is available on request. Based on the evidence base, all techniques are informed by protocols recommended by the National Institute for Health and Care Excellence, Behavioral Activation [35,36], alongside Cognitive Restructuring for low mood [37] and Worry Management for excessive worry [38]. All techniques are included within the most commonly adopted written self-help interventions [39], used by NHS Talking Therapies for anxiety and depression services [40]. Engagement with the app is maintained through an omnipresent chatbot to establish a “therapeutic approach” [41]. This promotes collaborative empiricism to encourage the user to explore and reflect upon outcomes arising from engagement for themselves [42].

The EVF [26] was adopted to systematically inform adaptation from the original Iona app to that specifically for Iona Ex-Servicewomen (IonaXSW). The app was later currently called “Iona Female Forces Veterans” (IonaFFV) based on results from phase 2. To ensure that cultural and contextual factors associated with the female forces veteran’s community are represented, the adaptation process focused on dimensions of the EVF [43] related to language, metaphor, context, and content. While dimensions related to concepts and methods were also adapted with respect to language, the theoretical rationale and technique-specific factors maintained high fidelity to the underlying CBT model. These dimensions were used to explicitly inform the focus group topic guide to ensure that structure was provided to ensure adequate coverage.

The adaptation process comprised 5 phases to adapt Iona and examine its acceptability for female forces veterans (Figure 2). Across the 5 phases, 4 panels were established to provide different perspectives and contributed in varied ways with each panel building upon the work of the others [44].

During the adaptation process, an Armed Forces Panel was established to guide the initial adaptations, focusing on the context and content dimensions of the EVF. Phase 1 aimed to tailor the Iona app for female forces veterans and support the development of IonaXSW. After initial adaptations, a female forces veterans panel used their lived experience to refine the app further in phase 2. Data were analyzed to create an adaptation framework, which confirmed acceptable phase 1 adaptations and identified areas requiring additional modification. The framework was then provided to the clinical or technical panel in phase 3, who made further changes to develop the IonaFFV app. In phase 4, the app was returned to the same female forces veterans panel from phase 2 to assess the acceptability of the adaptations and ensure that changes based on the adaptation framework were effectively implemented. Finally, in phase 5, the IonaFFV app was released online to a wider group of female forces veterans to explore acceptability.

The clinical or technical panel comprised a male app developer (JB) for Iona Mind, and a male expert (PF) in low-intensity CBT apps who holds national expert advisory positions in low-intensity CBT and digital technologies. They used the adaptation framework developed in phase 2 to develop IonaFFV. The panel ensured that all adaptations maintained fidelity with the underlying low-intensity CBT clinical protocols, and software remained conformant to appropriate software requirements.

Ethical approval was obtained for the focus groups (reference: 2611056) and online acceptability (reference: 5930558) studies from the University of Exeter psychology department. To ensure informed consent, all participants were provided with participant information sheets, which included information on the aims of the study, right to withdraw, data storage, and management processes. Participants were informed that their participation was entirely voluntary, and they could withdraw from the study without any negative consequences. No identifying information about participants is included. All data were reviewed to ensure that individuals cannot be identified, and no materials containing identifiable features were used. As all data were anonymized prior to analysis, participant consent for publication of identifying information was not required. Phases 2 and 4 focus group participants were offered compensation for travel costs, and phase 5 questionnaire participants were offered the opportunity to enter a prize draw for a US $133 electronic voucher.

Overall, feedback provided by the female veterans focus group combined with responses added to the free-text response box included within the MAUQ [53] questionnaire administered in phase 5 highlighted that IonaFFV was well received. Awareness that the app had been adapted for female forces veterans was positively responded to.

Adaptations appeared to promote continued engagement with IonaFFV as participants worked through each module.

However, there was 1 disconfirming case reported on the free response text box.

Analysis is presented in 2 parts. First, reflecting the attitudes expressed toward IonaXSW by the focus group during phase 2, alongside preferences regarding support. Second, attitudes expressed in phase 4 regarding the extent to which adaptations have been successfully implemented during phase 3 to create IonaFFV, alongside identifying outstanding areas requiring further adaptation. Categories identified as emerging were associated with “Context,” “App features,” and “No adaptations needed” (Table 2).

This category captured the attitude of participants toward features of the Iona app that have been adapted to represent female forces veterans. These were grouped into label, graphics, quotes, and avatar.

App features refer to attitudes held regarding general features included within IonaXSW to support help seeking, aid navigation and usability, and motivate continued engagement.

Three main screens and interactive functionality were positively responded to by all focus group members in phase 2 and did not need further adaptation. This potentially indicates that initial adaptations in these areas made to accommodate female forces veterans were compatible with the content, methods, and context dimensions of the EVF.

Median (IQR) values across each of the MAUQ subscales were as follows: ease of use 1.80 (1.20‐2.80), interface and satisfaction 2.50 (1.54‐3.33), and usefulness 2.75 (1.75‐4.00). Note that the lower values indicate higher agreement. Each subscale was further examined to determine percentage agreement with each statement (Table 3). Responses to some items were missing, resulting in varied sample sizes across items (n=19‐20). Overall, most participants agreed with each statement, indicating that the app was highly acceptable, except for “I could use the app even when the Internet connection was poor or not available,” rated by 63% of participants as neutral or disagreed.

Using a 7-point Likert scale (1=Strong Agree to 7=Strongly Disagree) a descriptive analysis was undertaken on questions related to 5 main areas associated with IonaFFV. Participants considered the language engaging (mean 1.85, SD 0.87), SOS button was required in the event of a mental health crisis (mean 1.95, SD 1.47), and the signposting information included within IonaFFV was helpful (mean 2.50, SD 1.38). However, it was considered that signposting details to regional organizations should be provided (mean 2.10, SD 1.15). Participants also indicated a need for human support to engage with the app (mean 1.90, SD 1.18). In follow-up questions about human support, where participants could select more than 1 option, there was a preference for weekly scheduled support provided by a professional or voluntary or community sector staff over the telephone (Table 4).

Interestingly, while there was least preference for weekly support to be provided by peers, using peers was identified as the most preferred in the event support was provided on demand. Where support was provided, it was considered that it should be focused on motivating continued engagement and supporting engagement with techniques used within the app as required.

Overall, adopting CBT as the basis of an app to support low mood and worry management in female forces veterans was acceptable. Participants found that the CBT model adapted to reflect female forces veterans both aided understanding of their emotional difficulties and helped tailor the app to them as a distinct group. Furthermore, there was good engagement with techniques used to support low mood and worry management, with initial engagement further enhanced by highlighting use of IonaFFV by other users at the start of engagement and the contextual factors. However, participants differ from most previous research studies with respect to the following characteristics: from the Navy, underwent unplanned discharge, and deployed. Although the average age is consistent with previous research examining access to the Op COURAGE high-intensity service [10].

Applying the EVF [43] to inform the development of an artificial intelligence–supported CBT app for the management of low mood and worry was, however, important to enhance acceptability and usefulness among female forces veterans. Analysis of phases 2 and 4 highlighted the need to make adaptations to ensure that IonaFFV specifically represented their transition from serving into civilian life. Focusing on the EVF context dimension highlighted the need to redirect adaptations from serving to female forces veterans to adequately capture this transition while maximizing acceptability. This was exemplified by reactions to the term “ex-servicewomen,” which was not considered to represent a new civilian identity while acknowledging their past service in the armed forces. This highlights the importance of the language dimension of the EVF when considering adaptations for female forces veterans.

Of interest, “ex-servicewomen” was the term originally proposed during phase 1 of the development process comprising senior members of the armed forces community and service providers. Furthermore, while phase 1 was used to identify the graphics and voices for use by the avatars, these were not considered to represent female forces veterans as a group and therefore removed during development. Adaptations were also required to enhance usability with respect to the structure of the SOS page, enhance navigation through the app where there was a desire to change CBT technique, and enhance the way in which progress was captured.

Consistent with previous research, the findings of this study reinforce the significant role that serving in the armed forces plays in shaping the identities of both male and female veterans [56]. The results emphasize the particular challenges faced by female forces veterans in negotiating their identities during the transition to civilian life [6]. This highlights the need for tailored approaches that acknowledge their unique experiences in the armed forces. While female forces veterans are often “hypervisible” during their military careers, they may simultaneously suppress behaviors traditionally considered feminine [7]. Failure to recognize these adaptive strategies supports the need identified in this study for appropriate modifications to IonaFFV. Implementing adaptations was recognized to improve acceptability by better reflecting the experiences of female forces veterans as they transition from life in the armed forces to becoming a civilian [6].

While the initial adaptations to IonaXSW used graphics aligned with the armed forces to enhance engagement, participants emphasized the importance of gradually reducing the focus on the armed forces as they progressed through the app to better reflect civilian life [7]. They also recommended replacing quotes associated with the armed forces context with those drawn from civilian life. A lack of recognition of the transition from military to civilian identity could have hindered the process of “rediscovery” that supports successful reintegration [7]. Furthermore, failing to change the term “ex-servicewomen” [1,7], proposed in adaptation phase 1 to female forces veterans, would have further reflected a transition oversight. However, the preference to maintain “veteran” is not consistent with previous research [57], where it is traditionally considered to reflect previous generations of male veterans [57]. Potentially, these differences have arisen from study demand characteristics, given that recruitment materials adopted the term “veteran” [58].

Features of IonaXSW that did not require adaptation are consistent with those commonly identified as acceptable with CBT [59]. For example, the CBT model both enhanced understanding regarding the nature of the mental health difficulty faced and aided an understanding of the way the low-intensity CBT technique may help [60]. Furthermore, consistent with a metasynthesis of computerized CBT, personalization of CBT-specific factor content enhanced relevance [61], promoting connection and engagement. The screen simulating the location of other female forces veterans also served to normalize mental health support [62] and encouraged lateral social comparisons [63].

While the chatbot successfully maintained user engagement with IonaXSW, consistent with findings from CBT-based self-help interventions more broadly, participants continued to express a preference for human support [64]. Among the various forms of support considered acceptable, the most preferred option was scheduled weekly sessions with a professional from a veteran-specific NHS service. This finding suggests potential benefits of integrating support for IonaFFV through services such as Op COURAGE [10]. However, despite the overall acceptability of this service [10], its evaluation included only male veterans, which limits the generalizability of the findings to female armed forces veterans.

Alternatively, nearly half of the participants preferred on-demand peer support, indicating potential advantages similar to those observed with peer-support groups [65]. Implementing IonaFFV alongside support from female forces veterans within veteran-specific services [9] could serve to enhance help-seeking and address existing barriers to engagement [66]. Moreover, as low-intensity CBT techniques can improve access to care, adoption could enable services to manage increased demand without extending waiting times [67]. While IonaFFV incorporates in-app support to promote engagement [68], incorporating peer support could mitigate low levels of engagement with a stand-alone app following download [69].

Adopting an approach toward adaptation comprising focus groups to guide initial modifications and using the MAUQ [53] to assess acceptability offers a means to evaluate how well these adaptations are acceptable to a broader group of female forces veterans. This methodology could help address challenges related to the limited generalizability of focus group data [70]. However, the small number of participants completing the MAUQ limited further analysis regarding the acceptability of IonaFFV across different demographic groups. Specifically, only a single participant had served in the Royal Air Force, identified as non-White, or had experience as a reservist. Recruitment challenges for phase 5 may have been influenced by the requirement for participants to engage with IonaFFV for 4 weeks before completing the questionnaire. Therefore, future research should consider strategies to enhance recruitment and questionnaire completion before assessing the acceptability and usability of IonaFFV and generalizing study results [71].

It may also have been of benefit to collect additional information in phase 5 regarding the service experiences of female forces veterans. Female forces veterans experience higher levels of in-service adversity than their male colleagues, including exposure to military sexual trauma [72]. Although IonaFFV was explicitly designed for female forces veterans experiencing low mood or difficulties with worry management, some questionnaire responses may have been influenced by past service-related adversity. This possibility was suggested by a free-text questionnaire response which indicated that the participant may have been experiencing moderate to severe mental health difficulties. Gathering more comprehensive data on participants’ service experiences could have supported a deeper interpretation of the findings and informed further adaptations, such as incorporating broader signposting information.

There is a clear need to recognize female forces veterans as a distinct group from their male peers. While sharing many values and traditions common across the armed forces, they also hold a collective identity that differs from that of male veterans [73]. Addressing the current inequities in access to, and appropriateness of, mental health services for female veterans [8,14] requires a focus on enhancing the acceptability of veteran-specific services. In particular, using dimensions of the EVF to structure the focus group topic guide was valuable to identify the types of adaptations necessary to maximize acceptability. The EVF also helped highlight how these dimensions shift during the transition from serving to armed forces veteran, offering insight into how mental health services can be better tailored for female forces veterans. When applied to the development of a CBT-based mobile phone app for managing low mood and worry, the EVF informed adaptations that produced a digital mental health approach with strong usability, acceptability, and perceived usefulness among female forces veterans.