This study represents Phase 1 of a multiphase project that aims to co-design and evaluate a social media–based universal prevention intervention for perinatal mental health in regional, rural, and remote communities, using Northern Queensland, Australia, as an exemplar. The project uses a participatory co-design approach informed by the Double Diamond model [34] through divergent exploration and convergent refinement to develop and evaluate a preventive perinatal mental health intervention program. Specifically, Phase 1 of the project aimed to discover the problems, insights, user needs, and ideas generated by stakeholders and consider how social media may afford or constrain the mental health needs of North Queensland perinatal women, resulting in a set of design principles for the prototype intervention. In Phase 2, these principles will be applied to develop a prototype intervention integrating evidence-based psychological techniques for preventing perinatal depression and anxiety, which will be iteratively refined through additional focus groups and interviews with stakeholders. In Phase 3, the prototype intervention will be evaluated in a pilot trial to assess its feasibility and acceptability with the target population ahead of a future randomized controlled trial.

The research team comprised mental health care researchers and clinicians (a clinical psychologist, a perinatal psychiatrist, a psychologist, a mental health nurse, midwives, and a social worker) as well as a human-computer interaction researcher. A community advisory group of representatives with lived experience of perinatal depression and/or anxiety diagnoses, perinatal mental health clinicians, and local community leaders provided ongoing consultation throughout the design process. The study also adhered to the GRIPP2 (Guidance for Reporting Involvement of Patients and the Public) checklist (Checklist 1) [35].

Participants were perinatal women and mental health care professionals residing in Northern Queensland, Australia. Northern Queensland is a regional, rural, and remote area within Queensland, Australia, with a population of approximately 732,000 people spanning over 810,000 km² from Sarina on the central Queensland coast up through Mackay, Townsville, and Cairns to the Torres Strait islands and westwards into inland and remote areas of the Gulf of Carpentaria and Mount Isa [6,36].

Perinatal women were eligible if they (1) were aged ≥18 years, (2) were pregnant or ≤12 months postpartum, (3) resided in Northern Queensland, (4) were proficient in English, (5) did not currently meet criteria for a mental health diagnosis, (6) had prior experience with social media, and (7) did not report moderate/severe symptoms or suicidal ideation on standardized screening measures. Mental health status was assessed using the Edinburgh Postnatal Depression Scale (EPDS), with scores >11 indicating moderate-to-severe distress and endorsement of the suicidality item prompting exclusion and referral to supports [37]. For Aboriginal and Torres Strait Islander mothers, Part 1 of the Kimberley Mum Mood Scale (KMMS) was offered as a culturally appropriate adaptation of the EPDS, with the same scoring and exclusion threshold [38]. Both the EPDS and KMMS are recommended as universal perinatal mental health screening instruments in Queensland Health clinical guidelines, with evidence suggesting approximately two-thirds of women are screened using the EPDS in pregnancy statewide [39,40]. Mothers were also asked whether they had a previous diagnosis of a mental health condition in the demographics survey, although this did not form part of the eligibility criteria. No additional prior mental health data were available for participants.

Mental health care professionals were eligible if (1) they were registered as a mental health care professional, (2) they had experience supporting perinatal mental health, and (3) they lived or worked in Northern Queensland. All participants received an AUD $50 (US $31.48) Prezzee voucher to reimburse them for their time.

Sample size was not predetermined but was guided by the principle of information power, which holds that the adequacy of a qualitative sample is determined by its capacity to illuminate the specific phenomenon under study rather than by a fixed numerical threshold [41]. Information power is shaped by the study aims, the specificity of the participant group, the quality of dialogue, and the analytical approach used. Data collection continued until the research team determined that sufficient information power had been achieved, taking into account the clearly defined study aims, the high specificity of the participant groups, and the depth and richness of data generated across sessions. The final sample of 26 perinatal women and 8 mental health care professionals was considered to reflect this determination.

Participants were recruited using convenience and community-based sampling methods. Recruitment advertisements were distributed via social media, flyers, community organization newsletters, and media releases inviting women and professionals to “help design a digital program for Northern Queensland women’s perinatal mental health.” Social media marketing was targeted using location and user preference information to reach the intended audience and was used to reach existing users. Snowball recruitment was also encouraged through participants’ social networks.

Interested individuals were directed to an online survey hosted in Qualtrics (Qualtrics International, Inc). After reviewing the participant information sheet, individuals provided informed consent, completed eligibility screening, and answered questions on their demographic information. Participants also provided a preferred nickname for use in group discussions and selected their availability for online or in-person sessions.

Focus groups and individual interviews were conducted from May 6 to July 4, 2025. Eight focus groups (2-4 participants each) and 15 individual interviews were held, with perinatal women and mental health care professionals participating in separate sessions. Sessions lasted approximately 60 minutes on average. Sessions were offered online via Microsoft Teams or in person in Mackay, Townsville, Cairns, and Mt Isa to maximize accessibility.

A semistructured interview guide was developed, consisting of 7 open-ended questions exploring experiences related to perinatal mental health, social support, and digital technologies (Multimedia Appendix 1). Each session began with an acknowledgment of country, an overview of study aims, and a reminder of participant rights, confidentiality, and voluntary participation.

In-person focus groups were cofacilitated by KJL and either ST or JM. Online sessions were facilitated by the lead author (KJL). Sessions lasted 60‐90 minutes, were audio recorded, and were supplemented by facilitator field notes to capture contextual information.

Ethics approval was obtained from the James Cook University Human Research Ethics Committee (HREC 24H-9686), and the study was conducted in accordance with the National Statement on Ethical Conduct in Human Research. All participants provided informed electronic consent before completing the demographic survey and attending any focus group or interview session. Consent materials clearly outlined the voluntary nature of participation, the right to withdraw at any time without consequence, and how data would be stored and used. To protect participant privacy and confidentiality, all data were deidentified before analysis. Participants were invited, though not required, to use pseudonyms during focus group sessions, and any identifying information was removed from study records. Sessions were audio recorded solely for transcription, and recordings were destroyed once transcription was complete. No images of individual participants are included in the paper or additional materials. Participants were compensated for their time with an AUD $50 (US $31.48) Prezzee digital gift voucher upon completion of their focus group or interview session.

A total of 26 perinatal women participated in the study (Table 1). Of these, 5 were pregnant and 21 were postpartum. Fifteen participated in focus groups and 11 participated in interviews. Participants ranged in age from 25 to 48 years (mean 32.2, SD 4.6 years). Most were multiparous (n=17, 65.4%), partnered (n=24, 92.3%), and living in regional (n=9, 34.6%), rural (n=8, 30.8%), or remote (n=9, 34.6%) areas classified by Modified Monash Model (MM) categories (MM2-MM7). Nearly half (n=12, 46.2%) reported a previous mental health diagnosis, with mean EPDS and KMMS scores indicating low levels of current distress. Most participants used social media daily, primarily Facebook and Instagram, with typical usage between 1 and 3 hours per day.

Eight mental health care professionals also participated (Table 2). Four engaged in focus groups and 4 in interviews. Participants represented a range of professional backgrounds and had a mean of 16.6 years of clinical experience. They worked across public, private, and community settings in MM2-MM7 regions, and 5 (62.5%) had received specialist perinatal mental health training. Most used social media for less than 1 hour per day, primarily Facebook and Instagram, and reported moderate familiarity with using social media as a psychoeducational tool.

This study sought to co-design a universal social media–based preventive intervention for perinatal mental illness among women living in regional, rural, and remote Northern Queensland. We identified 5 core mental health needs: social connection and support, personalized and respectful health care, information that empowers, place-based support, and accessible digital formats. Social media shows promise for addressing these needs through peer support, psychoeducation, and culturally responsive content, yet constraints such as misinformation, limited digital literacy, and inequitable access must be addressed in program design. Our findings demonstrate how digital platforms can be leveraged to deliver preventive mental health support in regional, rural, and remote communities, with implications for clinical practice, service design, and public health policy.

The first 3 themes mirror the broader perinatal experiences across diverse contexts. Previous literature indicates that mothers value connection and peer support during this period, which is often marked by loneliness [48]. The need for personal and respectful interactions that are clear, evidence-based, and empowering within health care is frequently cited as crucial to perinatal mental health, yet lacking within available supports [48]. Our study extends these findings by demonstrating that these needs are magnified in regional, rural, and remote communities, where social and systemic inequities amplify their salience. While these communities are resilient, this does not deem them exempt from requiring mental health support [9]. Acknowledging and addressing these needs is essential to reducing the compounding effects of inequity and mitigating intergenerational impacts of poor perinatal mental health [4].

Alongside these broader needs, our study highlights the unique systemic and social barriers faced by women in regional, rural, and remote regions. Unlike metropolitan centers, these regions experience workforce shortages that create longer wait times and a lack of individualized care, exacerbated by geographic isolation that has considerable financial and travel-related burdens. The unique social dynamics of smaller communities, characterized by shame, stigma, and a lack of anonymity, created barriers to seeking support beyond that of service availability. Our findings highlight the need for place-based approaches to appropriately address systemic and social inequities and cultural and contextual differences within these communities [11].

Further, our findings suggest that a well-designed social media intervention may help to mitigate some of these challenges. The mothers were supportive of a social media intervention due to its capacity to provide accessible information and act as a “digital village” to foster peer support. This is consistent with previous literature demonstrating that social media is an important source of health information for perinatal women [20]. However, social media cannot address all barriers, including limited digital literacy and structural inequities that remain prominent in these regions. Previous literature demonstrates how social media can perpetuate unhealthy comparison, identity distress, and parenting anxiety, negatively influencing mental health [23]. These concerns were shared by both mothers and mental health care professionals in this study and must be addressed in both the design and functionality of digital platforms. The needs-affordances analysis allowed us to translate these insights into actionable, relevant, and feasible design guidelines. It emphasized the need for a place-based approach that fosters social support while mitigating harmful comparison through moderation. While previous digital health interventions have reduced symptoms of anxiety and depression in perinatal women, the use of social media as a preventive mental health tool remains novel [49]. This research demonstrates that a social media health intervention is a feasible and acceptable platform for delivering preventive, accessible psychoeducational mental health support to perinatal women at a population level.

The needs and design guidelines identified in this study align with prior interventions but add nuance that addresses key gaps. Most previous interventions, including those that were effective in reducing depression and anxiety, incorporated peer connection via chat groups, peer worker support, or forums, highlighting the value of social support [27,30]. However, these interventions were predominantly psychoeducation-based, delivered uniform content, and rarely allowed users to self-direct learning and engagement. Only Hatamleh et al [32] explicitly considered diverse cultural needs, leaving most programs limited in their capacity to engage culturally diverse or regionally isolated mothers [32]. Importantly, none provided options for users to filter or select topics of interest. This was a feature that the mothers in this study identified as key to fostering meaningful engagement and supporting mental health outcomes. Our guidelines directly address these limitations by emphasizing a user-led, autonomous approach grounded in place-based content and design. This approach not only increases the likelihood of engagement and relevance to mothers in regional, rural, and remote communities, but also mitigates the risk of inadvertently excluding or disadvantaging mothers from diverse cultural and linguistic backgrounds. In doing so, these findings extend prior work by illustrating how social media interventions might be tailored to a regional, rural, and remote context to maximize both efficacy and equity, providing actionable guidance for future mental health programs.

The social media intervention could integrate into existing health care systems in a way that alleviates, rather than duplicates, existing burdens. Mental health care services remain predominantly reactive and treatment-oriented, constrained by workforce shortages, increasing health demands, and higher health care costs [50]. In the context of the continuum of care, this program could provide preventive, universal support for perinatal mothers [51]. The potential for a social media intervention to complement existing services was highlighted by the mental health care professionals in this study, which would be particularly beneficial for both their practice and client outcomes. For those experiencing mild distress, peer support and psychoeducation may be sufficient to prevent symptom escalation. The program can also provide a mechanism for mental health screening and referral to formalized supports, complementing existing services and potentially decreasing the need for more intensive services downstream.

The findings of this study challenge the assumption that digital mental health interventions developed in metropolitan or international contexts can be straightforwardly applied to regional, rural, and remote settings. The mental health needs identified in this study were experienced with a distinctly place-based character, shaped by geographic isolation, workforce shortages, small community social dynamics, and systemic inequity, which generic interventions are unlikely to adequately address. Place-based co-design is, therefore, not merely a methodological preference but a practical necessity for intervention relevance and equity in these communities.

For clinicians and health care services, a well-designed social media intervention holds potential to complement existing perinatal mental health care by providing psychoeducation, peer connection, and normalization of help-seeking at a population level, alongside a mechanism for referral to formalized supports where needed. For policymakers and funders, these findings make a case for sustained investment in co-designed digital mental health infrastructure, including the structural enablers of equitable access such as reliable connectivity, device affordability, and digital literacy support.

More broadly, this study demonstrates that social media is a viable and acceptable intervention platform for adult perinatal women—a population largely absent from the existing evidence base for social media interventions [52]. Our findings highlight the potential for social media to function as an equalizer of access to mental health support, offering an engaging, low-cost, geographically unconstrained, and stigma-reducing touchpoint in settings where structural barriers would otherwise prevent timely care. However, the findings also emphasize that social media is not a neutral delivery channel for mental health support. Realizing this potential requires affordance analysis to be adopted as a standard step in social media intervention development, ensuring platform selection and design are driven by user need.

The next phase of this research is a pilot study involving the development and evaluation of the universal preventive intervention informed by the prototype design guidelines generated here. This pilot phase will use a feasibility and acceptability design, examining whether the intervention can be implemented as intended in regional, rural, and remote communities, and generating preliminary data on outcomes including mental health and well-being, social connectedness, and help-seeking behavior, to inform the design of a subsequent fully powered effectiveness trial. Future studies should also explore the potential for predictive and personalized digital tools, including adaptive algorithms and artificial intelligence, to anticipate psychological distress and triage mothers into appropriate care [53,54].

The co-design process generated rich, contextually grounded data and was strengthened by the active involvement of a community advisory group comprising lived experience representatives, clinicians, and local community leaders, who provided ongoing input throughout the research process. The dual online and in-person format was effective in expanding accessibility across a geographically dispersed region, and participants across both formats engaged substantively with the focus group and interview questions, generating detailed and meaningful contributions to the design process.

Notably, the co-design process substantially shaped the direction of the intervention in ways that were not anticipated at the outset. The original intervention concept envisaged a broadcast-style social media presence through which preventive psychotherapeutic content would be disseminated to a largely anonymous audience (eg, a public Instagram or Facebook page). However, the needs-affordances analysis revealed that peer interaction, mutual support, and personalized connection were central priorities for participants, fundamentally reorienting the intervention toward a more interactive and community-driven model. This shift reflects the genuine influence of patient and public involvement on the research outcomes and underscores the value of co-design approaches for intervention development.

However, several challenges were also encountered. Higher-than-anticipated dropout was observed among mothers attending in-person sessions, with cancellations frequently occurring at short notice, consistent with the practical realities of caring for a newborn or infant, including feeding schedules, sleep deprivation, and unpredictable child health. Future co-design research with perinatal populations should anticipate this from the outset, for example, by offering shorter or more flexible session formats or providing childcare support where feasible.

This study’s strengths lie largely in its co-design approach, which provided contextually grounded insights supported by lived and clinical experience. Using face-to-face and online methods across diverse regions enhanced depth and representation within the data. The inclusion of a community advisory group with lived experience of perinatal depression and anxiety, alongside a sample in which nearly half of participants reported a prior mental health diagnosis, further strengthened the relevance and authenticity of the findings.

However, several limitations should be noted. First, the sample underrepresented First Nations and LGBTQIA+ (lesbian, gay, bisexual, transgender, queer (or questioning), intersex, and asexual (or agender), and other diverse sexual orientations, gender identities, and sex characteristics) voices, with only 2 participants identifying as Aboriginal and/or Torres Strait Islander despite recruitment through targeted social media advertising, posters in libraries, health care services, community spaces, and media releases. Notably, a number of Aboriginal and Torres Strait Islander mothers expressed interest in participating but were disproportionately screened out at eligibility due to reporting moderate-to-severe symptoms on the KMMS, reflecting the well-documented higher rates of perinatal distress among First Nations mothers in Australia [8,55], and representing an important intersection between our eligibility criteria and First Nations representation. This also highlights a broader tension inherent to universal program co-design: although the intervention is intended to be broadly accessible across diverse communities, the co-design process itself may not have equally captured the perspectives of all subgroups within the target population. First Nations perspectives were nonetheless present at the level of research governance, with Aboriginal and Torres Strait Islander members contributing to both the chief investigator group and community advisory group throughout the study. Future research should prioritize community-embedded recruitment developed in genuine partnership with Aboriginal Community Controlled Health Organizations, employ First Nations researchers and community liaisons, and consider whether supplementary targeted co-design processes are needed to ensure the intervention is culturally safe and relevant for First Nations and LGBTQIA+ communities specifically [56,57].

Second, the exclusion of women currently experiencing moderate-to-severe mental health symptoms or suicidal ideation limits the study’s capacity to capture insights from those at highest risk. This exclusion was intentional and reflects the preventive rather than clinical focus of the intervention; the co-design sample was deliberately aligned with the intended end-user population of the program, which is explicitly preventive and psychoeducational in nature rather than a clinical or therapeutic tool. Accordingly, the findings and design guidelines generated by this study should not be generalized to clinical programs aiming to treat moderate-to-severe perinatal mental health presentations; such programs serve a distinct population with distinct needs and would require their own dedicated co-design processes to determine what women in those contexts want and need from a social media-based intervention. Nevertheless, we acknowledge that women who have experienced symptom escalation may have offered valuable perspectives on what earlier digital support could have prevented. This limitation is partially mitigated by the fact that 46% of participants reported a prior diagnosis of depression and/or anxiety, and by the inclusion of lived experience representatives of perinatal depression and/or anxiety on the advisory group. Future work could consider engaging women with current or recent clinical presentations as a distinct consultative group to complement the perspectives captured here.

Third, the eligibility criterion requiring prior social media experience was included to ensure participants could engage meaningfully with the affordances analysis. While this criterion could theoretically filter out the most digitally excluded members of the target population, a tension worth acknowledging given that digital literacy and affordability emerged as barriers in our findings (Theme 5), no participants were excluded on this basis in practice. This reflects the near ubiquity of social media use even among populations with limited digital literacy or constrained financial resources. Recruitment through both online and offline channels further reduced the likelihood that digitally excluded women were unable to learn about or express interest in the study. Nonetheless, future evaluation phases should monitor whether the intervention is equally accessible and acceptable to women with lower levels of digital literacy to ensure equitable reach.

Fourth, social media affordances are not static and may shift over time as platform designs, algorithms, community norms, privacy policies, and access conditions evolve. The design principles generated by this study reflect the affordances landscape at the time of data collection, and future intervention development and evaluation phases will need to remain responsive to changes in the social media environment to ensure continued relevance and safety.

Fifth, recruitment via convenience sampling and snowball methods, including social media advertising and participant referral networks, may have preferentially engaged women who are already digitally active and connected to community networks. Women with lower digital literacy, limited internet connectivity, or reduced engagement with community organizations, including those in the most geographically remote parts of Northern Queensland, may therefore be underrepresented. Future co-design and implementation research should incorporate targeted, community-embedded recruitment strategies to reach these subgroups.

This study aimed to identify the mental health needs of perinatal women in regional, rural, and remote communities and to co-design a framework for a preventive social media–based intervention informed by platform-specific affordances and constraints. Using a participatory co-design approach with 26 perinatal women and 8 mental health care professionals in Northern Queensland, Australia, 5 core mental health needs were identified: social connection and support, personalized and respectful health care, information that empowers, place-based and culturally safe support, and accessible low-burden digital formats. A needs-affordances analysis demonstrated that social media holds promise for addressing these needs through peer support, psychoeducation, normalization, and timely information delivery, provided that known constraints are proactively addressed in program design, including misinformation, privacy risks, harmful social comparison, and inequitable digital access. These findings were synthesized into prototype design guidelines to inform subsequent phases of intervention development. Taken together, these findings address a significant gap in the literature and offer a replicable methodological approach for co-designing digital mental health interventions with underserved perinatal populations.