Developing a Guided Web App for Postpartum Depression Symptoms: User-Centered Design Approach

Introduction

Background

Information and communication technologies offer a significant opportunity to overcome logistical and economic challenges in mitigating the burden of perinatal depression [1]. Most existing psychological internet-based interventions (IBIs) for the prevention and treatment of perinatal depression incorporate evidence-based techniques, such as cognitive behavioral therapy (CBT). Despite promising outcomes, these interventions often face high attrition rates, a problem persisting even with the inclusion of human support [2]. Addressing attrition is crucial, as it significantly impacts the overall effectiveness of the interventions [3]. The tendency to overly rely on theory- or expert-driven development methods without sufficient user involvement is a potential factor contributing to higher attrition and lower adoption rates [4]. Early and integrative inclusion of user involvement in the development process has been suggested as a strategy to address these challenges [1,4,5]. The Center for eHealth Research and Disease Management (CeHRes) Roadmap, developed by van Gemert-Pijnen et al [4], offers a guideline for eHealth development, combining evidence-based methods from different disciplines and approaches, such as user-centered design and persuasive technology. It specifically emphasizes the importance of integrating user feedback throughout the development process, ensuring that the intervention aligns closely with user needs and preferences, thereby potentially improving adherence and reducing attrition rates.

This study focuses on developing a psychological IBI aimed at addressing symptoms of perinatal depression through an iterative, user-centered approach. Initially, we evaluated the broader perinatal context to determine the most suitable period for intervention. On the basis of the exploration results and practical considerations, we decided to specifically target postpartum depression symptoms. The study aligns its objectives with the first 3 phases of the CeHRes Roadmap: contextual inquiry, value specification, and design. The overarching goal was to create a user-friendly system that reflects the values of potential users and other stakeholders. In the contextual inquiry phase, we aimed to explore the existing knowledge about IBIs for perinatal depression and determine specific issues where the IBI can add value in our context. This phase established the initial set of values and requirements. During the value specification phase, the focus shifted to further detailing these identified issues and translating the values into specific technological requirements. This phase involved understanding the needs of key stakeholders for an IBI to address postpartum depression symptoms and determining the main values and features that should be included in such an intervention. The outcomes of the value specification phase served as a comprehensive blueprint for the technology development in the design phase. This final phase involved developing the prototype, integrating persuasive elements to enhance user engagement, and conducting usability tests with potential users. We explored whether an early digital prototype and the prototype’s final version met the needs and wishes of potential users and gathered their suggestions for possible improvements.

Objective

Thus, the user-centered approach of this study, guided by the CeHRes Roadmap, inherently aims to enhance adherence by ensuring that the intervention aligns closely with user needs and preferences, fostering sustained engagement and potentially increasing its effectiveness. The project management team comprised researchers in perinatal mental health and eHealth technologies, clinical psychologists specializing in perinatal mental health, web designers, and software developers. In addition, the feasibility and acceptability of this intervention have already been assessed through a small-scale randomized controlled trial, the results of which have recently been published [6]. We are currently planning a large-scale trial to evaluate intervention efficacy.

Methods

Ethical Considerations

The Health Sciences Ethical Committee of Pontificia Universidad Católica (Santiago, Chile; protocol ID 210824004) approved all the development process methods. Informed consent was required for participation in each research activity. The development process took place from February 2020 to February 2023.

Contextual Inquiry

In this phase, we aimed to gain a better understanding of the Chilean context regarding perinatal mental health; the international evidence on IBIs for depression, specifically perinatal depression; and the perspectives and opinions of perinatal women and health professionals in Chile on perinatal apps. This comprehensive exploration enabled us to create our initial list of values and requirements.

Literature Scan

We conducted desk research to gain a deeper understanding of perinatal mental health within the broader Chilean context, as well as international evidence on IBIs for depression, particularly perinatal depression. In addition, we performed a systematic review, which aimed to explore the range and characteristics of psychological IBIs designed to prevent or treat perinatal depression, as documented in randomized controlled trials involving adult women.

The systematic search was conducted in 2020 in CENTRAL, Embase, PsycINFO, and PubMed databases using a combination of perinatal, depression, and internet terms in English. Article abstracts were screened for eligibility by 2 researchers, and eligible papers were retrieved for full-text screening. In developing the intervention, our objectives encompassed gaining insights into the effectiveness of the interventions assessed in the studies included in the review. We aimed to discern the therapeutic approach or theory underpinning these interventions, ascertain the content they covered, determine their length and the presence of modular components (if applicable), evaluate whether they involved human support, and review the adherence to the intended use of the intervention.

Web-Based Survey

In addition, we conducted a survey in this phase aimed at assessing views toward pregnancy and parenting apps among perinatal women and health care professionals in Chile. In perinatal women, we explored pregnancy and parenting app use, what they value in the apps they use, and what they considered an “ideal app” for supporting them during this period. We explored health professionals’ opinions on perinatal apps used by women and what they believed a perfect app for their clients should entail. The survey was conducted through a secure web-based platform between September 2020 and October 2020. Participants were recruited through posts and advertisements on social media. Further details regarding this survey can be found in a separate paper [7]. The survey results were thoroughly discussed with the development team, with a particular focus on the information concerning perinatal mental health and user preferences for app features.

Value Specification

The value specification phase aimed to explore the need for an IBI for perinatal depression and further elaborate on the main values and features that should be included in such an intervention. It involved conducting 8 web-based focus groups to explore the need for a psychological IBI for perinatal depression symptoms and to identify central values, attributes, and features that should be included in such an intervention. We conducted the focus groups with (1) pregnant women, (2) postpartum women, (3) pregnant couples, (4) postpartum couples, (5) pregnant women and perinatal health professionals, (6) postpartum women, (7) women who had undergone treatment and remission for postpartum depression, and (8) perinatal health professionals. Initially, we planned to conduct 7 groups, with 2 mixed groups involving perinatal women and health professionals (groups 5 and 6). However, after observing the dominance of health professionals in the discussions of group 5, which overshadowed the perspectives of pregnant women, we adjusted our approach for group 6. As participants were already recruited and interested in contributing, we chose not to cancel group 6 but instead conducted it exclusively with postpartum women to ensure their views were adequately represented. To still gather insights from health professionals without influencing the women’s responses, an additional exclusive focus group (group 8) was formed. These adjustments were made to optimize stakeholder engagement and data quality.

We aimed for 5 or 6 participants for each group. For focus groups 1 to 6, the inclusion criteria were as follows: perinatal women must be residing in Chile and being pregnant or having a baby aged 0 to 12 months (depending on the focus group). For focus groups 3 and 4, we asked women to attend with their partners. For focus group 7, the inclusion criteria were as follows: women who were residing in Chile, had a baby in the last 3 years, received a diagnosis and treatment for postpartum depression, and were discharged from therapy due to symptom remission at least 6 months ago. The inclusion criteria for health professionals were that they must be residing in Chile and being involved in the care of perinatal women. We provided the following examples in the recruiting flyer: psychologists or psychiatrists specializing in perinatal mental health, midwives, gynecologists, and doulas. All participants were recruited through posts and advertisements on social media.

The first 6 focus groups followed the same structure. After a brief introduction to the research project, we described what psychological IBIs are (as there are no psychological IBIs for adults in the country) and what they might entail. Participants were then asked for their thoughts on accessing an IBI to enhance their psychological well-being and address depression symptoms during the perinatal period. They were also asked to provide input on the characteristics and features they would like to see in the intervention. The focus group with women who had undergone treatment and remission for postpartum depression involved questions about their experience with face-to-face psychological treatment and what characteristics and features they would like to see in an IBI and the e-coaching. The focus group, consisting solely of perinatal health care professionals, also aimed to create descriptions of 5 “typical” Chilean mothers to exemplify the techniques taught throughout the intervention. At the end of each focus group, a book (illustrated about characteristic pregnancy moments) was raffled among the participants.

The 8 focus groups were transcribed and analyzed by 2 independent coders (one of whom is the first author, and the other is a researcher independent of the project). Inductive content analysis was used to identify response patterns [8]. The second and third authors reviewed the categories and subcategories map and provided feedback. The information collected was summarized based on the main themes and discussed with the project development team in requirement sessions (refer to Multimedia Appendix 1 for the focus groups script).

Design Phase

Overview

In this phase, we explored whether certain features of an early digital prototype and its final version met the needs and preferences of potential users. We assessed its user-friendliness and functionality and gathered suggestions for possible improvements. For this purpose, we conducted 2 focus groups, 5 usability interviews (UIs), and a technical pilot (TP). The inclusion criteria comprised women residing in Chile with babies aged 1 to 6 months. The criterion for the baby’s age was determined based on the findings from the previous phase (mentioned in the Results section). Participants were recruited through posts and advertisements on social media.

Focus Groups

During the focus groups, we presented a set of photographs and illustrations for the prototype to gather feedback on general impressions, represented situations, representativeness of Chilean mothers, and emotional tone. We also presented screenshots of a “Workbook” section (with therapeutic exercises) to assess the value of having this function and gathering feedback on the design. Subsequently, participants were asked to explore the prototype on their smartphones and provided feedback on its sections, user-friendliness, and design. We also assessed whether participants found it feasible to complete 1 module per week. Finally, participants were presented with a list of themes for developing infographics and asked to rate their relevance on a scale from 1 (not relevant) to 5 (very relevant).

For the focus groups, the following materials were prepared: (1) a set of photographs selected from an open database, (2) one developed illustration for each 1 of 5 example cases of mothers, (3) screenshots of the Workbook, and (4) a high-fidelity clickable short version of the prototype of the intervention. The prototype was developed on the Figma website (Figma [9]) and included the registration procedure, the first 2 intervention modules (Introductory module and Psychoeducation on postpartum depression), 1 infographic, the reading for the support network, and the Workbook section. The illustrations, the photographs, and the overview of the Workbook are presented in Multimedia Appendix 2.

The head of designers, who was present in the 2 focus groups from the design phase, took notes along with the interviewer. After each focus group, a debriefing meeting was held between the interviewer and the designer. The focus groups were transcribed and analyzed using inductive content analysis [8]. The information collected was summarized based on the main themes and discussed with the project development team in requirement sessions (refer to Multimedia Appendix 1 for the focus groups script).

UIs: Think Aloud Method

An extended version of the clickable prototype, encompassing all the modules and sections of the platform, was developed for the UIs. Participants accessed the prototype on their smartphones while using the think aloud method [10]. We encouraged participants to navigate the prototype on their smartphones and speak their thoughts aloud as they explored it. The interviewer’s questions primarily served as prompts to keep them talking and gain a deeper understanding of any points raised by the participants.

Given the extensive amount of information in the prototype, we instructed participants not to concentrate on the content but on aspects such as design, user-friendliness, and features. Toward the conclusion of the interview, participants were asked 3 questions: (1) “On a scale from 1 to 10, how would you rate the user-friendliness of the prototype, and why?” (with 1 indicating “not user-friendly” and 10 signifying “very user-friendly”), (2) “Would you consider using this intervention if you were experiencing low mood? If so, why?” and (3) “Would you recommend this intervention to a friend with low postpartum mood? Please explain your rationale” (refer to Multimedia Appendix 1 for the interview script).

The interviews were transcribed, and all comments were categorized into 2 groups: positive aspects highlighted by the users concerning the prototype and suggestions for improvement. These categorized comments were then organized into main themes and discussed during the requirements sessions.

TP Phase

After developing the technology, a small TP was conducted to ensure the functionality and reliability of the web app. This step is crucial to identify and address any programming flaws before proceeding with the feasibility and acceptability study. Ensuring that the app operates smoothly is essential to maintaining the integrity of the research and providing a reliable user experience. Given that the primary goal of this pilot was not to assess feasibility, acceptability, or effectiveness, it was not necessary to statistically calculate the sample size for this phase. Instead, we aimed to recruit a practical number of participants to allow for adequate testing of the technology’s functionality. Consequently, 10 women with infants aged between 1 and 6 months were recruited. Having symptoms of depression was not an inclusion criterion; however, we requested participants to complete the Chilean version of the Edinburgh Postpartum Depression Scale (Edinburgh Postnatal Depression Scale [EPDS]; 10 items, range 0-30; cutoff values of ≥10 identify women who might have depression) [11] before accessing the technology to characterize the sample. We also documented whether the participant was presently receiving mental health treatment. Recruitment was conducted using a convenience sampling approach, asking researchers and psychotherapists not affiliated with the study to refer potential participants.

Throughout this pilot phase, participants were granted full access to the technology, including e-coach support, for 6 weeks, enabling them to explore all its features. The primary objective was ensuring the functionality and reliability of the technology rather than collecting data on its impact or user opinions. Our main focus was to gather information related to log data and identify potential bugs (ie, coding errors) or issues. If participants encountered a bug or a significant usability issue, they promptly reported it to the study coordinator, who relayed this information to the software developers for timely resolution. We kept records of the number of completed modules and exercises submitted and tracked the identification and nature of any issues discovered.

As a secondary objective, we assessed satisfaction and engagement with the technology. Upon the conclusion of the 6 weeks, participants were asked to complete 2 assessment tools: the Spanish version of the Client Satisfaction Questionnaire-8 (CSQ-8; 8 items) [12] and a Spanish version (using back and forward translation) of the Twente Engagement with eHealth Technologies Scale (TWEETS; 9 items) [13]. Descriptive analyses were conducted to analyze the scores obtained from the CSQ-8 and TWEETS.

Formative Evaluation: Requirement Sessions

After each research method, the research team (comprising researchers in psychotherapy who are also psychotherapists and researchers in eHealth) discussed the results. The results were then communicated to the design and development team through reports written by the researchers, and those reports were discussed with them in several meetings. The results were translated into requirements by the designers and software developers, and the researchers verified whether the requirements met the expected needs.

Results

Contextual Inquiry

Literature Scan

In Chile, perinatal women attending public health facilities are universally screened for depression symptoms by the EPDS during the second pregnancy check-up and the baby’s 2- and 6-month checks [14]. If there is a suspicion of depression diagnosis, they are referred for further evaluation. In the country, adults diagnosed with depression have guaranteed access to treatment at low or no cost, which includes psychotherapy, pharmacotherapy, or both [15]. Despite these measures, perinatal depression prevalence in Chile remains high, and it is often undertreated [16]. The barriers for perinatal women seeking mental health care include misconceptions about mental disorders and psychotherapy, fear of being judged as a “bad mother,” lack of time or energy, and difficulties attending health care centers in person [17]. Providers’ barriers include a lack of training on mental health for perinatal health care professionals [17], long waiting lists for mental health services, frequency and duration of sessions below international recommendations, and high staff turnover [18].

Moreover, there is a lack of interventions in the country for perinatal women exhibiting depressive symptoms who do not meet the criteria for a major depression diagnosis, such as subthreshold depression. Subthreshold depression is more prevalent than major depression, making its associated costs comparable, and has a significant impact on quality of life and health status, with an increased risk of developing major depression [19,20]. Psychological interventions aimed at reducing subthreshold depression symptoms and preventing the onset of major depression [21] could also help reduce the treatment gap. By contrast, studies in the country have reported a high prevalence of anxiety symptoms among perinatal women, with about 40% of women presenting moderate to severe symptoms of anxiety according to the Perinatal Anxiety Screening Scale [22]. This finding is significant, as evidence indicates that comorbid perinatal anxiety and depression are common and associated with greater symptom severity [23].

Regarding the systematic review of psychological IBIs for preventing and treating perinatal depression, only the most relevant information for the intervention development is presented here, with detailed conclusions published elsewhere [24]. The systematic review identified 15 effectiveness studies on interventions for perinatal depression [25-39]. The majority were conducted in high-income countries. We identified a multicenter study conducted by a research team in the United States involving participants who spoke both Spanish and English and included individuals from 23 countries, including South American nations [27]. However, we could not identify a study specifically conducted within South America.

In 11 studies, significant differences favoring the intervention were observed in depression symptoms in the intervention group compared with the control group, which predominantly consisted of treatment as usual and, in a minority of cases, a waitlist group. Most of the interventions used a CBT approach. Approximately half of the interventions included some form of human guidance or support, typically delivered by a clinical psychologist. The interventions were usually structured into sequential modules meant to be completed weekly. Most incorporated multimedia-presented therapeutic content and homework assignments for practicing the acquired skills. Case examples were frequently integrated to illustrate therapeutic concepts. Most studies identified in the systematic review that assessed intervention acceptability showed good participant satisfaction levels. However, many studies reported a high rate of intervention attrition.

The literature scan showed some characteristics that might enhance the effectiveness of an IBI: providing some form of human support [3,40,41], ensuring cultural sensitivity [42], incorporating persuasive technology features [4,43,44], and having a 7-module length [3]. By contrast, research on face-to-face treatments for perinatal depression highlights the importance of improving the quality of the mother-infant relationship as a protective factor for both the mother and child, achieved through enhancing maternal sensitivity and the mother-infant bond [45]. Theories rooted in mentalization and attachment play a vital role in interventions for mothers with infants by promoting secure bonding, improving comprehension of the baby’s needs, facilitating emotional regulation, and nurturing healthy infant development [46,47].

Our literature scan also identified self-help books and manuals for perinatal distress, depression, and anxiety.

Web-Based Survey

The survey was completed by 451 perinatal women and 54 perinatal health professionals. Here, we provide a summarized overview of the most important findings for the development process; detailed results are published elsewhere [7]. The results indicated that perinatal women in Chile frequently use perinatal apps, and both women and health care professionals hold a positive view of them. The most highly valued aspect of these apps was the information they offer, which mainly targets baby development in the uterus or after birth, week-by-week changes during pregnancy, and guidance on maintaining health throughout pregnancy and postpartum. The survey further unveiled various requisites for an “ideal app.” A visually appealing interface with visuals such as images, illustrations, graphs, or videos, as well as user-friendliness and personalization options, were sought-after features. Content encompassing both pregnancy and postpartum aspects, grounded in evidence or endorsed by health care professionals, and offering relevant and up-to-date tips was deemed crucial. Participants desired the capability to interact with both perinatal health care professionals and other perinatal women in the community.

Addressing women’s mental health emerged as a prominent concern among both perinatal women and health care professionals, an area seemingly underaddressed in the apps they use. Recommendations for mental health care included strategies for mental health promotion, psychoeducation, well-being monitoring through questionnaires, advice on managing perinatal depression and anxiety symptoms, and guidance on seeking professional help.

Both perinatal women and health care professionals expressed a desire for app sections targeting partners. In addition, perinatal women emphasized the importance of the app addressing the couple’s transition to parenthood.

Requirements Sessions

At this stage, it was decided that the IBI would primarily serve as a research tool. This decision was motivated by practical considerations, as there was no specific implementation setting for the intervention, such as a participating mental health care organization or institution. The absence of institutional participation also meant that the budget for developing the intervention was very limited. In addition, the project team aimed to experiment with the IBI, and this approach was deemed most feasible as a research tool. Consequently, the most critical stakeholder groups were identified. Given the focus on a research implementation setting, care providers played a less prominent role in the development process during this stage.

Several decisions were reached during the researchers’ meetings, drawing from insights gathered during the contextual inquiry phase. The literature scan on the Chilean context has highlighted several critical needs that our intervention aims to address. First, there is a pressing need for interventions that are easily accessible and flexible, particularly for managing major perinatal depression, to reduce the treatment gap. Second, there is a requirement for interventions designed for the indicated prevention of perinatal depression, which could prevent the progression from subthreshold to major depression. Finally, given the high prevalence and impact of comorbid anxiety symptoms among perinatal women, it is crucial for the intervention also to address anxiety if comorbidity is present. These specifications are essential for developing a comprehensive and practical approach to perinatal mental healthcare. The intervention will be rooted in CBT principles, supplemented with elements from attachment and mentalization theories, and will include homework exercises with personalized feedback provided by an e-coach. The intervention should compass psychoeducation, strategies for promoting mental health and managing perinatal depression and anxiety symptoms, and guidance on seeking professional assistance. In addition, it should include information on how to enhance the couple’s relationship functioning during the transition to parenthood and provide information for partners.

The language, examples, and visual materials must be culturally sensitive. The technology must have a visually appealing interface, be user-friendly, and allow some personalization. Persuasive elements will be incorporated following the persuasive systems design framework [43].

Value Specification

Participants

A total of 39 perinatal women and health care professionals participated in the 8 focus groups. None of the women from the pregnant and postpartum couples’ groups attended with their partners, so these groups were conducted with only women. Perinatal women had an average age of 32.6 (SD 4.6; range 25-42) years. Pregnant women were, on average, at 28.7 (SD 7.3; range 15-39) weeks gestation. The average age of postpartum women’s babies was 4.2 (SD 2.4; range 1-9) months. More than half of the perinatal women (22/31, 71%) were primiparas. Most perinatal health care professionals (6/8, 75%) were psychotherapists specialized in perinatal mental health. Table 1 shows the participants’ characteristics presented by group across all phases.

Participants’ Expected Needs

Table 2 shows the perceived value of an IBI for perinatal mental health and the desired characteristics and features from the participants’ perspective, as derived from focus groups 1 to 7, with quotes to exemplify each category.

In focus group 8 with perinatal health professionals, the discussion explored the most frequent problems or challenges and depression or anxiety symptoms reported by postpartum women in consultations. The debate served as input to create the intervention’s fictitious characters (eg, the “diligent mother” with perfectionist traits and high self-expectations who invested significant effort in educating herself extensively about infant care during pregnancy but then experienced frustration when she realized that motherhood does not unfold as she initially anticipated).

Requirements Sessions

The focus groups’ results were discussed initially with the research team and later with the designers. Some categories overlapped with those that emerged in the survey study. The focus groups reinforced the choice to include therapist support. Despite strong demand for support from other perinatal women in both focus groups and surveys, it was not included at this stage, given the intervention’s research focus and the desire to assess its psychotherapeutic components in the future.

Opinions among participants were divided regarding the preferred timing of the intervention during the perinatal period. All pregnant women (17/17, 100%) wanted the intervention to focus on the pregnancy period or both pregnancy and postpartum. Regarding postpartum women, about half (8/14, 57%) preferred the intervention to focus exclusively on the postpartum period, while the rest preferred it to cover both pregnancy and postpartum. When postpartum women (who had recent experience with both periods) were asked to choose only one period, the majority expressed a preference for the postpartum period (13/14, 93%). In addition, postpartum women emphasized the importance of receiving this intervention in the early months following childbirth. Considering these preferences and economical and practical factors and the specific barriers associated with the postpartum period (eg, related to baby care and attending in-person health centers), the decision was made to target the intervention at postpartum women, specifically during the first 6 months after childbirth. During this stage, the intervention modules and sections were defined. The content was divided into short, easy-to-read modules that can be read in 10 to 15 minutes. Modules include providing psychoeducation on depression and anxiety, promoting mental health care help seeking, and providing CBT strategies for dealing with negative emotions. The homework exercises were carefully chosen and adapted from the therapy manual for perinatal depression and anxiety by Green et al [48], which covers the core principles and skills of CBT.

It was decided to include infographics on relevant postpartum topics (such as body changes), a shareable guide on supporting mothers with depression symptoms, an “About” section detailing the development team and funding sources, a “Resources” section with links to maternal and mental health support, and a “Favorites” section for saving preferred content.

The insights from the focus group of perinatal health care professionals informed the creation of fictional characters, who serve as practical examples to illustrate the content and therapeutic techniques. Our design team crafted illustrative representations of these characters in everyday scenarios to ensure a realistic portrayal. To ensure inclusivity for lesbian couples, the gender of the partners of the fictional women was left ambiguous in the text and illustrations. Moreover, we have carefully chosen inclusive language throughout the written content, consistently using “your partner” instead of “the father of the baby.” In addition, one of the female characters in our narratives does not have a romantic partner, reflecting the diversity of family configurations.

Subsequent meetings with designers and software developers focused on refining the interface to ensure it is user-friendly and easy to navigate. It was also defined how elements of persuasive systems design [43] would be applied to the intervention (Table S1 in Multimedia Appendix 3 [44]). Table 3 summarizes the key values, attributes, and requirements for the intervention that emerged from the contextual inquiry and value specification phases.

Design

Participants

A total of 10 women with babies aged 1 to 6 months participated in focus groups 9 and 10, 5 in the UIs, and 10 in the TP. The mean age of the women and their babies was 33.7 (SD 6.3; range 25-43) years and 3.6 (SD 1.4; range 1-6) months, respectively, in the focus groups; 33.6 (SD 6.9; range 26-39) years and 5.2 (SD 1.2; range 4-7) months, respectively, in the UIs; and 34.5 (SD 3.5; range 31-40) years and 3.8 (SD 1.3; range 2-6) months, respectively, in the TP. The majority (16/25, 64%) were primiparas.

In the TP group, women had an average score of 9.60 on the EPDS (SD 5.30; range 3-18); 4 (40%) women were currently undergoing psychological treatment for postpartum depression symptoms. A total of 7 (70%) women completed the CSQ-8 and the TWEETS after the 6-week access period.

Focus Groups: Participant Feedback

Feedback on the Set of Illustrations

Participants appreciated that the illustrations depicted challenging daily situations faced by mothers. Moreover, 1 participant expressed how the images shattered unrealistic myths about motherhood:

I find that they are very realistic and also tear down those myths of motherhood or pregnancy, of moms always happy and almost making postpartum depression invisible, when in truth, one is very tired, one experiences a lot of emotions, one has little energy.

In both focus groups, 1 participant mentioned wanting the mothers to be more accompanied:

[I]n the end, maternity is quite lonely, but eh, I think that too, but I don’t know, the message could be that I need to be more accompanied, to have more tribe, more family around.

The participants generally agreed that the mothers represented the “Chilean mother visually,” with good diversity among them. While most participants found the facial expressions realistic and relatable, 2 felt that the sole display of overwhelmed expressions was not ideal:

The facial expression came to me; it was the first thing I looked at... “Oh, the afflicted face, it’s me all the time,” I saw myself in all their faces... It does come to me, and I connected with the images, how I am in bed, lying with the baby on top.

I know that motherhood is difficult, but seeing a super sad, super frustrated mom makes me sad because I try to be there for my baby and not be sad.

Feedback on the Set of Photographs

Participants appreciated the portrayal of a wide range of emotions, positively valued the esthetics of the photos, and especially praised the depiction of family diversity and coparenting situations in the various photographs:

[H]ere you can see different situations, not only the negative side, but the positive side of motherhood, enjoying her baby or also with friends or with family, or having time for themselves, as writing or meditating.

However, some participants (3/10, 30%) felt that the photographs did not fully represent the diverse profiles of Chilean mothers, as they primarily depicted young and slim women, thereby representing mainly high-income mothers and not the broader population. In addition, a few participants (2/10, 20%) felt that some photographs appeared too professional and posed, making it difficult to identify with them.

Feedback on the Workbook

Participants positively evaluated the inclusion of a Workbook, highlighting the value of receiving feedback from a psychologist, which conveyed a sense of personalized conversation. They found the idea of having a professional behind the technology to be innovative and supportive, which could increase the use of it as a meaningful support for those struggling with postpartum depression:

For me personally, the fact that one has the option of writing and receiving feedback already gives me the feeling that I am speaking to someone. It’s not like a predetermined survey, and then they give you an automated result, so I like the idea of having someone who is actually going to read and respond something personalized somehow.

Feedback on the Short Version of the Prototype

All participants (10/10, 100%) found the short version of the prototype user-friendly and intuitive. The colors and design were well received and perceived as “pleasing to the eye.” The font size and amount of text were considered appropriate and easy to read. The quality, depth, and comprehensiveness of the information provided and the feasibility of reading 1 module per week were well evaluated:

It is super friendly, the text does not seem so much to me, and it is easy to read; It is not saturated with things; sometimes they [apps] are saturated with images and icons; this one is simple, so it makes me want to read it.

You can tell it [the information] was written by someone who wanted it to be easy for you to read and understand.

The inclusion of fictional characters and the presence of features like the favorites or saved button, the infographic (guide for healing), and the reading for loved ones received positive feedback:

I found it super positive that there are cases about depression or symptoms, such as stories, because that also helps you to identify yourself with some stories and thus also say, “heck, I’m not the only one that happens to me,” or “maybe I have these symptoms.”

[I]t is important [the reading for loved ones] because often relatives, instead of helping, make it worse, make you feel bad or sometimes say super unfortunate things.

Some participants expressed a desire for specific information or features that were not included in the prototype they explored. These desires involved accessing references for the information provided, including a resources section containing articles or books, and integrating a depression and anxiety symptoms questionnaire with automated feedback.

Feedback on the Topics of the Infographics

Participants found most of the mentioned topics (1-6) relevant for an infographic. In each group, on average, every subject was rated above 4.3 (on a 1-5 scale), except for “Coping with postpartum rage,” which had an average rating of 3.3 in 1 focus group and 5 in the other. A few participants (2/10, 20%) suggested adding infographics on returning to work after maternity leave and the baby’s sleep.

Requirements Sessions

In response to the feedback received, we adjusted the illustrations, ensuring that they depict a broader range of emotions and incorporate other members of the mother’s social network. We also introduced an infographic addressing returning to work after maternity leave. Furthermore, we have included a section that contains references for all the information used and another one that provides a list of websites and hotlines offering support for maternity and mental health, including websites with evidence-based information about baby’s sleep.

User-Based Usability Evaluation

Feedback on Design, Sections, and Features

The color palette and the inclusion of illustrations and photographs accompanying the text were positively evaluated:

I find the presentation of colors kind of tender, calm... that pastel purple is a calm color; it is not an aggressive color because there are applications that have aggressive colors, but not this one.

It is nice to have photographs between the text, so it’s not so overloaded with text.

All participants (5/5, 100%) liked the illustrations and photographs, finding them “pretty” and “attractive.” They also highly appreciated the fictional mothers narrating their stories and exemplifying the content.

The length of the text was deemed appropriate. All participants (5/5, 100%) found the text size suitable (none had vision problems). All the sections of the prototype (Modules, Favorites, Workbook, Infographics, About, Resources, and Reading for loved ones) were evaluated as valuable.

Feedback on Prototype’s User-Friendliness

All participants (5/5, 100%) mentioned that they found the prototype user-friendly. When asked to rate the prototype’s user-friendliness from 1 (not friendly) to 10 (very friendly), participants gave ratings between 7 and 10 (mean 8.6, SD 1.1). Some suggestions arose when asked what was needed to rate it the maximum score: including a night mode, increasing the text font size to be more inclusive to visually impaired people, enabling a quick return to where they left off in the module when returning to the platform, including a depression or anxiety symptoms questionnaire, and displaying the logos of the institutions listed in the resources section to make them easier to identify. Each suggestion was mentioned by 1 or 2 participants:

It is friendly, easy to log in, easy to follow the steps.

It seems very friendly to me; the only thing I would need would be the night mode because I will probably read it at night... 9, as I told you, the only thing lacking for me is that.

Willingness to Use the Platform if Struggling With Depression Symptoms

All participants (5/5, 100%) said they would use the platform if struggling with low mood or depression symptoms. They valued its user-friendliness; its friendliness in tone or images; the content provided; having all the information on perinatal mental health and management in one place; having the option to interact with an e-coach; and trusting the platform because it was designed by a Chilean team of professionals, perceived as trustworthy:

Yes, because it looks like it has all the practical information on how to deal with those symptoms and in a platform made by a team that makes you trust them... instead of Googling it.

Willingness to Recommend the Platform to a Friend

All participants (5/5, 100%) expressed strong receptivity toward recommending the prototype to a friend struggling with postpartum low mood or depression symptoms, using words like “definitely,” “totally,” and “sure.” The information provided was highlighted as its primary value.

Yes, so she can inform herself and read about her psychological process, understand herself more, and at what point she should seek help or how to communicate it to her family.

Requirements Sessions

The results of the interviews were discussed with the project development team. Given the positive feedback on the design and user-friendliness of the Figma prototype, it was decided to develop the actual version of the IBI using the Figma prototype as the blueprint. The intervention had to be accessible by smartphone, enable communication with the e-coach, and not necessarily require installation. It was decided to create a web application, which is a software application that runs on a web browser and is accessed over the internet. Unlike traditional applications that must be installed on a computer or mobile device, web applications do not require installation. Users can access them directly through a web browser. Web applications are designed to provide interactive functionality and offer various features and services, similar to mobile apps. For the TP, we requested the participants to add a shortcut to the web app on their phone’s screen, enabling them to access it quickly, such as a mobile app.

The points for improvement were discussed. While the project team agreed that adding a night mode would be beneficial, it was decided not to include it in the current version due to time and resource constraints, though it is planned for future updates. The logos of the websites and hotlines from the resources section were included next to their names.

TP Phase

Detection of Bugs and Usability Improvements

All 10 participants registered on the web app. A total of 4 (40%) participants completed all main modules, and 2 (20%) completed the additional modules. While 7 (70%) participants submitted at least 1 exercise, 2 (20%) submitted the exercises of all the main modules, and 1 (10%) also sent the exercises of the extra modules. The therapist provided feedback on the exercises within 48 hours. Throughout the study, participants reported 8 bugs, all related to nonfunctional clicks. The software developers promptly addressed and resolved each issue.

In addition, 2 suggestions were made to improve the web app’s usability concerning the exercises. At that moment, users could only know they received the e-coach’s feedback on their exercises by navigating to the Workbook and finding the specific exercise they submitted, where the feedback is displayed beneath their text. In response, 2 participants expressed their desire for a notification system to promptly inform them of feedback pending reading upon accessing the web app. The other suggestion came from 2 participants, who proposed enabling an autosave function when doing the exercises. They mentioned encountering instances when they could not finish and submit an exercise, which meant redoing the entire exercise from scratch later. The first suggestion has been addressed with a visual notification system within the web app, while the second suggestion will be considered for a future update.

Satisfaction and Engagement With the Web App

The overall CSQ-8 sum score averaged 28.14 (SD 4.29), with only 1 participant assigning the lowest possible score to a single item. The items, ranked from the highest to the lowest mean (in a range from 1 to 4), are as follows: overall satisfaction (3.71), quality of service (3.71), deal with problems (3.71), amount of help (3.71), recommend to a friend (3.57), come back (3.57), type of service (3.28), and met needs (2.85) (see Table S2 in Multimedia Appendix 3 for CSQ-8 item response frequency).

The mean total TWEETS score was 3.20 (SD 0.55), and no participant assigned the lowest possible score to any of the items. The items, ranked from the highest to the lowest mean (in a range from 0 to 4), are as follows: “I enjoyed using MTE” (3.57), “MTE takes me little effort to use” (3.57), “I was able to use MTE as often as needed to improve my mood” (3.42), “I enjoyed seeing the progress I made in MTE” (3.36), “MTE helped me to get more insight into my thoughts and emotions” (3.28), “MTE fits me as a person” (3.23), “MTE made it easier for me to work on improving my mood” (3.28), “MTE motivated me to improve my mood” (3.14), and “MTE was part of my daily routine” (2.00; see Table S3 in Multimedia Appendix 3 for TWEETS item response frequency).

The Internet-Based Intervention

In this subsection, we describe the resulting IBI. Mamá, te entiendo (“Mom, I get you”) is a CBT web app intervention that includes elements from attachment and mentalization theories. The intervention consists of 6 main sequential modules and 3 extra modules (Table 4). Five case examples of mothers illustrate depressive symptoms and techniques. There is also an introductory module, and a guide about maintaining changes (relapse prevention) is displayed when the last main module is completed. Women can select any part of the content as “add to favorites” and save it in a “Favorites” section. The optional homework exercises are included in the modules alongside the related content and in a separate “Workbook” section. Participants receive feedback from an e-coach through the web app. They can also contact the e-coach to resolve doubts about the content and exercises. A sample of screenshots of the web app is presented in Figure 1.

Another section features infographics covering various pertinent postpartum mental health topics, including (1) a guide to healing; (2) strengthening your support network; (3) coping with postpartum rage; (4) navigating changes in your postpartum body; (5) handling doubts, comparisons, and information overload (empowering yourself as a mother); (6) mentalizing your baby; (7) nurturing the couple’s relationship during the postpartum period; and (8) returning to work after maternity leave. In addition, the app includes a shareable document with insights on supporting a loved one struggling with symptoms of depression. A resources section provides links to websites and hotlines for mental health and maternity support (most of them from institutions within the country). The app also dedicates a section to introduce the team responsible for its development and references for its content. Finally, a contact section enables participants to send messages to the IT support team in case of technical issues.

‎

Discussion

Principal Findings

This study aimed to create a user-friendly technology for perinatal depression symptoms that aligns with stakeholders’ values and to evaluate the development process. The intervention development followed a user-centered design approach, actively involving perinatal women and health care professionals in providing input and feedback. The process was iterative and used a variety of research methods. The invaluable insights from stakeholders, combined with a literature study, shaped the web app’s content, layout, and features.

The overall development process was satisfactory. The development team gained valuable insights into the importance of iterative testing and user input. Different research activities were used to achieve a wide range of goals, providing crucial insights into users and the context, which guided the refinements of the web app. Each step was built upon the knowledge gained from earlier stages and informed the next steps. The literature study helped us understand the potential of addressing perinatal depression through a psychological IBI in Chile; solved questions related to therapeutic content; and offered design recommendations, such as including fictional characters. Surveys provided insight into attitudes toward existing perinatal health apps and identified the needs and preferences for an “ideal” technology for the perinatal period. The first round of focus groups offered insights into stakeholders’ attitudes toward a potential intervention for reducing perinatal depression symptoms and enhancing well-being as well as identifying values and requirements for such an intervention. The second round of focus groups helped collect input to improve the prototype. UIs provided valuable feedback on the web app’s user-friendliness and users’ initial impressions. The TP assessed the app’s functionality and provided initial feedback on women’s satisfaction and engagement.

However, it is essential to recognize that a one-size-fits-all approach does not apply to development processes. Our method, which included several standard methods, could have been further enhanced by integrating innovative and participatory techniques, as Kip et al [5] recommend. By contrast, we faced challenges in harmonizing contributions from multiple participants, selecting relevant information and materials, and working within time and resource constraints. These challenges were addressed through proactive problem-solving and collaborative efforts among team members. For example, while there was a high demand for a feature allowing interactions with other mothers, its implementation was not feasible not only due to logistical constraints and limited resources but it diverged from our objectives. The multidisciplinary research team and requirement sessions were crucial in addressing these emerging challenges. However, not all team members could always be present at meetings due to time, language, and time zone constraints. In some cases, debriefing sessions were held with specific team members to discuss project progress and decisions, which were then addressed in subsequent meetings with other team members. Nevertheless, it would not have been practical to have all the meetings with all team members, as some were highly technical or involved decisions that did not require input from the entire team. However, the project’s principal investigator attended all meetings.

We also learned important lessons regarding the evaluation of materials. For instance, negative feedback regarding the photographs mainly emerged when they were presented in isolation during focus groups, but no such comments arose when viewing them within the prototype context. The inclusion of these photographs was well received, likely because they resembled those used in other perinatal mobile apps or websites, providing a different perspective when viewed in context. The last underscores the importance of not relying solely on out-of-context feedback and the need to incorporate multiple methods while verifying earlier findings.

Postpartum women appreciated the integration of realistic illustrations and photographs, which enhanced their connection with the material. Furthermore, including 5 fictional characters throughout the content effectively conveyed the requirements of showing the “real motherhood” and incorporating testimonials from women. This approach aligns with other IBIs for perinatal depression that have shown potential efficacy [32,33,38]. Such inclusion helps women dispel “motherhood myths” and normalize and validate their thoughts and experiences with motherhood, as highlighted in qualitative studies related to IBIs [29,49].

In addition, the sense of “being on the same page” may have reduced this target group’s need for personalization and tailoring. Aside from the personalized feedback provided by the e-coach, the optional nature of the exercises, and the availability of extra optional modules, participants did not emphasize the importance of personalization in our study. Instead, “group-based” content personalization, designed to target the intervention as a whole for postpartum women with depression symptoms, appeared to be more relevant. The last suggests that a general intervention for depression (without explicitly addressing postpartum issues) may not appeal to this target group. The absence of personalization requests may also be due to the novelty of the intervention, as there are no similar web-based interventions for adults in the country, leaving women uncertain about what to expect regarding personalization. Nevertheless, we believe that exploring additional personalization strategies in the future, as suggested by Hornstein et al [50], is a worthwhile endeavor. Adapting the content and design of an intervention to users’ needs aims to maximize adherence, engagement, and, ultimately, treatment efficacy, as emphasized by Ludden et al [51] and Kelders [52].

Future plans also include the consideration of implementing a night mode for reading in low-light conditions and adding an autosave function to prevent users from losing progress. In addition, we will explore the possibility of conducting a photo shoot with “Chilean moms” to enhance the representation of images in the intervention. Despite these considerations, the positive feedback on content and features, along with favorable assessments of user-friendliness, satisfaction, and engagement, supports the web app’s readiness for further testing. Notably, a feasibility and acceptability study has already been conducted, yielding promising results that have recently been published [6]. Motivated by these findings and participants’ acceptability feedback, we are updating the intervention and proceeding with a larger-scale study to continue evaluating its impact.

Study Limitations

Our study has some limitations concerning the samples of perinatal women. One limitation is that the presence of depressive symptoms was not used as an inclusion criterion for participants in the various research activities. While at least 1 participant in each focus group, interview, and TP mentioned undergoing treatment for depression, actively including a more significant number of women with depressive symptoms might have yielded more focused input from this group. Nevertheless, given that the recruitment material emphasized an intervention to enhance perinatal mental health, we believed that it would naturally attract a significant proportion of individuals grappling with this issue. Another limitation is that the target users who participated in the contextual inquiry (and likely in the value specification and design phases) were primarily highly educated women. We did not assess the education level during the value specification and design phases, but since the recruitment process resembled that of the contextual inquiry, the samples may have similarities. Another limitation concerning the study participants is that all procedures were performed with small groups of participants, potentially limiting generalizability. Nevertheless, it is important to note that our findings should be considered a preliminary step and must be confirmed within the specific audience for the intervention. This iterative approach aligns with the fundamental principles outlined in the CeHRes Roadmap [4].

Another limitation is the limited involvement of health care providers in the development process. This decision was made during the contextual inquiry, where we focused on research as the implementation setting. However, if we choose to expand the development of this intervention for a different implementation context, it is imperative to include health care providers in the process. In addition, fathers or lesbian partners did not contribute to the development process. Partners were invited to participate in 2 focus groups during the value specification phase, but the women attended these sessions individually. The last occurred despite our efforts to involve partners, highlighting a gap in our recruitment strategy. We did not initially plan to organize a focus group exclusively for partners. We acknowledge that this may limit the scope of our findings and potentially introduce bias, as partners play a significant role in the perinatal experience. Although participants appreciated the fact that the web app was designed “for them,” considering the importance of the couple’s relationship during the transition to parenthood and the expressed desires of mothers to involve their partners in some way, it may be beneficial to include partners as stakeholders in the future. Future formative research should involve partners as active stakeholders to enrich the intervention’s content and appeal and to ensure that it supports the dynamics of the couple’s relationship during the transition to parenthood.

Finally, in this study, we did not evaluate the impact or effectiveness of the intervention, so we cannot determine if it successfully reduces depressive symptoms in the target group. Future research is necessary to assess if users will use the intervention and if it achieves the intended effects. To improve the intervention’s design and effectiveness, it is crucial to use those methods that objectively measure use, gather qualitative feedback on user satisfaction and acceptability, and assess clinical effectiveness [53].

Conclusions

In conclusion, the IBI Mamá, te entiendo is a significant step toward addressing the pressing issue of perinatal depression in Chile. Through a comprehensive and user-centered design process, we aimed to create an accessible and culturally sensitive platform to support postpartum women experiencing depression symptoms. The integration of CBT principles and elements from attachment and mentalization theories and the inclusion of realistic case examples and infographics have resulted in a well-rounded intervention that targets the unique challenges that mothers face during the postpartum period. The web app offers evidence-based strategies and psychoeducation in a user-friendly format, potentially empowering women to manage depression and anxiety symptoms. An e-coach feature is included to bridge the gap between professional care and self-help, offering additional support. Positive feedback from perinatal women during the intervention development indicates its potential as a valuable mental health tool. If effective, it could significantly improve perinatal mental health in Chile and serve as a model for similar interventions in other Latin American countries.