MB materials are publicly available [30]. Once downloaded, MB’s curriculum for course leaders and the participant workbook served as the starting point for MBapp. The draft wireframe was created by this study’s principal investigator (AKL), who was supported by his coauthors—a multidisciplinary team with expertise in perinatal mental health, digital health interventions, and qualitative research methods. To create the draft wireframe, MB’s curriculum was first adapted to be delivered asynchronously on a smartphone app. Prior digital adaptations of MB reported that end users requested additional alterations to the digital draft content by further simplifying language and adding more visuals [20,31,32]. Thus, in MBapp’s initial draft wireframe, paragraphs of text or worksheets in MB were re-restructured to include short, direct phrases, colorful graphics (some from the MB curriculum), and pictures of pregnant and postpartum people and young infants. Some of MB’s videos were included within MBapp to allow users to learn the curriculum via a combination of text, graphics, and videos. The draft wireframe also contained ecological momentary assessments and gamification, features that have been shown in other perinatal mental health apps to increase the likelihood of sustained user engagement [33-35]. Specifically, MBapp included daily push notifications at users’ preferred times that prompt them to complete a mood scale. Their scores on this scale send them to pertinent content within the app. In terms of gamification, responding to at least 3 of the 7 daily ecological momentary assessment requests to complete the mood scale generated a digital reward, which can be visualized in a digital trophy case. Original MB creators confirmed the draft wireframe was adherent to CBT principles and retained fidelity to MB. Once the draft wireframe was completed, qualitative feedback was obtained from target end users to further adapt the intervention.

Participants were recruited from among those receiving perinatal care at a clinic that exclusively serves patients with government pregnancy-related health insurance (Medicaid), as having perinatal Medicaid insurance is only available for low-income patients who are pregnant in our state (Rhode Island, United States) [36]. To be eligible for this study, participants were English- or Spanish-speaking, aged at least 18 years, owned a smartphone, and were either in the third trimester of pregnancy or had an infant that was born within the prior 6 months. As MB prevents—but does not treat—PPD, volunteers were excluded if they reported prepregnancy or current anxiety or depression, an active (or within last 5 years) prescription for psychotropic medications, active engagement in psychotherapy, untreated substance use disorder, or if they were unable to consent. Potential participants were approached in person during routine antenatal or postpartum care visits. Those interested, who remained eligible after screening, provided written informed consent in person before the interview was scheduled. Participants who did not complete their interview after 3 no-shows at interview appointments were not rescheduled.

Individual in-depth interviews were all conducted via video teleconferencing at the participant’s preferred schedule and in their preferred language (English or Spanish). Each interview was conducted by the same investigator (AKL), who followed a semistructured interview agenda that divided the interview into two components. First, the interview discussed the participant’s personal experiences, preferences, and challenges with apps, particularly those that focus on perinatal or mental health. Then, the investigator shared his computer screen to review distinct components of MBapp’s wireframe with the participant and obtain page-by-page feedback. All interviews were audio-recorded, transcribed by a professional transcription service, and reviewed for accuracy with identifiers redacted. Participants were asked their race and ethnicity.

Similar to other web-based digital adaptations of MB [18,20], a human-centered design approach was used to optimize MBapp based on end user feedback about the intervention itself and other, optimal perinatal or mental health apps. We recorded MBapp’s evolution according to the FRAME (Framework for Modification and Adaptation) [37], an established method of systematic reporting adaptations and modifications to evidence-based interventions via end user feedback. The FRAME provides insight about when MBapp’s wireframe was modified during the development process and the reason for the modification. Feedback on MBapp’s themes, content, and function were extracted from the data and, after 5 to 7 interviews, the draft wireframes were revised to incorporate this feedback. The new iteration of the wireframe was then presented to subsequent participants to allow revisions suggested by earlier participants to be seen and commented on by subsequent participants. We anticipated recruiting up to 36 participants and planned to cease recruitment with content saturation, defined as 3 successive participants providing only positive feedback on MBapp’s wireframe, without further suggestions for improvement.

Authors AKL, MG, and KMG generated a preliminary codebook for the interviews based on the semistructured interview agenda. This initial version of the codebook was limited to deductive codes. Authors AKL, MG, KU, RB, and LL then collaborated via consensus coding in stages. First, two transcripts were coded collaboratively, and additional codes were added inductively to the codebook. Some of these codes were added de novo whereas other inductive codes were based on deductive codes that were reorganized to better capture data. Second, each of these authors independently coded a third transcript, and the authors met as a group to reconcile coding discrepancies. In this manner, consensus coding was achieved and a codebook comprising deductive and inductive codes was finalized. Then, all transcripts—including those that had already been reviewed—were coded independently by two authors (AKL and either KU, RB, or LL). A third coauthor reviewed transcripts and codes when coding disagreements could not be resolved (MG). All coding was then inputted into NVivo (Lumivero), and coded excerpts were organized and synthesized into thematic memos and discussed by the coding team. This study was reviewed and approved by Women & Infant’s Hospital of Rhode Island’s institutional review board prior to the consenting of the first participant.

This study was approved by the institutional review board of Women & Infants Hospital of Rhode Island (submission WIH 21-0021; approved on September 7, 2021). All participants were assigned a study identification number, which was stored on a password protected file and was only utilized to coordinate the $50 electronic gift card, which was provided to each participant after the study interview was completed. Otherwise, all data generated during the interview was deidentified and stored in a password protected computer network.

From August 2022 to August 2023, a total of 33 patients were consented, and 25 individual in-depth interviews were conducted. There were no differences in language of scheduled interview, patient-reported race, or parity between those who consented and did versus did not complete the interview. The sociodemographic and obstetric characteristics of participants who completed their interviews are described in Table 1. In brief, 24% of participants were aged 35 years or older, 24% self-reported as White, and nearly half (n=12, 48%) of interviews were conducted in Spanish. Approximately the same proportion of participants were pregnant versus post partum, and over half of participants were first-time parents (ie, either about to deliver or had recently birthed their first child).

Twenty-one of the 25 participants reported using at least one pregnancy-related app, and 2 of the 25 participants accessed a mental health related app during the perinatal period. The general consensus among perinatal app users was they were more likely to engage with an app if it provided pertinent, updated, and easy-to-understand text and video education either via a push notification or an email and contained digital rewards or games. Many perinatal app users noted that they turned to apps to receive perinatal education between prenatal care appointments or to supplement perinatal education received during these appointments:

Another consistent theme that emerged was that many participants engaged with perinatal apps to learn from people similar to them (those who were pregnant or had recently given birth):

Though apps were consistently described as useful throughout the perinatal period, the majority of users who had delivered at least one child at the time of the interview noted that apps were particularly helpful after delivery. The need for perinatal apps in the postpartum period was described in language suggesting that apps help fill the gap between in-person medical care from birth until 6 weeks after delivery. The long interval between medical appointments left participants feeling unsupported by their prenatal care team, and apps helped to provide support:

While prenatal apps were often used for additional perinatal education, there was a recurrent theme that these apps—such as prenatal care—focus more on the baby than on the mother:

Participants in this sample were much less likely to use mental health apps than perinatal apps. Mostly, participants attributed this difference to a lack of knowledge that such resources were available, not due to lack of interest in perinatal mental health apps. For example, one participant shared that she had depressive symptoms after delivery and informed her midwife of these symptoms at her postpartum visit:

Once the interview shifted toward talking about how MBapp aimed to provide app-based mental health support for perinatal people, the majority of participants responded very favorably, even before viewing any of MBapp’s wireframe:

The MBapp wireframe was adapted in multiple ways through end user feedback. Participants were asked about specific features in perinatal or mental health apps, and whether each feature was appealing (or not) and useful (or not). In addition, each participant viewed at least two distinct sections of MBapp’s wireframe and provided direct feedback on the appearance, appeal, content, and usability of each page. This feedback led to minor changes in MBapp’s font size or color scheme. Each participant also was asked general questions on how MBapp should be modified to encourage initial and sustained user engagement from the third trimester until 6 months post partum. All feedback from participants was considered by the study team and adaptations were incorporated into MBapp’s wireframe. Table 2 highlights the major adaptations made to MBapp’s wireframe during qualitative interviews and categorizes these changes according to the FRAME [37]. Of note, participant feedback resulted in both significant and nuanced changes to MBapp but did not alter MBapp’s core CBT content; thus, these adaptations did not affect MBapp’s fidelity to the evidence-based MB curriculum.

MBapp’s wireframe was originally created to mirror MB’s programmatic structure, in which a module on parenting education and infant care occurred midway through the curriculum and participants were instructed to complete daily mood scales between each in-person session, regardless of content. Based on end user feedback, MBapp was restructured such that the parenting module was first. We also edited the wireframe to provide users with the option to proceed with the daily mood scales for one week before accessing the next module or to defer these assessments in lieu of moving on immediately to the next module.

Participants recommended adding multiple new features to increase the likelihood of sustained engagement. These included modest financial reimbursement tied to responding to the daily mood scale and the option for audio narration, as well as an “ask a clinician” feature in which users can ask a nonurgent question to a member of the MBapp clinical team, who would respond through the platform. End users also desired short module summaries to be added after each module so they could review high-yield content without repeating the entire module. Module summaries now appear automatically on a digital bookshelf upon completion of each module. Figure 1 provides an example of how MBapp’s wireframe was adapted from the initial to final versions.

To our knowledge, this study represents the first to systematically report the process of adapting an in-person curriculum to a perinatal mental health app before the intervention was deployed. Indeed, this paper presents qualitative data from 25 diverse pregnant or postpartum perinatal patients with Medicaid insurance, describing their perspectives on perinatal or mental health apps and design recommendations for MBapp, a novel app based on an in-person CBT program shown to effectively prevent PPD among low-income patients [15,16]. The majority of participants used perinatal apps to receive on-demand pregnancy or postpartum education, and many preferred learning from others with similar experiences. Comparatively, mental health app use was less common, but this lack of uptake was attributed to lack of awareness, not disinterest.

Participants’ perspectives and recommendations were incorporated into MBapp’s wireframe, which led to minor adjustments in MBapp’s aesthetics and major adaptations to MBapp’s structure and features. All major adaptations were categorized according to the FRAME [37]. As each modification was derived directly from qualitative feedback from individual end users, incorporating minor and major adaptations to MBapp’s appearance, structure, and features should increase its appeal to future users.

MB was originally created as 6 in-person 2-hour group sessions [17] and has been implemented successfully in nontrial, community settings [38]. Since its inception, MB has been adapted into multiple formats. For example, 1 in-person MB curriculum comprises 10 individual sessions occurring during prenatal visits or at home [19] and can include supplemental text messages to participants [32]. There are also two asynchronous, online adaptations of MB, one for perinatal women and one for pregnant adolescents at risk for PPD [20]. Each digital adaptation of MB (text-based or online) included mixed methods research to optimize the intervention according to end user perspectives, and this feedback led to simplified language and more visuals in the final digital MB intervention [20,31,32]. Based on prior investigators’ experience with adapting MB to a digital intervention, we assumed that adapting MBapp would also need to include simplified and direct language and visual engagement via graphics and pictures of perinatal individuals and infants. The initial wireframe was developed with these features. Thus, all participants could provide granular feedback on features that were included or should be added to MBapp to encourage engagement, instead of focusing on giving the investigative team feedback on reducing the amount of text or adding visual appeal to the wireframe. It is important to note that only the online MB program for high-risk perinatal adolescents used formative qualitative interviews to adapt MB via a structured framework such as the FRAME, as is described in this paper. Incorporating this evidence-based approach ensured that all major MBapp modifications were well-justified and provided transparency to inform how future data on MBapp’s effectiveness may be interpreted.

Our study had many strengths. Each aspect of intervention creation and adaptation was grounded in evidence-based practices: (1) MBapp’s core content was derived from an in-person CBT curriculum known to be highly effective for PPD prevention [17-19]; (2) the individual in-depth interviews were performed according to established and recommended qualitative research practices as described by coauthors of our paper [39-41]; and (3) MBapp was adapted using a standard, commonly used framework for modifying established interventions [37]. In addition, our participants were diverse in terms of race, ethnicity, language, and parity, helping to ensure that different perspectives and opinions were included in our digital perinatal mental health intervention. Lastly, MBapp was created with a multidisciplinary team with expertise in perinatal mental health, digital health, and qualitative research, ensuring the digital health intervention was created via best practices.

Nevertheless, our study is not without limitations. First, our study population was comprised of 25 participants. Though a standard sample for a qualitative research study, this study population cannot possibly represent all opinions and perspectives of perinatal patients. Second, recruitment was limited to one obstetric clinic, which may reduce transferability of findings to individuals within different types of clinics (eg, academic vs community-based obstetric clinics). Third, MBapp will be a fully functional native app available on iTunes (Apple Inc) and Google Play (Google LLC). Compared to a web-based intervention that can be easily and cheaply modified after creation, native apps such as MBapp incur significant software developer costs for any modifications postcoding. Though we cannot alter MBapp postproduction as has been carried out in web-based perinatal mental health interventions [42], we are optimistic that using the FRAME and end user feedback has optimized MBapp’s wireframe to the point that additional edits are unnecessary. Lastly, though MBapp was deemed to have preserved fidelity to MB’s curriculum by MB’s original creators, it remains an empirical question whether an asynchronous digital education delivered via smartphone app is comparable to a 12-hour in-person group education led by a trained facilitator. Thus, the effect of MBapp on PPD will need to be explicitly and intentionally tested in future randomized trials.

In conclusion, the majority of our participants—pregnant or postpartum individuals with public health insurance—used apps to obtain on-demand perinatal education, and this expertise was leveraged to update and enhance a novel smartphone adaptation of MB. MBapp’s wireframe was tailored to the unique needs of app-based interventions and to feedback from target end users, which may result in sustained user engagement once the app is deployed. Our team’s next steps will be to confirm that MBapp is a feasible, acceptable intervention among English- or Spanish-speaking pregnant and postpartum people at risk of PPD through a pilot randomized trial. Ultimately, we aim to demonstrate that MBapp is an effective, scalable intervention that can prevent PPD on a population level.