To adequately modify our existing tool to address domains of disparity common to Latino pregnant people in a specific community, we drew from the Transcreation approach developed by Nápoles et al.(22) Transcreation outlines the steps for a community-engaged process of planning, delivering, and evaluating interventions to reduce health disparities in specific underserved communities. The focus is on community fit and interventions that fit the targeted community and also improve outcomes.
There are seven steps described by the Transcreation Framework(22): 1) identify community infrastructure and engage partners, 2) specify theory, 3) identify multiple inputs for the new program, 4) design intervention prototype, 5) design study, methods, and measures for community setting, 6) build community capacity for delivery, and 7) deliver the intervention. In this paper, we focus on the first four steps to identify design components for the new Spanish version of the MyHealthyPregnancy app.
Identify community infrastructure and engage partners
Our initial step was to identify current infrastructure and partners with relevant experience to understand the context in which the App would be delivered. For this purpose, we engaged several stakeholders who treat, teach, and support Latino women. We began with community outreach and word-of-mouth to identify stakeholders w in Pittsburgh, PA, which is currently considered a U.S. Emerging Latino Community.(23) Our focus was on healthcare and community organizations working with Latinos and educators who serve Latino patients. These stakeholders then recommended contacts they believed could offer additional insight (e.g., local Spanish-speaking doulas). Included in these informational interviews were three healthcare providers, three prenatal educators, two doulas, two leaders from community organizations, one academic researcher, and one social worker.
1. Specify a Theory
In this step, we reviewed published literature to determine areas to address in order to understand health disparities and the specific context of this population. The Health Equity Implementation Framework by Woodward et al.(24) was used to organize literature around the different components believed to predict a successful and equitable implementation (Figure 1). This framework was designed to assess both implementation and health equity determinants at the same time. Along with the contextual factors primarily considered in implementation frameworks, Woodward et al. place a focus on the clinical encounter. This inclusion reflects how the unique patient-provider relationship can affect health disparities for vulnerable populations. Finally, they recognize how societal influence, including economies, physical structures and sociopolitical forces affect a patient’s ability to access and experience healthcare services.
Review of the literature in these areas identified eight primary sources of disparity for the pregnant Latino population: language (low English proficiency), access to health insurance, immigration status, cultural factors, location of services and transportation, healthcare accessibility, and health literacy.(23,25–32) These domains reveal how societal influence and context can conflict with communication goals, interpersonal objectives and exacerbate potential discrimination, eventually lowering participation and engagement with healthcare interventions, including mHealth apps.
3. Identify and Use Multiple Inputs
In transcreation, the purpose of the third step is to identify evidence-based interventions and evidence-based guidelines to guide the design of the intervention. MyHealthyPregnancy draws from behavioral decision research methods to offer a personalized medical communication smartphone app. The objective of this tool is to assess and communicate pregnancy risks related to preterm birth. This study focuses on refining and tailoring the MyHealthyPregnancy app for a new population. For this purpose, we conducted interviews with Latino peripartum individuals.
Patient Interviews
We conducted qualitative semi-structured interviews with Latino peripartum individuals to characterize their experience of the domains identified in the literature and clarified by the stakeholder interviews. The semi-structured interview instrument included open-ended questions about the themes identified (e.g., Are you worried about your ability to depend on a support system?), along with additional open-ended questions about the participant's pregnancy experience to identify any other domains or themes (e.g., What are some things that are done differently around pregnancy here in the U.S. in contrast to where you or your family came from?).
Participants
We recruited 15 peripartum individuals for the semi-structured interviews, with 14 completing the entire process (one participant withdrew after enrolling due to a scheduling conflict). All interviewees were recruited using community recruitment strategies that included posting flyers in Federally Qualified Health Centers visited by Latino populations, Facebook groups, and snowball sampling from recruited participants. To participate, individuals had to 1) be 18 years or older, 2) speak Spanish as their primary language, and 3) be pregnant or have given birth within six months of the interview. Interviews were conducted in person or by telephone between June and August 2019. For their participation, interviewees were compensated USD 50 for approximately 1 hour of their time. All participants had access to a smartphone.
Interview procedures
The semi-structured interviews started with open-ended questions formulated to suggest potentially relevant topics but not desired answers. As the interview progressed, questions became increasingly focused on the areas identified by our literature and stakeholder interviews. If necessary, responses were followed up with prompts for clarity (e.g., "How does that work?" "Can you explain what you mean a little more?"), as well as prompts that facilitated discussion of the issues. All the interviews were conducted in Spanish.
Coding
After an initial round of open coding to identify themes by the primary author, complete thematic analysis coding was conducted by a second Spanish-speaking coder. The second coder conducted a deductive analysis, identifying several subcodes within each domain.
4. Design intervention prototype
After analyzing the interviews, a series of app updates were outlined to address the interviewees' healthcare barriers, concerns, and inequities and used it to design a prototype of the Spanish language MyHealthyPregnancy app that addressed the specific concerns for Latino pregnant people in this community.