Development of patient information tool for maternal depression
Client: M2B research group
My role: UX researcher
Team: Co-design group (various stakeholders), project coordinator, frontend/backend developer
Background
The M2B* research group at Uppsala University works with local maternity clinics to implement and evaluate early detection and treatment programs for maternal depression.
​
*The name of this research project has been changed as per project requirements.
Challenge
Previous insights identified challenges women face in gathering information regarding mental health during the maternity period and related support. Relevant information is often scattered across various websites, sometimes too brief or too long, sometimes too technical, hard to verify for accuracy, and lacking a patient-perspective.
Kickoff meeting
I met with the project coordinator and the communication representative from clinics.
The aim of the meeting was to establish the main goals for the project, to establish a plan, determine roles and responsibilities, and recruitment strategy.
Goals
1. Ideate and identify informational/educational needs: what info is important
for women based on their experiences? What is already out there? What
information out there lacks?
2. Ideate and design a patient-centered informational resource: what format
do women prefer information in? How to design the resource?
3. Develop: what characteristics and qualities are valued by users? What
additional features do users feel will complement or support information
seeking? How do other tools compare?
4. Validate: What do women think about the UX/UI of the new tool?
Recruitment strategy
Participants for the co-design group included 6 women who were in various stages of the perinatal period, and with varying experiences of depression, as well as healthcare professionals (2 midwives, 1 perinatal psychologist) and a communications representative from the clinics.
Participants for the survey and questionnaires were recruited from the clinics via fliers and staff referrals.
PROCESS
Exploratory Survey
Goal:
Learn about women's preferred formats for receiving information, barriers to information attainment, desirable features of a resource, and satisfaction with existing resources.
​
Participants:
Women in the maternity period (up to 1 year postpartum) attending clinic appointments in all participating clinics were referred to an anonymous online survey via fliers and their midwives. 148 women completed the survey.
Key Outcomes:
-
Apps are most preferred format for an informational resource.
-
Women with low socioeconomic background prefer paper-based resources most, followed by apps.
-
Readability and accessibility is better in digital forms.
-
Websites give ability to search for information tailored to user, and can be translated.
-
Most common barriers: readability, relevance, concerns about accuracy, content length, connectivity, inability to ask questions, language barriers, impractical or unhelpful advice.
-
Most desired features: various lanaguge options (including simple Swedish), accessibility features (dyslexia font, UI adaptations), personalization, chat or forums to ask questions and discuss, symptom tracker, notes, links to publications with info, daily affirmations.
-
Satisfaction: room for improvement in satisfaction related to information accuracy and richness, UI, supporting features.
Participant preferences for information format
Affinity map for key survey outcomes regarding pain points and gaps in current resources
Stakeholder Interviews
Participants: Women were selectively invited to participate from survey respondents (11 women interviewed), and 3 midwives and 1 psychologist were interviewed.
Goals:
-
Gain further insights on women's information-seeking behaviors, their experiences and challenges in doing so.
-
Gain insights regarding the constraints, concerns, and frustrations of from the perspective of professionals involved in maternity care.
Process:
-
Women were asked about their mental health experiences, their efforts to seek information to understand their experience and get support, any frustrations or obstacles they faced while doing so, how they felt their midwife supported them with accessing information outside healthcare appointments. We also asked whether the information women found/recieved was ultimately helpful, what it lacked, and how they felt abou the whole experience.
-
Healthcare professionals were asked about the roles they play in providing informational support, their concerns or opinions about existing information resources, what they think is lacking and can be improved, and what changes would improve efficiency in their work flow.
Personas were created representing interview subjects. We included a stress case, which is especially important in healthcare.
For each persona, a journey map was created to break down their experiences throughout their information seeking process.
Interaction scan
The co-design working group and I performed a rapid interaction scan to identify existing tools and resources, including those mentioned by users in the surveys/interviews, to seek inspiration for information content and associated features. We pulled inspiration from both the healthcare space and comparative industries, and both digital and paper-based sources. The result was a collection of over 100 sources that we used to inspire our ideation sessions.
Feature ideation and prioritization
We completed a series of ideation sessions with our stakeholders (women representing the end group and healthcare professionals) using our survey results, personas, and interaction scan as inspiration.
After coming up with a long list of possible features, we began conducting prioritization activities to narrow the list to the most desirable and feasible features. We also established a list of design principles based on our discussions and insights so far to help guide our UI and content.
MVP prioritization
To decide on the features to integrate into an MVP prototype, I sketched various ideas in collaboration with the research team for how the main features and flows could be developed. Feedback was gathered from end-user representatives to inform further prioritization and design.
Content analysis
Goal: Identify existing topics of information, assess the quality of available information and gaps in content.
​
Process:
-
We compiled an exhaustive list of existing Swedish or Sweden-based information sources related to perinatal mental health, including websites, articles, books, and pamphlets.
-
Quality Assessment: we assessed the accuracy, relevancy, completeness of information, as well as author credentials, references, and alignment with evidence-based guidelines to investigate what needs to be prioritized and what is missing.
-
We extracted all subjects and themes of information from the top 10 high quality content sources.
-
Content topics were put on post-its, and participants were encouraged to introduce new topics based on user insight reflections and personal experiences.
Categorization
Participants sorted all topic cards into categories or themes as part of a guided discussion-based group activity.
An affinity map was created with the final output from all participants to represent the most popular and consistent ideas, and to structure the information.
Content development
Health care professionals, target group representatives, and members of the research team drafted information for all topics, and developed related visuals. It was reviewed continuously for alignment with the established design principles.
4 additional versions of the draft are delivered tailored to users with multiple pregnancies, with common pregnancy complications, with low or no social support, and with a history of mood disorders.
The final draft of text are reviewed by medical experts, as well as language reviewers and editors, for accuracy, actionability, scope, readability, relevance, tone, and cultural relevance.
Prototype
The MVP prototype illustrated the prioritized features, taking into account the insights we gathered from patient research. In particular, we prototyped the reading materials, notes, self-assessment tools, and statistics sections, as well as the signup/login process. Some examples are shown below.
​
We focused especially on keeping the design uncomplicated and uncluttered. One key innovaion within this app-based information guide involved a shift in how we present information regarding support resources for patients.
​
Unlike previous iterations, which categorized resources based on the provider's role within the organization, our approach in this version focused on identifying patient needs and concerns, organizing them as such. This allowed us to pivot from an organization-centered perspective to one centered around the needs of patients.
Validation
We recruited patients from the clinic to use the app for 1 month, whilst giving ecological momentary feedback via feedback surveys sent via text message. A mix of quantitative and qualitative data was collected, and insights for further improvements and developments were highlighted. ​
​
"I have a few apps and the information is always focused on fetal development and pregancy, so it's nice to see information focused primarily on the mother and her wellbeing" ​
"I love that it refers me to a lot of reliable and trustworthy information, since I know that the people behind the app are in healthcare and research, I can trust that I'm not being misled. A lot of times it's just a few words here and there on mental health on other apps and websites, and I feel like I already know all this."
​
"I am happy to see a lot of information available on antidepressant usage during pregnancy and that it goes into the risks and benefits clearly. I also saw some [scientific papers] linked which is good to have. There is just so much conflicting information about it online and you don't know what to believe"
"I like organization of it, it feels easy to read. It feels like all important information is in one place, so you don't have to go from one place to another to read more."
OUTCOME
The new patient information guide app has been implemented for 9 months at the point of writing this report, and is currently available specifically to patients of our collaborating clinics. Reports from both patients and healthcare workers regarding it's impact on empowering women and reducing anxiety in the perinatal period are largely positive.
Usage trends indicate a satisfactory retention rate, with 16% of registered users uninstalling the app at the 30-day timepoint, and 22% uninstalled at 60 days.
​
The key metric for the project team was Monthly Active Users, which made up approximately 80% of registered users. Active users were considered those who registered to the app, and engaged with at least their weekly perinatal updates and at least one reading material.
​
Self-assessment and statistics features were found to have an unexpectedly high engagement rate, with almost 90% of users completing at least one survey a month, and 93% engaging with their statistical charts at least one a month.
Ongoing user experience research will inform subsequent redesigns, and the app will then be evaluated to determine if it is fit for launch for a wider user base.