Increasing Accessibility of Maternal Health Information for Kenyan Mothers
Project Summary
PROBLEM
Expectant mothers in Kenya currently underutilize a text-based application that provides them with pregnancy health information, even though using this resource has been shown to result in healthier outcomes.
GOAL
Design a  system using telecommunication technology that will increase engagement with maternal health information.
RESULT
My research found the lack of engagement using the text-based application was due to  negative perceptions of the Kenyan maternal healthcare system. This gave rise to a lack of trust because of limited access, privacy concerns, and lack of respect from care providers. I addressed this medical hesitancy by designing a voice user interface service that featured birth stories, frequently asked questions, and regular nurse-patient condition updates.
ROLE
Lead Experience Designer, User Researcher, Communications,
Graphic Designer
DELIVERABLES
Research Report, Design Proposal, Project Poster.
TIMELINE
January 2022 — June 2022
TEAM
Microsoft Global Health Access, Microsoft Africa Research Institute, Amen Aiwekhoe, Adoniah Carmeline, Drew Nevins
Process
+
Initial Research
We started this project with no access to users, the existing text-based application, or any user data. Also, we operated under the assumption that the existing application worked perfectly and any issues with it were cultural. In addition, we used secondary research methods, as well as subject matter expert interviews and academic literature, to conduct our research.
"Preventative care is not a concept believed to be important. On less than ten dollars per day and barely scraping by, preventative care is only if it affects your life. It costs opportunity and travel to care." -Microsoft Health
Three Delay Model
The three delay model shows that the goal of the application is to persuade the mother to seek medical care.
Patient Activation Measure
The patient activation measure is a measure of engagement and consists of 4 elements as shown below. In practice, text messaging has not allowed expectant mothers to habitually use it as a form of healthcare information. Also, mothers in Kenya do not have confidence in  the application because they do not primarily rely on using skilled birth attendants, which is what the application advocates.
Research Synthesis -Journey Mapping
Journey mapping was used to combine the different research findings from across the team. This method emphasized that Kenyan women culturally do not reveal much about their pregnancy, traditional birth attendants are a preferred homeopathic alternative, and delivery sites are often different facilities from their antenatal care check-ups.
Ideation - Storyboarding
Giving the patient a voice was integral to the concept of patient-centered care and was a core aspect of increasing one's patient activation measure. I created the following storyboards as an idea of how the system should be used.
Ideation - User Flow
Browsing behaviors were a perceived benefit of a voice user system compared to its text-based counterpart. Therefore, features were conceived to allow the mother to interact with the system even when they have no questions in mind.

Eric: With browsing behaviors in mind, as well as the contrast of a voice user system to its text-based counterpart, we developed more accessible features to allow the mother to interact with the system even when they had no specific health-based questions in mind. This, in turn, would allow them to learn more about healthier lifestyle practices before and after pregnancy.
Prototyping - Role playing Conversation
To understand the conversational user interface, I asked my team to role-play what it would be like to be a mother who uses the app and the voice response. This method revealed questions such as: how do we know if the mom is picking up the phone, how long until the mom gets a response, and how should the mom receive an answer to a previously asked question.
Ideation & Prototyping - Co-Creation Session
I moderated a remote, co-design session with Kenyan mothers to identify how they approach pregnancy related information and the viability of voice user interface. The responses are blurred for privacy reasons..
"Is it a human or a robot? I would like a robot so other people don't hear about my problems."‍-Kenyan Mother
Prototyping -
Voiceflow
I built a chatbot in Voiceflow but was considered ineffective without using the medical content currently used by the text-based application.
Final Deliverables
Primary User Persona
A persona was created from interviewing people who work in public health in Sub-Saharan Africa and a literature review of health seeking behaviors among Kenyan women. I used the persona to synthesize research findings and understand the needs of the Kenyan women. It summarizes patterns that we viewed through our research methods.

Many mothers had a negative perception of public health clinics, despite maternal health care being a free service in Kenya. There are prevalent rumors of lack of privacy, overcrowding, and insensitive providers. Additionally, it is often difficult for mothers to share their own symptoms during pregnancy as there is a slight risk of being blamed for any issues by their community. It's important to note that mothers are very passionate about their baby's health. But many believe that other resources, such as homeopaths, are more reliable and convenient resources.
The Voice User Interface
Text-based platforms cannot be a stand alone solution as they lack engagement and privacy (especially in a social context where many families share phones). However, the biggest threat to a text-based system that promotes maternal health occurs when a mom believes that the she doesn't want to go to the medical provider.  There are no safeguards to let the mom know that anti-hospital sentiments are valid while emphasizing that despite rumors, delivering with a skilled birth attendant will decrease risk of maternal mortality by 50%.
‍
I designed the integration of a voice user interface to transmit pregnancy-information while targeting medical hesitancy. Giving the system a voice allows the mother to build trust with formal medical care. Additional features help in making the health information more actionable and provide opportunities to explore information.
Ecosystem Map
The ecosystem map describes how to build the interactive voice response system, the goals of each function, and how it interacts with the existing text-based texting system. The implementation is left broad to allow for flexibility with the existing system.
Service Blueprint
Despite, maternal health care being a free service in Kenya, it is a widely underutilized resource. Interviews with stakeholders and academic research  revealed a negative connotation associated with the public health clinics. I designed the integration of a voice user interface to help transmit maternal health information for Kenyan mothers.
References‍

Isaac Amankwaa, Daniel Boateng , Dan Yedu Quansah, Cynthia Pomaa Akuoko, and Catrin Evans.2018. Effectiveness of short message services and voice call interventions for antiretroviral therapy adherence and other outcomes: A systematic review and me-ta-analysis. PLOS ONE, 13, 9 (Sep. 2018). DOI: https://doi.org/10.1371/journal.pone.0204091

Atik Ahsan and Ananya Raihan. 2013. Under-standing mHealth impact among Aponjon (MAMABangladesh) subscribers through a phone survey inBangladesh. In Proceedings of the Sixth InternationalConference on Information and Communications Technologies and Development: Notes - Volume 2 (ICTD ‘13).Association for Computing Machinery, New York, NY,USA, 1–4. https://doi.org/10.1145/2517899.2517920

Vanessa Boudewyns, Itai Himelboim, Derek L.Hansen, and Brian G. Southwell. 2015. Stigma’s Effect on Social Interaction and Social Media Activity. Journal of Health Communication, 20, 11 (June 2015) 1337- 1345.DOI: https://doi.org/10.1080/10810730.2015.1018604

Sherri Bucher, Olive Konana, Edward Liechty, Ana Garces, Peter Gisore, Irene Marete, Constance Tenge, Evelyn Shipala, Linda Wright, and Fabian Esamai.(2016). Self-reported practices among traditional birth attendants surveyed in western Kenya: a descriptive study. BMC Pregnancy and Childbirth, 16, 219 (Aug.2016). DOI: https://doi.org/10.1186/s12884-016-1007-8

Jessica Cohen, Ginger Golub, Margaret E. Kruk, and Margaret McConnell. 2016. Do active patients seek higher quality prenatal care?: A panel data analysis from Nairobi, Kenya. Prev Med., 92 (Nov. 2016) 74 - 81.DOI: https://doi.org/10.1016/j.ypmed.2016.09.014

Jessica Crawford, Erin Larsen-Cooper, ZachariahJezman, Stacey C Cunningham, and Emily Bancroft.2014. SMS versus voice messaging to deliver MNCHcommunication in rural Malawi: assessment of delivery success and user experience. Global Health: Science and Practice, 2, 1 (Feb. 2014) 35-46. DOI: https://doi.org/10.9745/GHSP-D-13-00155

Amit Das, Preethi Jyothi, and Mark Hasega-wa-Johnson. 2016. Automatic Speech RecognitionUsing Probabilistic Transcriptions in Swahili, Amharic,and Dinka. Proceedings of the Annual Conference of the International Speech Communication Association(INTERSPEECH ‘16), Sep 8 - 16, 2016. San Francisco,California, United States, 3524 - 3528. https://doi.org/10.21437/Interspeech.2016-657

Valentina Actis Danna, Carol Bedwell, SabinaWakasiaka, and Tina Lavender. 2020. Utility of the three-delays model and its potential for supporting a solution-based approach to accessing intrapartum care in low- and middle-income countries. A qualitative evidence synthesis. Global Health Action, 13, 1 (Oct. 2020).DOI: https://doi.org/10.1080/16549716.2020.1819052

Mary L Greaney, Elaine Puleo, Kim Sprunck-Harrild, Gary G Bennett, Moira A Cunningham, MatthewW Gillman, Molly Coeling, and Karen M Emmons. 2012.Electronic reminders for cancer prevention: factors associated with preference for automated voice reminders or text messages. Prev Med., 55, 2 (Aug. 2012) 151 - 154.DOI: https://doi.org/10.1016/j.ypmed.2012.05.014

Lucy Gilson. 2005. Editorial: building trust and value in health systems in low- and middle-income countries. Social Science & Medicine, 61, 7 (Oct. 2005) 1381 -1384. DOI: https://doi.org/10.1016/j.socscimed.2004.11.059

Judith H. Hibbard, Jean Stockard, Eldon R.Mahoney, and Martin Tusler. 2004. Development
of the Patient Activation Measure (PAM): Conceptualizing and Measuring Activation in Patients andConsumers. Health Services Research, 39, 4p1 (June2004) 1005 - 1026. DOI: https://doi.org/10.1111/j.1475-6773.2004.00269.x

HygieneHub. 2021. Using the Leap mHealth plat-form remotely to train Community Health Volunteers about COVID-19. Retrieved from https://www.hygiene-hub.info/en/case-studies/kenya-using-leap-mhealth-platform-remotely-to-train-community-health-volunteers-about-covid-19

Nicole L. Johnson, Susanna F. Scott, and Maria Brann. 2020. “Our Birth Experiences are What BindsUs”: Women’s Motivations for Storytelling about Birth to Build Mother wisdom. Communication Studies, 71, 4(June 2020) 649 - 668. DOI: https://doi.org/10.1080/10510974.2020.1771391

Michael W. Kraus. 2017. Voice-only communication enhances empathic accuracy. American Psychologist, 72, 7 (Oct. 2017) 644 - 654. DOI: https://doi.org/10.1037/amp0000147

Evaline Lang’at, Lilian Mwanri, and Marleen Temmerman. 2019. Effects of implementing free maternity service policy in Kenya: an interrupted time series analysis. BMC Health Services Research, 19, 645 (Sep. 2019). DOI: https://doi.org/10.1186/s12913-019-4462-x

Kenneth L. Leonard. 2014. Active patients in rural African health care: implications for research and policy. Health Policy and Planning, 29,1 (Jan. 2014) 85 - 95. DOI: https://doi.org/10.1093/heapol/czs137

Rodgers O. Moindi, Moses M. Ngari, Venny C. S. Nyambati, and Charles Mbakaya. 2016. Why mothers still deliver at home: understanding factors associated with home deliveries and cultural practices in rural coastal Kenya, a cross-section study. BMC Public Health, 16, 114 (Feb. 2016). DOI: https://doi.org/10.1186/s12889-016-2780-z

Nirmala Murthy, Subhashini Chandrasekharan, Muthu Perumal Prakash, Aakash Ganju, Joanne Peter, Nadi Kaonga, and Patricia Mechael. 2020. Effects of an mHealth voice message service (mMitra) on maternal health knowledge and practices of low-income women in India: findings from a pseudo-randomized controlled trial. BMC Public Health, 20, 820 (June 2020). DOI: https://doi.org/10.1186/s12889-020-08965-2

Kenneth Ngure, Susan B. Trinidad, Kristin Belma-Sofie, John Kinuthia, Daniel Matemo, Grace Kimemia, Anne Njoroge, Lillian Achiro, Jillian Pintye, Nelly R. Mugo, Elizabeth A. Bukusi, Jared M. Baeten, Renee Heffron, Grace John-Stewart, Maureen C. Kelley. 2021. Perceived Social Influences on Women’s Decisions to use Medications not studied in Pregnancy. A qualitative Ethical Analysis of Preexposure Prophylaxis Implementation Research in Kenya. Journal of Empirical Research on Human Research Ethics, 16, 3 (July 2021) 225 -237. DOI: https://doi.org/10.1177/15562646211012296

Osamuedeme Odiase, Beryl Akinyi, Joyceline Kinyua, and Patience Afulani. 2021. Community Perceptions of Person-Centered Maternity Care in Migori County, Kenya. Front. Glob. Womens Health, 8 (Oct. 2021). DOI: https://doi.org/10.3389/fgwh.2021.668405

Michelle Reissig, Cynthia Fair, Bethany Houpt, and Vicki Latham. 2021. An Exploratory Study of the Role of Birth Stories in Shaping Expectations for Childbirth among Nulliparous Black Women: Everybody is Different (but I’m Scared). Journal of Midwifery & Women’s Health, 66, 5 (Oct. 2021) 597 - 603. DOI: https://doi.org/10.1111/jmwh.13282

Roselyter Monchari Riang’a, Anne Kisaka Nangulu, and Jacqueline E. W. Broerse. 2018. Perceived cause of adverse pregnancy outcomes and remedies adopted by Kalenjin women in rural Kenya. BMC Pregnancy and Childbirth, 18, 408 (Oct. 2018). DOI: https://doi.org/10.1186/s12884-018-2041-5

Casper S. Shikali and Refuoe Mokhosi. 2020. Enhancing African low-resource languages: Swahili data for language modelling. Data in Brief, 31 (Aug. 2020). DOI: https://doi.org/10.1016/j.dib.2020.105951.

Haohan Wang, Agha Ali Raza, Yibin Lin, and Roni Rosenfeld. 2013. Behavior analysis of low-literate users of a viral speech-based telephone service. In Proceedings of the 4th Annual Symposium on Computing for Development (ACM DEV-4 ‘13). Association for Computing Machinery, New York, NY, USA, Article 12, 1–9. https://doi.org/10.1145/2537052.2537062