In the humid tropics of Makeni, Sierra Leone, the AISHA team huddled together in a secluded space of their fieldwork office at World Hope International. The discussions were lively, filled with overlapping conversations surrounding Wi-Fi, qualitative data collection, budgets, and contract agreements for translators. For the last eight months, the AI Strengthening Healthcare Access (AISHA) team has been diligently working on programming Amazon Alexa devices with the hope of using them as health education tools in rural peripheral health units (PHUs). To prepare for such a monumental task, the team has two working subgroups to ensure a smooth implementation: skill developers and use case researchers. The skill development team creates and writes code in Amazon’s Developer Console, to bring the use cases identified by researchers to fruition. However, all teammates have the knowledge to create Alexa skills.
With the fresh use cases identified and (almost) ready to demo, the team’s PI Khanjan Mehta, told the team their first PHU visit would be that afternoon. While the research team was focused on creating the budgets, a protocol for visiting PHUs, or the limitations of qualitative data collection sans-IRB, the skill developers were heard asking “What is preeclampsia?” or “How do I know if I have breast cancer?” from across the room, increasingly anxious with the upcoming visit soon approaching. With a half hour left and all contracts and protocols done, the entire team was hunched over the Alexas and code trying to debug the Question and Answer skill. As the clock turned 2:10, Professor Mehta informed the team that the meeting will be at 4:30 instead. A collective sigh of relief and a wave of hope was felt by the team, as they realized they’d be able to finish debugging the skill.
On the short ride to Makump Bana Community Health Post (CHP), a confused laugh was shared by the team as they followed Professor Mehta out of the land cruiser preemptively at a pit stop. When the team arrived to the CHP, they were greeted by the primary nurse at the unit, Mabinty. After an exchange of warm greetings, the team got to work. Splitting into three primary roles, the team set to work. Sofía was the primary communicator and spoke to Mabinty about her experiences working at the CHP. Acting as observational documentarians, Rachel, Elif, Aabi, and Eric were interacting taking pictures of reference manuals, treatment plans, as well as interacting with community members who gave them a tour of the nearby flora. As the technology implementers, Omar and Zavier were using the power outlet in Mabinty’s room to set up the Alexa devices. It was finally time to put the device to the test.
Using Sofía’s phone as a hotspot, Zavier guided Mabinty through the process of using the AISHA answers skill, which is linked to a spreadsheet of over one thousand questions relating to various topics including but not limited to STIs, Contraceptives, HPV, and Women’s health. Zavier prompted Mabinty to ask the device a question about cervical cancer:
Zavier: “Open AISHA answers”
Alexa: “Welcome to AISHA Answers. I can answer questions from numerous health and disease topics. Selecting a topic will help me give you the best answer. Would you like to hear the topics, select a topic, or ask a question? If you need help, say help.”
Z: “I want to ask a question.”
A: “I’m here to help. What’s your question?”
Z: “Go ahead and ask a question.”
M: “What is the leading cause of death for women in Sierra Leone?”
A: “Maternal mortality is a pressing concern, encompassing direct and indirect causes. Direct deaths result from complications during pregnancy, labor, or postpartum, while indirect deaths stem from pregnancy-related complications affecting longer-term health. Direct obstetric deaths are tied to pregnancy stages, while indirect ones arise from worsened pre-existing conditions due to pregnancy changes. Fatal complications like hemorrhaging, hypertension, preeclampsia, and infections contribute significantly. This classification guides interventions for improved maternal health outcomes.”
After a few rounds of questions and interactions, the hotspot connection began to dwindle, and the group came together again outside to ask how the demo went. Mabinty was expressing hesitancy at the proposed technology and the team speculated that she was unlikely to use it.
Team Highlights
Elif → Learning how to troubleshoot seemingly daunting problems in a time crunch.
Omar → It was great to see the device working in an environment of low connectivity after being unsure it’d work or not.
Zavier → Successfully debugging the code after hours of frustration and being able to see it work in the PHU.
Rachel → Writing a formal contract and learning document design shortcuts from Eric, as well as being in a PHU after researching it for so long.
Aabi → Interacting with community members and being shown the mimosa flower.
Sofía → Connecting with Mabinti and learning about her experiences working in the CHP.
Eric → Writing a contract for the first time.
Average rating of the team’s day: 3.14/5