Text Messaging AI Helps Pregnant Women in Kenya Access Care

In rural parts of Kenya, a simple text message can now serve as a lifeline for pregnant women facing medical uncertainty. What began as a basic system for sending appointment reminders has grown into a sophisticated health service that handles thousands of daily inquiries. This shift addresses a significant gap in healthcare access for women who cannot easily travel to hospitals or rely on modern smartphones.

The service, known as PROMPTS, is operated by Jacaranda Health and serves patients in 24 counties across the country. It allows users to ask questions about symptoms, nutrition, or general well-being using standard SMS technology. For many, this is the fastest and most accessible way to determine if immediate medical attention is required, removing the barriers of cost and connectivity that often delay critical care.

AI triages urgent symptoms efficiently

The system uses artificial intelligence to read and understand patient messages, prioritizing those that indicate potential danger. Approximately seven percent of incoming texts are flagged as urgent by the algorithm. These messages bypass automated responses entirely and are routed directly to a human nurse who can review the patient’s history and call them immediately. This rapid diversion to human care is designed to catch serious conditions like pre-eclampsia or complications in newborns before they become life-threatening.

According to reports from HENNET, a coalition of Kenyan healthcare providers, this direct line to professionals helps prevent the delays in seeking care that contribute to a significant portion of maternal deaths. By providing a reliable referral mechanism, the service empowers women to make informed decisions about when to seek hospital care. Health experts note that this approach centers the conversation on the patient's specific concerns rather than broadcasting generic public health advice.

Local language support improves reach

A key feature of the platform is its ability to understand and respond in multiple local dialects. The system is trained to process messages in English, Swahili, and Sheng, a popular urban dialect in Nairobi that blends both languages. Jacaranda Health has also expanded the service to neighboring countries, including Ghana, Nigeria, and Tanzania, where versions are available in local languages such as Twi and Hausa. This linguistic sensitivity ensures that the advice provided is culturally and contextually relevant to the user's specific environment.

Access remains limited by connectivity

While the technology offers a valuable safety net, it is not without limitations. The service relies on basic mobile network coverage, meaning women in areas with poor signal strength may still struggle to connect. Additionally, the system is designed to triage and refer, not to provide full medical treatment or diagnosis. It functions as a bridge to professional care rather than a replacement for it. As reported by GN technics/ai (en-US), the tool demonstrates the potential of AI in global health, but its effectiveness remains dependent on the existing infrastructure and the willingness of users to trust automated systems with sensitive health data.