Across dozens of counties in Kenya, expectant mothers navigating the uncertainties of pregnancy are finding timely guidance and vital emergency support through an innovative text-messaging platform driven by artificial intelligence. What originally began as a routine administrative setup has evolved into an essential digital communication system for thousands of women seeking accessible, reliable health advice.
The initiative traces its roots to an unintentional development within a network of maternal health clinics operating in Kenya. Healthcare organizers had initially established an automated short message service (SMS) framework intended solely to broadcast clinic appointment reminders and scheduling notices to registered patients.
Instead of merely reading the automated appointment notices, expectant mothers began replying directly to the broadcast numbers with their own health inquiries. The sudden influx of inbound messages revealed a substantial, previously unmet demand among patients for continuous, accessible health communication during pregnancy.
The incoming questions spanned a wide spectrum of everyday maternal concerns, ranging from routine lifestyle inquiries to distressing physical symptoms. Expectant mothers frequently wrote in asking about dietary safety, including whether eating avocados was safe during pregnancy, which is completely safe for pregnant women.
Alongside dietary queries, patients frequently sought clarity regarding bodily changes and potentially concerning symptoms. Women regularly sent messages asking about severe headaches, bodily swelling, and other discomforts, attempting to determine whether their experiences were standard parts of gestation or warning signs of serious health complications.
In the early stages of this messaging influx, Jacaranda Health—the healthcare organization behind the clinic initiative—relied on human staff to address the incoming volume. The organization hired clinical nurses whose primary responsibility was to manually review, assess, and answer every individual patient query one text at a time.
However, the volume of incoming texts quickly outgrew the capacity of manual handling. Jay Patel, the director of technology at Jacaranda Health, recalled that during the early period after he joined the organization, the system handled fewer than 100 patient questions each day.
Over the years, public participation expanded rapidly, and the daily message traffic climbed to approximately 15,000 inquiries per day. This dramatic increase in volume made it impossible for a modest nursing staff to maintain individualized manual responses while keeping response times fast and effective.
To address this challenge, software developers began designing an artificial intelligence solution starting in 2020. The resulting platform, known as PROMPTS, was specifically engineered to read, comprehend, categorize, and prioritize the massive stream of inbound health inquiries sent by pregnant women.
Today, the PROMPTS platform is active across public health clinics situated in 24 Kenyan counties. This wide implementation ensures that expectant mothers who visit government-operated clinics can access continuous digital communication and clinical guidance throughout their pregnancy journey.
A critical design advantage of the PROMPTS service is its accessibility for underserved populations with limited technology. Because the platform operates entirely through basic SMS rather than internet-reliant smartphone applications, it is accessible to anyone owning a standard mobile phone, functioning without internet data and at zero financial cost to the user.
This basic technological framework is especially crucial for women living in rural environments who lack reliable internet connections or search engine access when unexpected health crises occur. As Patel pointed out, a mother in a remote village experiencing sudden bleeding in the middle of the night cannot simply search the web for answers, making a free SMS service an essential lifeline.
To safeguard maternal well-being, the PROMPTS platform incorporates a specialized triage algorithm. The artificial intelligence system continuously evaluates inbound text messages and flags approximately 7% of all incoming communications as potentially urgent medical situations requiring immediate attention.
Whenever a message is categorized as an emergency, it completely bypasses the automated AI interaction and moves directly to a specialized help desk queue monitored by medical professionals. On-duty nurses can immediately review the mother's past medical question history, evaluate the emergency, and place a direct telephone call to provide urgent instructions.
Rapid clinical response remains critical in Kenya, where maternal mortality presents a major healthcare challenge. Approximately 6,000 women die every year across the country from pregnancy-related issues and childbirth complications, a statistic that translates to roughly 16 maternal deaths every day.
Public health data indicates that almost one-third of these maternal deaths are directly linked to delays in seeking medical attention. By helping mothers evaluate warning symptoms and eliminating hesitations around seeking emergency clinic care, the text-based service plays a pivotal role in reducing fatal delays and improving maternal outcomes nationwide.
(0)