A nurse at a public clinic in Accra pulls a patient file from a metal cabinet. She is searching for a phone number that might be three years old. The patient has not returned in eight months. The last blood pressure reading on the page was already high. Nothing in the file explains why the patient stopped coming, because nobody asked.
Ghana's National Health Insurance Authority tracked hypertension and diabetes patients at accredited facilities between 2017 and 2019. Thirty-seven percent were lost to follow-up. Within that group, seventy-three percent had attended only one visit before disappearing from the record. In Kenya, a primary care diabetes program reported a thirty-one percent dropout rate among registered patients.
Researchers studying the Ghana cohort found the strongest predictors were not clinical. Patients unaware a follow-up visit was expected dropped out at higher rates than patients with more severe disease. So did patients at facilities with inconsistent physician availability. The gap sits in communication, not medicine.
Reminder systems already work where they exist. A systematic review found SMS appointment reminders cut no-show rates by thirty-eight percent. One pediatric clinic saw its no-show rate fall from 38.1 percent to 23.5 percent after adding them. WhatsApp performs better still. Message open rates run near ninety-eight percent, against roughly eighty-two percent for SMS. Reminders sent through it already qualify as low-cost utility templates under Meta's 2025 pricing tiers.
The tooling gap is what stops this from scaling past pilot studies. Community health workers tracking chronic patients still fall back on paper when phones break or power fails. Data gets re-entered later, by hand. Clinics rely on a phone call, a home visit, or nothing, once a patient misses a date. None of that is a treatment-access problem. It is a tracking problem, running on human memory in a system built for hundreds of patients per nurse.
The build is not a hospital electronic medical record. It is not a global telehealth platform trying to serve every condition at once. Both already exist, and neither has closed this gap. The wedge is narrower - a WhatsApp-native follow-up layer for one chronic condition, in one country's primary care network. It flags a missed visit the day it happens, not the month a chart review finds it.
The clinic in Accra does not need a system that replaces the nurse's judgment. It needs one that remembers what she does not have time to.
