The Project
In 2023, we partnered with the Kenyan Ministry of Health to deploy telemedicine services in Turkana County—one of Kenya's most underserved regions. The goal was ambitious: bring specialist healthcare to communities with no doctors.
The Context
Turkana County:
Population: 900,000Doctors: 15 (1 per 60,000 people)Nearest hospital with specialists: 500km awayInternet: Limited 3G/4G in towns, none in rural areasElectricity: Solar only in most areasTraditional telemedicine wouldn't work here.
Our Approach
Hub-and-Spoke Model
We established:
1 central hub in Lodwar (county capital) with specialists10 spoke clinics in remote villages staffed by clinical officersMobile units for hard-to-reach communitiesTechnology Stack
Hardware:
Ruggedized tablets for clinicsSolar charging stationsSatellite uplinks for remote locationsBasic diagnostic devices (stethoscopes, BP monitors, glucometers)Software:
Offline-first mobile app for data collectionStore-and-forward for async consultationsVideo calling for synchronous casesAI triage for prioritizationWorkflow
Patient visits local clinicClinical officer conducts assessment using tabletData syncs to hub when connection availableAI triages cases by urgencySpecialist reviews and provides recommendationsClinical officer implements care planFollow-up via SMS remindersChallenges We Faced
Connectivity
Problem: No reliable internet in most locations.
Solution:
Store-and-forward model (doesn't need real-time)Satellite backup for urgent casesData compression and efficient protocolsWeekly sync visits to villages without connectivityPower
Problem: Solar provides limited, inconsistent power.
Solution:
Low-power devicesAggressive battery managementBackup power banksClinical officers trained to manage powerLanguage
Problem: Many patients speak only Turkana or Swahili.
Solution:
All interfaces in SwahiliLocal translators for Turkana speakersVisual/audio explanations for low-literacy patientsTrust
Problem: Communities unfamiliar with technology, suspicious of remote doctors.
Solution:
Community sensitization before launchLocal health workers as ambassadorsStarting with respected community membersTransparent explanations of how it worksClinical Quality
Problem: Telemedicine diagnosis is harder than in-person.
Solution:
Structured assessment protocolsAI-assisted diagnostic supportClear escalation criteriaRegular quality auditsResults After One Year
Quantitative:
15,000 consultations completed2,000 specialist referrals avoided (patients treated locally)Average time to specialist opinion: 24 hours (down from 2 weeks)500 high-risk patients identified and escalatedQualitative:
High patient satisfaction (4.5/5)Clinical officers report increased confidenceSpecialists appreciate efficient case preparationMinistry planning expansion to other countiesKey Learnings
Technology Is the Easy Part
The technical challenges were solvable. The hard parts were:
Building trust in communitiesTraining clinical officersIntegrating with existing health systemsSustaining operations long-termDesign for the Worst Case
We designed for:
No internetNo powerLow literacyNo smartphone experienceThis made the system robust enough for any condition.
Local Partnership Is Essential
We couldn't have done this without:
Ministry of Health buy-inCounty government supportLocal NGO partnersCommunity health workersStart Small, Learn, Scale
We started with 2 clinics, learned intensively, then expanded. Many fixes were discovered only through real deployment.
What's Next
Based on this success, we're:
Expanding to 5 more counties in KenyaAdding AI diagnostic support (image analysis, symptom checking)Developing training programs for clinical officersWorking toward sustainability through government integrationTelemedicine alone won't solve Africa's healthcare challenges. But combined with AI, local capacity building, and system integration, it's a powerful tool for health equity.