HIV infections fall in lakeside counties but rise in arid north
The national HIV prevalence stands at three per cent, with an estimated 1.3 million people living with HIV.
What you need to know:
- Siaya, Kisumu and neighbouring lakeside counties have reduced infections through sustained investment, targeted prevention programmes and strong community engagement.
- In contrast, many arid and semi‑arid counties have moved in the opposite direction, with limited health infrastructure, chronic staff shortages, and low testing coverage allowing transmission to continue largely unnoticed.
Samburu, Mandera and Wajir counties are recording rising HIV infections, placing Kenya's arid and semi-arid (ASAL) regions at the centre of a growing geographic divide within the epidemic, even as lakeside counties such as Siaya and Kisumu have spent years bringing their caseloads under control.
The Kenya HIV Estimates 2024, cited in the Kenya AIDS Integration Strategic Framework (KAISF) 2025–2030, reveal that new infections are emerging in regions that have historically reported low rates of transmission. The national HIV prevalence stands at three per cent, with an estimated 1.3 million people living with HIV.
However, these national figures mask widening regional disparities.
The National Syndemic Diseases Control Council (NSDCC), which released the framework, states that the epidemic is evolving and that new infections are no longer exclusively concentrated in the traditionally high burden counties surrounding Lake Victoria and major urban centres.
Siaya, Kisumu and neighbouring lakeside counties have reduced infections through sustained investment, targeted prevention programmes and strong community engagement. In contrast, many arid and semi arid counties have moved in the opposite direction, with limited health infrastructure, chronic staff shortages, and low testing coverage allowing transmission to continue largely unnoticed.
“While counties such as Siaya and Kisumu recorded significant reductions in new HIV infections, others, particularly in ASAL and marginalised regions such as Mandera, Wajir and Samburu, saw reversals. This requires county specific strategies rather than uniform national approaches,” said the NSDCC in the framework.
The framework identifies population movement as a key driver. Pastoralist communities regularly cross county and international borders in search of pasture and water, while major transport routes connect local populations to networks extending into Ethiopia, Somalia and South Sudan.
These movements create transmission patterns that fixed health facilities struggle to address. Testing and treatment models designed for urban and peri urban populations often fail to reach communities that may travel hundreds of kilometres between visits to health facilities.
“The challenge is exacerbated by limited county resources, with some counties allocating over half of their health budgets to salaries, leaving insufficient funds for outreach, medication, supplies and community programmes,” the framework stated.
A review of the previous Kenya AIDS Strategic Framework concluded that geographic inequality was one of the biggest weaknesses in the country's HIV response, finding that health authorities had not done enough to address the different infection patterns emerging across counties.
The review recommended that arid and semi arid counties receive interventions designed around their unique circumstances, rather than scaled down versions of national programmes. It also highlighted weak digital health systems, poor information technology infrastructure, and limited data sharing, all of which made it harder to track infections and inform decision making.
To address these gaps, the new framework requires all 47 counties to develop county operational plans based on local HIV data. These plans will be overseen by County
Integration Steering Committees, chaired by County Directors of Health, with the aim of bringing planning, funding decisions and programme management closer to the communities where infections are occurring.
