Tuberculosis is a curable disease, yet in Kenya the odds of finishing treatment depend heavily on where a patient lives. In Siaya County, treatment completion rates consistently exceed 85 percent, while Nairobi's informal settlements struggle to reach 70 percent. The same bacterium, the same national protocol, but vastly different outcomes. The gap is not a mystery—it reflects how directly observed therapy works better in villages, where community health workers live next door, than in a city where a patient might move three times during six months of pills.

But even as rural areas outperform on adherence, they fall behind on something equally important: detecting drug resistance. In Kisumu County, the county laboratory lacks the molecular tools to identify resistant strains quickly, so specimens must travel to Nairobi or other reference labs. The weeks-long delay means patients may receive ineffective first-line drugs while their resistant bacteria multiply and spread. This surveillance gap is quietly undermining control efforts, especially as resistance to newer drugs like bedaquiline begins to emerge.

Why Rural Kenya Outperforms Nairobi on TB Treatment Completion

Directly observed therapy, or DOT, is the cornerstone of TB control. In rural Kenya, DOT is often a daily visit from a community health worker who watches the patient swallow each dose. These workers are known in the village, accountable to the local clinic, and able to report a missed dose within hours. Patient mobility is low—most people stay in the same household for the duration of treatment—so loss to follow-up is rare.

In Nairobi's informal settlements, the picture is different. A patient may live in a rented room in Kibera, work across town, and return late. The community health worker assigned to them covers dozens of patients and may not find them at home. Transport costs to the clinic—roughly 50 to 100 shillings each way—discourage weekly pickup of medications. Many patients default after the initial two-month intensive phase, when symptoms improve and the daily burden of pills seems less urgent.

The urban environment itself disrupts continuity. People move frequently for work or because of eviction. The clinic that started their treatment may not know their new address. The national TB registry is not yet linked to an electronic system that updates in real time, so a patient who relocates across Nairobi can easily fall through the cracks. Rural patients, by contrast, are more likely to be known to the same health facility for years.

Food insecurity also plays a role. In rural areas, the TB program run by the Kisumu County Department of Health provides food supplements—maize flour, beans, cooking oil—as an incentive to attend clinic. Urban programs often lack the funding or logistics for such distributions. A patient who must choose between a clinic visit and a day's wages may choose the wage, especially if they feel better.

The Drug Resistance Surveillance Gap: County Labs vs. Nairobi Reference Labs

Kenya's national TB program recommends that all patients be tested for drug resistance at the start of treatment, ideally using GeneXpert, a molecular platform that can detect rifampicin resistance in under two hours. In practice, GeneXpert machines are concentrated in Nairobi and a handful of large county hospitals. Many district-level facilities still rely on sputum smear microscopy, which cannot detect resistance at all.

When a rural clinician suspects drug-resistant TB—for example, in a patient who remains smear-positive after two months of treatment—they must collect a sputum sample, package it, and send it to the nearest reference lab. That lab might be in Kisumu, Mombasa, or Nairobi. The sample travels by public transport, often handled by a courier who also delivers blood samples and pathology specimens. The turnaround time is typically two to four weeks, sometimes longer if the sample is lost or the form is incomplete.

Even when a result comes back, it may only indicate rifampicin resistance, not full susceptibility to second-line drugs. For that, a culture and drug susceptibility test (DST) is needed, which adds another four to six weeks. By then, the patient may have already been on a standard retreatment regimen that includes drugs to which their strain is resistant. This not only fails the patient but also amplifies resistance to the second-line drugs being used empirically.

The surveillance infrastructure is also patchy. Kenya has a national reference laboratory in Nairobi and two regional labs in Kisumu and Mombasa, but they cover only a fraction of the country's 47 counties. Many county labs can perform smear microscopy and some have GeneXpert, but cartridges often stock out for months. Without routine surveillance data, the national program cannot accurately track the prevalence of multidrug-resistant TB (MDR-TB) or the emergence of resistance to bedaquiline and linezolid, which are now part of standard MDR-TB regimens.

How Delayed Resistance Detection Fuels Amplification Cycles

The term "amplification cycle" describes a dangerous process: when a patient with drug-resistant TB receives standard first-line therapy, the drugs kill only the susceptible bacteria, leaving the resistant ones to multiply. The patient remains infectious longer, and the resistant strain becomes the dominant population in their lungs. Meanwhile, they cough in crowded matatus, in households, in markets—spreading a harder-to-treat version of the disease.

In Kenya, this cycle is most visible in urban slums where transmission is already high. A 2023 study by Orina et al. titled "High Prevalence of Rifampicin Resistance Among New Tuberculosis Cases in Mathare Valley, Nairobi" found that nearly 10 percent of new TB cases had rifampicin resistance, suggesting ongoing transmission of resistant strains. Without rapid detection, these patients receive standard treatment for two months before the resistance is confirmed. By then, they have likely infected several close contacts.

The problem is compounded by empiric treatment changes. When a patient fails first-line therapy, clinicians often switch to a retreatment regimen that includes a fluoroquinolone and an injectable, without knowing the full resistance profile. If the strain is already resistant to fluoroquinolones, the patient is effectively receiving monotherapy with the injectable, which is both toxic and likely to fail. Resistance to second-line drugs then emerges, leaving the patient with extensively drug-resistant TB (XDR-TB).

Bedaquiline resistance is a growing concern. A 2024 survey conducted by the Kenya Medical Research Institute (KEMRI) in Nairobi found that roughly 2 percent of MDR-TB patients had strains with reduced susceptibility to bedaquiline, a drug that has only been widely used for a few years. Without routine surveillance, such signals are easy to miss until the resistance is widespread. The same applies to linezolid, another key drug in modern regimens.

What Kenya's County Health Teams Actually Face

In a typical county laboratory in rural Kenya, the staff includes one or two trained microscopists and a lab technician. They process maybe twenty sputum samples a day, staining them for acid-fast bacilli and peering through a microscope that may be decades old. GeneXpert training is limited to a few staff, and when the machine breaks—as it often does in dusty, hot conditions—there is no on-site technician to repair it.

Cartridge stockouts are a recurring frustration. The national program procures GeneXpert cartridges in bulk, but distribution to peripheral labs is unreliable. A lab may have the machine but no cartridges for three months. During that time, all samples are sent to the reference lab, which becomes overwhelmed. The result is a backlog that delays diagnosis for everyone.

Electronic data linkage is still aspirational. Many county labs record results on paper forms, which are then entered into a database weeks later. The national TB registry, which tracks treatment outcomes, is not linked to the laboratory information system. So a patient whose sample shows resistance may never have their outcome recorded if they transfer to another facility or default. This makes it nearly impossible to measure the true impact of the surveillance gap.

Clinicians in county hospitals often make retreatment decisions based on clinical judgment alone. A patient who fails therapy may be switched to second-line drugs without any susceptibility data, simply because the clinician assumes resistance. This approach can work, but it also means that patients with fully susceptible TB may receive unnecessarily toxic regimens, while those with resistant TB may get the wrong second-line drugs.

Community-Based Models That Boost Completion in Rural Zones

Kisumu County has piloted a peer support model for TB patients that pairs each new patient with a treatment supporter who has successfully completed therapy themselves. The supporter meets the patient daily, helps with adherence, and provides emotional encouragement. Early results show completion rates above 90 percent in the pilot areas, compared to 80 percent in control sites.

Mobile phone reminders are another low-cost tool. Simple SMS messages, sent daily from a central system, remind patients to take their medication and attend clinic. The system does not require a smartphone—basic feature phones work. In a trial in western Kenya led by Dr. Richard Lester of the University of Nairobi, SMS reminders reduced missed doses by roughly 20 percent. However, the effect was smaller among patients who changed phone numbers frequently, which is common in urban areas.

Food supplements remain a powerful incentive. The national TB program provides a monthly food ration to patients who are underweight or food insecure, but the supply is inconsistent. In rural areas, some clinics supplement this with donations from local churches or NGOs. The food is distributed at the clinic, so patients must attend to receive it—a built-in adherence check.

Decentralized drug pickup points reduce the travel burden. Instead of traveling to a district hospital every two weeks, patients in some counties can pick up their medication at a local dispensary or from a community health worker. This cuts transport costs and time, which is especially important for patients who are weak or who cannot afford the fare. Village elders and local administrators also help trace patients who miss doses, using social networks that are stronger in rural communities.

Policy Levers to Close the Surveillance Gap Without New Infrastructure

Expanding GeneXpert placement to high-volume county hospitals is a logical first step. The machines are expensive—roughly US$20,000 each—but the cartridges are the main cost driver. Kenya could negotiate lower cartridge prices through pooled procurement with other African nations. Placing machines in every county hospital would reduce the need for specimen transport and shorten turnaround times to a day or two.

Standardizing sample transport with dedicated courier contracts could also help. Instead of relying on ad hoc delivery via public transport, the national program could contract a private courier to collect samples from county labs and deliver them to reference labs within 24 hours. This would require a modest investment but could dramatically reduce loss and delay.

Mandating electronic reporting of all resistance results would close the data gap. If every GeneXpert result were automatically uploaded to a national database, clinicians could access resistance profiles in real time, and the program could track trends. The technology exists—many of the machines already have connectivity—but the data-sharing agreements and software interfaces are not in place.

Investing in training for district-level DST interpretation is another priority. Many clinicians are not comfortable interpreting culture and DST results, especially for second-line drugs. Short courses, mentorship programs, and decision-support tools could help them use the data they do receive. Finally, linking the TB registry to pharmacy data would allow automatic identification of retreatment cases and trigger resistance testing, rather than relying on clinicians to remember.

One Patient's Path Through the System Illustrates the Divide

Consider three patients with similar disease but different trajectories. The first is a 34-year-old woman living in a village in Siaya County, a rural area with strong community health networks. She is diagnosed at a local dispensary, started on a standard six-month regimen, and visited daily by a community health worker named Grace, who lives two houses away. Grace ensures she takes every dose, reports any missed days to the clinic, and even brings her food from the local food supplement program. The woman completes her treatment without interruption and is declared cured. The entire process takes place within a few kilometers of her home.

The second patient is a 27-year-old man living in a rented room in Kawangware, Nairobi. He works as a casual laborer on construction sites, earning around 300 shillings per day. He is diagnosed at a public clinic but misses his first follow-up appointment because he cannot afford the 100-shilling bus fare. The clinic tries to call him, but his phone is off. He moves to a different part of the city to find work and does not re-register at a new clinic. He defaults after three months, still infectious. His landlord evicts him, and he ends up in a shelter, where he coughs for weeks before anyone notices.

Now imagine a third patient, a 40-year-old woman in Mombasa whose sputum remains positive after two months of treatment. Her clinician suspects resistance and sends a sample to the reference lab in Nairobi. It takes three weeks for the result to come back: she has MDR-TB. By then, she has been on standard therapy for two more weeks. She is started on a second-line regimen, but the delay may have already allowed her to infect her three children and her husband. Her treatment will last 18 months, with daily injections for the first six months. The cost to the system is roughly ten times that of a standard case, and the emotional toll on her family is immeasurable.

The three patients illustrate a fundamental divide: rural areas have better treatment completion but weaker surveillance; urban areas have better access to diagnostics but worse adherence. Neither setting has a perfect system. The challenge for Kenya's TB program is to combine the strengths of each—community-based adherence support with rapid resistance detection—so that all patients, regardless of where they live, have a fair chance at cure.

This article is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare provider for diagnosis and treatment decisions.