Evaluating tuberculosis treatment outcomes and predictors in five Southern African countries: A multi-country cohort analysis

This multi-country cohort analysis of 1,438 tuberculosis patients across five Southern African sites reveals that while 80% achieved successful treatment outcomes, the overall 15% rate of unsuccessful outcomes (death, loss to follow-up, or failure) exceeds WHO targets and is significantly predicted by multidrug-resistant TB, lack of formal education, and low BMI, highlighting an urgent need for improved retention and management strategies.

The Gist

Imagine the fight against Tuberculosis (TB) as a long, grueling marathon. For many runners in Southern Africa, the finish line is too far away, and too many people are dropping out along the way. This paper is like a race report from five different countries (Malawi, Zambia, Mozambique, Zimbabwe, and South Africa) that looks at who finished the race, who didn't, and why.

Here is the story of that race, broken down simply:

The Big Picture: The Race is Harder Than We Hoped

The World Health Organization (WHO) set a goal: 90% of runners should finish the TB treatment marathon successfully. In this study, only 80% made it. While that sounds like a passing grade, in the world of life-or-death diseases, it's a failing one.

The "unsuccessful" runners fell into two main categories:

1. The Dropouts (Loss to Follow-up): About 9% of people stopped showing up to the clinic. They didn't necessarily quit on purpose; they just got lost in the shuffle, moved away, or couldn't keep up with the schedule.
2. The Tragedies (Deaths): About 6% of the runners passed away before finishing.

Who Was Most Likely to Drop Out?

The researchers looked for "potholes" in the road that made the race harder. They found two major obstacles:

• The "Super-Bug" Trap (MDR-TB): Some runners were fighting a tougher version of TB called Multidrug-Resistant TB (MDR-TB). Think of this as running with a heavy backpack full of rocks while everyone else runs with a light backpack. These runners were more than twice as likely to drop out. The treatment is longer, harder to tolerate, and the side effects are worse, making it easy to give up.
• The Education Gap: People with no formal education were much more likely to drop out or die compared to those who had finished high school or university.
  • The Analogy: Imagine the treatment instructions are written in a complex language. If you have a high school diploma, you can read the map and know which turn to take. If you can't read the map, you might wander off the path and get lost. Education acts as a "GPS" for navigating the complex healthcare system.

The Surprising "Helpers"

The study also found some things that acted like "boosters" for the runners:

• The "Proof" Boost (Bacteriologically Confirmed TB): People who had a lab test that definitely proved they had TB were less likely to die than those who were just "suspected" of having it based on symptoms.
  • Why? It's like having a confirmed diagnosis of a flat tire. You know exactly what's wrong, so you fix it immediately. If you just "think" you have a flat tire, you might keep driving until the wheel falls off. Knowing for sure helps doctors treat the patient faster and more accurately.
• The "Weight" Factor: People with a healthy Body Mass Index (BMI) were less likely to die. Being underweight is like running a marathon with an empty fuel tank; your body doesn't have the energy to fight the infection.
• The Smoking Paradox: This was the weirdest finding. People who smoked a moderate amount were actually less likely to die than non-smokers.
  • The Explanation: The authors suspect this isn't because smoking is good (it's not!). Instead, smokers often have other health issues (like coughs or breathing problems) that make them visit the doctor more often. This extra attention might have meant their TB was caught earlier or they were monitored more closely. It's like a car that makes a weird noise and gets checked by a mechanic every week, whereas a quiet car might break down completely before anyone notices.

The Takeaway: How to Fix the Race

The authors conclude that to get more people to the finish line, we need to change the strategy:

1. Help the "Super-Bug" Runners: We need better support systems for people with MDR-TB. They need more hand-holding, reminders, and maybe shorter, less toxic treatments so they don't get overwhelmed and quit.
2. Hand Out More Maps: We need to improve health education, especially for those with less formal schooling. If people understand why they need to take their medicine every day, they are less likely to drop out.
3. Get Better at Diagnosis: We need to find more ways to confirm TB with lab tests (like the "Proof" boost) so treatment starts immediately and correctly.

In short: The race to cure TB in Southern Africa is currently running behind schedule. To win, we need to clear the potholes (MDR-TB and lack of education) and give every runner a better map and a lighter load.

Technical Summary

1. Problem Statement

Despite global efforts to end tuberculosis (TB), treatment outcomes in high-burden settings, particularly in Southern Africa, remain suboptimal. While global success rates for first-line TB treatment have improved, they still fall short of the World Health Organization (WHO) End TB Strategy targets (≥90% success rate by 2035). Unsuccessful outcomes—defined as death, loss to follow-up (LTFU), and treatment failure—are driven by complex factors including HIV coinfection, drug-resistant TB (MDR-TB), socioeconomic status, and diagnostic delays. There is a critical need to identify specific predictors of poor outcomes in the Southern African context to refine management strategies and policy interventions.

2. Methodology

• Study Design: A prospective, non-interventional multi-country cohort analysis.
• Setting & Population: The study utilized data from five sites within the International epidemiology Databases to Evaluate AIDS (IeDEA) Southern Africa consortium:
  • Zambia (CIDRZ, Lusaka)
  • South Africa (Themba Lethu Clinic, Johannesburg)
  • Malawi (Martin Preuss Center, Lilongwe)
  • Mozambique (Chiure Health Center)
  • Zimbabwe (Masvingo Health Center)
  • Inclusion Criteria: Patients aged ≥15 years diagnosed with pulmonary TB (bacteriologically confirmed or clinically diagnosed), with or without HIV, initiating TB treatment between October 2022 and March 2023. Data were analyzed as of March 31, 2025.
• Definitions:
  • Successful Outcome: Cured or completed treatment.
  • Unsuccessful Outcome: Death, loss to follow-up (LTFU), or treatment failure.
  • Other: Transferred out or unknown outcomes.
  • MDR-TB: Defined as resistance to at least rifampicin.
• Statistical Analysis:
  • Descriptive statistics summarized baseline characteristics.
  • Multivariable Mixed-Effects Models: Used to assess associations between patient characteristics (age, sex, BMI, education, HIV status, MDR-TB, diagnosis type) and treatment outcomes.
  • Adjustments: Models included country fixed effects to account for site-level heterogeneity.
  • Handling Missing Data: Multiple imputation by chained equations (10 imputations) was used for missing covariates.
  • Software: Analyses were performed in R (version 4.5.1).

3. Key Contributions

• Multi-Country Scope: This is one of the few studies to simultaneously evaluate TB outcomes across five diverse Southern African nations (Malawi, Mozambique, South Africa, Zambia, Zimbabwe) using a standardized IeDEA framework.
• Granular Outcome Analysis: The study distinguishes between specific types of unsuccessful outcomes (death vs. LTFU vs. failure) and identifies distinct predictors for each, rather than treating "unsuccessful" as a monolithic category.
• Focus on MDR-TB and Education: It highlights the disproportionate burden of MDR-TB on treatment failure and the protective role of formal education against mortality and LTFU.
• Surprising Finding on Smoking: The study identified a counter-intuitive association where moderate smoking risk was linked to reduced odds of death, prompting a hypothesis regarding increased healthcare utilization among smokers.

4. Results

• Cohort Characteristics:
  • Total participants: 1,438.
  • Median age: 39 years; 67% Male; 40% HIV-positive; 4% MDR-TB.
  • 84% had bacteriologically confirmed TB.
• Treatment Outcomes:
  • Success Rate: 80% (1,151 patients). This includes 606 cured and 545 completed treatment.
  • Unsuccessful Rate: 15% (221 patients), comprising:
    • Death: 6% (89 patients)
    • Loss to Follow-up: 9% (129 patients)
    • Treatment Failure: <1% (3 patients)
  • Other: 5% (66 patients; 49 unknown, 17 transferred out).
  • Variation by Country: Success rates ranged from 62% (South Africa) to 93% (Malawi). MDR-TB patients had a significantly lower success rate (56%) compared to non-MDR patients (81%).
• Predictors of Unsuccessful Outcomes:
  • MDR-TB: Strongly associated with increased risk of LTFU (aOR 2.4, 95% CI 1.17–4.97) and overall unsuccessful outcomes.
  • Education: Lack of formal education was a major risk factor. Secondary or higher education was associated with:
    • Reduced risk of death (aOR 0.3).
    • Reduced risk of LTFU (aOR 0.3 for secondary/higher; aOR 0.3 for primary).
  • Diagnosis Type: Bacteriologically confirmed TB was associated with a lower risk of death (aOR 0.45) compared to clinically diagnosed TB.
  • BMI: BMI ≥18 kg/m² was protective against death (aOR 0.6).
  • Smoking: Moderate risk of smoking was surprisingly associated with reduced odds of death (aOR 0.4), potentially due to earlier diagnosis or more frequent medical contact.
  • HIV Status: No significant difference in survival was found between HIV-positive and HIV-negative participants in this cohort, though HIV-positive patients experienced more non-fatal serious adverse events.

5. Significance and Implications

• Gap in Targets: The 80% success rate falls short of the WHO target of ≥85% (and the 90% End TB target), indicating a critical need for improved care strategies in the region.
• MDR-TB Management: The high rate of LTFU among MDR-TB patients (driven by long, toxic regimens) necessitates urgent adoption of shorter, safer, all-oral regimens and stronger retention strategies.
• Diagnostic Imperative: The protective effect of bacteriological confirmation suggests that strengthening diagnostic capacity (reducing reliance on clinical diagnosis) is vital to reduce mortality.
• Socioeconomic Interventions: The strong link between education and successful outcomes underscores that TB control is not just a medical challenge but a socioeconomic one. Interventions should target health literacy and support for patients with lower educational attainment.
• Policy Recommendations: The authors recommend enhanced patient education, counseling, direct observed therapy (DOT), and patient tracing systems to address the high rates of LTFU and death.

Limitations: The study notes that sites may not fully represent national health infrastructures, potential operational heterogeneity between sites, and the challenge of circular migration in Southern Africa affecting follow-up data completeness.