The potential impact, cost and cost-effectiveness of tuberculosis interventions - a modelling exercise

This modeling study across Brazil, India, and South Africa demonstrates that while only a few tuberculosis interventions—specifically vaccination, community-wide screening, and prison screening—can significantly reduce incident cases by 2050, many of these high-impact strategies are also cost-effective, necessitating a strategic shift in funding and policy toward broader population targets beyond clinic-diagnosed individuals.

The Gist

Imagine tuberculosis (TB) as a massive, stubborn fire burning in three different neighborhoods: Brazil, India, and South Africa. Each neighborhood has its own unique reasons why the fire keeps spreading—some have crowded prisons, others suffer from poor nutrition, and some have a high number of people living with HIV.

This paper is like a giant simulation game where researchers built a digital model to test different "firefighting strategies." They wanted to answer a simple question: If we have a limited budget, which tools will actually put out the most fire, and are they worth the cost?

Here is what they found, broken down into everyday concepts:

1. The "Small Bucket" vs. The "Big Bucket" Problem

The researchers tested many different tools. Some were like using a small cup to pour water on the fire, while others were like bringing in a fire truck.

• The Small Buckets (Limited Impact): Strategies like giving medicine only to people already diagnosed in clinics, or treating the immediate family members of those patients, were like using a small cup. Even though they helped the specific people involved, they only stopped about 0% to 5% of the total new fires. The paper explains that if you only treat people who are already coughing or their families, you are missing the "hidden" fires spreading silently in the community.
• The Big Buckets (High Impact): Three strategies acted like fire trucks, stopping 10% to 38% of new fires:
  1. Vaccination: A new vaccine for adults.
  2. Community Screening: Going door-to-door (or using mobile units) to check everyone in a neighborhood, even if they feel fine, using chest X-rays.
  3. Prison Screening: Systematically checking everyone in prisons (specifically in Brazil).

2. The Cost of Fighting the Fire

Just because a tool is powerful doesn't mean it's affordable. The researchers looked at the "price tag" for every "Disability-Adjusted Life Year" (DALY) saved. Think of a DALY as a "health year" lost to sickness.

• The "Free" Wins: Shortening the treatment for drug-resistant TB was actually cost-saving. It was like finding a way to fix a leak that not only stopped the water but also saved money on the plumbing bill.
• The "Cheap and Effective" Wins:
  • In Brazil, screening prisoners was incredibly cheap and effective (about $72 per health year saved).
  • In India, giving nutritional supplements to patients and their families was also very cheap and effective (about $167 per health year saved).
• The "Expensive but Worth It" Wins:
  • Vaccines and Community Screening were the most expensive upfront. However, in countries with very high TB rates (like South Africa), they were still considered a good value for money.
  • In India, however, the cost of community screening was so high compared to the local budget that it wasn't considered a "good deal" compared to other options.

3. The "Shorter Treatment" Trap

There was a specific finding about a new, shorter treatment for standard TB (Drug-Susceptible TB).

• The Expectation: People hoped a shorter 4-month treatment would be a miracle.
• The Reality: The study found that because the current 6-month treatment is already very good, making it shorter didn't stop many new fires. It just made the treatment slightly more convenient for the patient. Because it cost more to develop and implement but didn't stop the spread of TB any better, it was not a cost-effective use of money in this model.

4. The Big Lesson: Look Beyond the Clinic

The main takeaway from this study is a shift in perspective.

• Old Way: Waiting for people to get sick, come to a clinic, get diagnosed, and then treating them and their families. This is like waiting for a house to burn down before calling the fire department.
• New Way: Going out into the community, checking people who feel fine, and vaccinating people before they get sick. This is like installing sprinklers and checking for sparks before a fire starts.

The paper concludes that to really win the fight against TB, we need to stop focusing only on the people already in the clinic. We need to shift our funding and attention to broad, population-wide strategies (like mass screening and vaccines) and fixing the root causes (like nutrition and prison conditions). While these strategies require a bigger upfront investment, the study suggests they are the only ones capable of significantly reducing the number of new cases and deaths by 2050.

Technical Summary

Technical Summary: The Potential Impact, Cost and Cost-Effectiveness of Tuberculosis Interventions

Problem Statement
Despite an expanding toolkit for tuberculosis (TB) prevention, screening, diagnosis, and treatment, there is a scarcity of evidence directly comparing the population-level impact and cost-effectiveness of diverse interventions across settings with varying epidemiological backgrounds and structural drivers. Existing Target Product Profiles (TPPs) and modeling exercises often focus on single intervention classes or specific sub-populations, failing to account for the broader population dynamics or the interplay between intervention domains. Furthermore, few studies have compared interventions across prevention, screening, and treatment while simultaneously addressing key structural drivers of TB (such as incarceration, undernutrition, and HIV) in a standardized manner. This gap complicates policy decisions regarding funding allocation and product development in an era of constrained global budgets.

Methodology
The authors employed a deterministic, dynamic transmission model calibrated to epidemiological indicators in three distinct country archetypes: Brazil, India, and South Africa. These countries were selected to represent diverse TB burdens (incidence ranging from 49 to 427 per 100,000) and dominant structural drivers: incarceration (Brazil), undernutrition (India), and HIV (South Africa).

• Intervention Scope: Seven interventions were modeled across three domains:
  • Prevention: A prophylactic vaccine (prevention of disease) and Tuberculosis Preventive Treatment (TPT) for household contacts.
  • Screening and Diagnosis: Symptom-agnostic community-wide screening (using chest X-ray and Xpert Ultra), improved diagnostic approaches in clinics (replacing empirical diagnosis with Xpert Ultra), and expanded Drug Susceptibility Testing (DST).
  • Treatment: Shortened regimens for drug-susceptible (DS) TB and shortened, more efficacious regimens for drug-resistant (DR) TB.
  • Structural Drivers: Annual screening in prisons (Brazil) and nutritional supplementation for patients and household contacts (India). No separate structural driver intervention was modeled for South Africa, as ART availability is already widespread.
• Standardization: Implementation was standardized with a start year of 2025, a linear scale-up to 80% coverage of the target population by 2030, and a time horizon extending to 2050 to capture indirect effects. Community-wide screening was modeled as a five-year campaign followed by cessation.
• Model Structure: The model was age-stratified and incorporated specific natural history parameters for children and adults. It distinguished between bacteriologically positive symptomatic (sTB) and asymptomatic (aTB) infectious states, and non-infectious bacteriologically negative (nTB) states. Structural drivers were integrated as specific sub-populations with distinct mixing patterns and risk profiles.
• Economic Analysis: A health system perspective was adopted using 2023 USD. Costs were sourced from literature and public data, with uncertainty modeled via gamma distributions. Incremental Cost-Effectiveness Ratios (ICERs) were calculated as cost per Disability-Adjusted Life Year (DALY) averted relative to a Business-As-Usual (BAU) scenario, using country-specific cost-effectiveness thresholds.

Key Results

• Epidemiological Impact: Only three interventions prevented more than 10% of incident TB episodes by 2050 across the modeled settings:
  1. Vaccination: Prevented a median of 15–28% of sTB episodes.
  2. Community-wide Screening: Prevented 32–38% of sTB episodes (noting a rebound in incidence after the screening campaign ceased in 2035).
  3. Prison Screening (Brazil): Prevented 23% of sTB episodes.
  • Other interventions, including TPT, improved clinic diagnostics, DST expansion, and shortened DS-TB treatment, prevented fewer than 5% of total episodes. Shortened DS-TB treatment showed 0% impact due to the assumption of non-inferiority (no improvement in cure rates) over current standard regimens.
• Cost-Effectiveness:
  • Shortened DR-TB treatment was the only intervention that was cost-saving (negative incremental cost) across all settings.
  • Structural driver interventions demonstrated high cost-effectiveness: Prison screening in Brazil (72 USD/DALY) and nutritional supplementation in India (167 USD/DALY).
  • Community-wide screening and vaccination showed variable cost-effectiveness depending on the setting and cost scenario. In Brazil and South Africa, low-cost community screening and vaccination fell below cost-effectiveness thresholds. In India, high-burden settings favored these interventions, but the high cost of screening in India rendered it less efficient.
  • Shortened DS-TB treatment was not cost-effective in any country due to the lack of additional epidemiological impact combined with higher regimen costs.
• Budget Impact: Incremental costs varied widely, from cost savings (DR-TB treatment) to over 240 billion USD (high-cost community screening in India).

Significance and Claims
The paper argues that interventions with meaningful epidemiological impact must target populations beyond clinic-diagnosed individuals and their immediate households. The authors claim that the potential population-level impact is largely driven by the size and burden of the target population (Population Attributable Fraction). Consequently, restricting interventions to clinical settings imposes a "hard ceiling" on potential benefits.

The study highlights that addressing structural drivers (incarceration, undernutrition) offers value for money comparable to historical successes like ART roll-out. Furthermore, the authors assert that achieving meaningful TB control requires a priority shift in funding, policy, and product development to support interventions that reach high-risk, non-clinical populations. They suggest that while community-wide screening is impactful, its efficiency is context-dependent, and in lower-burden settings, a combination of vaccination and targeted screening (e.g., in prisons) may be more appropriate. The work underscores the necessity of a cross-silo approach in research and funding, urging that TPPs consider population-level impact rather than just individual-level benefits.