CXR/CAD Screening

TB Screening

Every year, more than 10 million people fall sick with TB worldwide. Out of these, about 2.4 million people are being missed, because they are either not diagnosed or treated properly. Systematic screening for active TB in high-risk populations, provision of preventive treatment (TPT) and prompt access to appropriate treatment for confirmed active TB cases are crucial to END TB. Digital chest radiography, when used as rapid triage to select individuals for TPT or for bacteriological testing, facilitates screening of large groups at very low costs per person screened.

“TB screening is often defined as “systematic identification of people with suspected active TB in a predetermined target group, using tests, examinations, or other procedures which can be applied rapidly and do not only target individuals seeking care for symptoms or signs”. [1]

Enhanced efforts with targeted screening strategies are much needed for vulnerable communities and to accelerate the national TB control efforts towards TB control and elimination. [2] Symptom questionnaires and chest radiography are the most available and best documented methods to screen for active TB disease [3]. While the use of Xpert® in programmatic settings has expanded in recent years, the WHO has also recommended use of more cost effective diagnostic algorithms through screening tools such as CXR [4]. Digital radiology overcomes the barriers of using traditional film based chest X-rays with human readers for (cost)effective digital screening in low resource-high burden settings. Computer Aided Detection for TB now surpasses the performance of a trained human reader and can be used for passive and active TB case finding as well as in prevalence surveys [5]. Combining CAD and clinical information to estimate the risk of active disease is a promising tool for TB screening [6].

Ministries of Health and funders increasingly focus on systematic screening for TB in high risk groups. Where operationally feasible multi-disease screening can be included in programs for cross-disease impact. Digital X-ray and Artificial Intelligence support these screening demands in resource-constrained settings with innovative and proven CAD systems to simultaneously detect and quantify abnormalities suggestive of TB, other lung diseases and cardiomegaly within less than 20 seconds.

Stop TB Partnership – on use of CAD

Artificial intelligence (AI) technologies offer unprecedented opportunities within a healthcare context. AI is increasingly being applied in the field of medical imaging for the computer-aided detection (CAD) of diseases, including cancer, COVID-19, and TB.

A diverse range of AI products for the recognition of TB-related abnormalities from chest X-rays are now commercially available. Evidence produced by the Stop TB Partnership informed the World Health Organization’s TB screening guideline update in April 2021, when, for the first time, AI was recommended as a triage tool for TB in adults. The potential of AI to accelerate TB detection, particularly in rural and low-resource contexts, is tremendous.

CAD presents an opportunity to improve the detection of TB by circumventing inefficiencies in the interpretation of CXR images, automating and standardizing X-ray interpretation, and supplementing existing human health workers. When used in combination with ultra-portable X-ray systems, the promise of CAD technology can be extended to hard-to-reach key populations.

The Global Fund – on procurement of CXR

“Consider the use of CAD where human readers may not be available to read Chest X-rays & to decrease the workload of radiographers. Consider CAD technologies that offer multi-disease potential given the system strengthening potential of CXR and CAD. Consider the associated cost for the entire ecosystem like installation cost, PACS, site preparations, recurrent cost, service and maintenance, local calibration of CAD thresholds”

IOM – CAD4Silicosis

Centro de Saúde Ocupacional de Ressano Garcia, Moçambique at the border with South Africa provides essential screening services for Mozambican migrant mine workers. Occupational Health Consultation Rooms include screening on symptoms and Chest X-ray abnormalities suggestive of silicosis and/or TB using CAD for presumptive silicosis and TB detection.

Lungs' CAD image

Silicosis is an occupational lung disease caused by chronic exposure to silica dust, primarily affecting current and former miners. As silicosis is irreversible and can be fatal, early and timely diagnosis is crucial. While there is no specific diagnostic test for silicosis, radiological abnormalities on chest X-rays, occupational history, and physical examinations can help confirm the diagnosis.

Individuals with silica dust exposure are at an increased risk for Tuberculosis (TB): the TB incidence among silicosis patients can be significantly higher than in the general population. Moreover, some develop Silicotuberculosis, being affected by both silicosis and TB at the same time.

Routine screening for silicosis and TB is essential for those exposed to silica dust. However, challenges persist due to a shortage of medical experts to assess chest X-rays accurately.

GHS front runner in applying Artificial Intelligence to screen for TB and other lung diseases

WHO on screening

The available evidence suggests that screening, if done in the right way and targeting the right people, may reduce suffering and death.

Knut Lönnroth, WHO - PSI/GTB UNION Liverpool, October 2016.

  • CXR alone cannot establish a diagnosis for TB – bacteriological confirmation must always be attempted
  • In absence of bacteriological confirmation, sometimes clinical diagnosis is needed
  • If patient is not critically ill, a wait-and-see approach can be used
  • WHO has no specific guidance on use of CXR in clinical diagnosis
  • Proportion of TB patients with non-confirmed TB is a possible indicator for quality control

The primary objective of systematic TB screening is early TB case detection, which requires a sensitive screening tool

  • High sensitivity of CXR for TB screening (87-98%)
    • Highly sensitive if using “any abnormality consistent with TB” as criteria for an abnormal result
    • Means the vast majority of those with TB who undergo CXR screening will have an abnormal result
  • Low specificity of CXR for TB screening (46-89%)
    • Means the test will also be abnormal in individuals with other lung abnormalities besides TB that also need to be followed up (e.g. cancer, pneumonia, emphysema) -> benefit to individual being screened
    • Also identifies individuals with inactive TB/fibrotic lesions, who are at high risk of developing active TB in future and require follow-up
    • However, means it needs to be coupled with a bacteriological test with high sensitity and specificity for diagnosis

WHO recommends that priority be given to populations with:

  • High TB prevalence (risk factor profile and/or poor access/delay)
  • High risk of severe negative consequences if diagnosis is missed
  • High transmission risk

In particular systematic screening for TB disease should always be conducted for:

  • Households and close contacts of TB patients
  • People living with HIV
  • Miners exposed to silica dust
  • Prisoners

Conditionally recommended systematic screening for TB disease among:

  • People with clinical risk factors for TB seeking healthcare, in settings with TB prevalence of 0.1% or higher
    • Malnourishment, diabetes, history of previous TB, chronic lung disease, etc.
  • Populations with limited access to healthcare
    • Urban poor, refugees, homeless, other vulnerable or marginalized groups
  • General population in settings with TB prevalence of 0.5% of higher

For the above conditional screening populations consideration should be given to weighing benefits and risks of screening and prioritizing groups that have the greatest burden of vulnerability in a particular setting.

WHO Operational Handbook on Systematic screening for tuberculosis disease – March 2021

In particular people living with HIV are 16 (uncertainty interval 14–18) times more likely to fall ill with TB disease than people without HIV. TB is the leading cause of death among people with HIV. HIV and TB form a lethal combination, each speeding the other's progress. Without timely detection, hence no access to proper treatment, 60% of HIV-negative people with TB on average and nearly all HIV-positive people with TB will die, according to WHO.

Stop TB partnership

A Paradigm Shift is needed in the way we fight TB. Unless we speak about Active Case Finding, X-ray, contacts, prevention (amongst others) we will not reach our targets.

“We are committed to providing an inclusive and supportive platform, including our technical expertise and catalytic funding, to strengthen the innovation ecosystem and accelerate the introduction and scale-up of novel approaches and new tools”

ACF innovations to support Paradigm Shift implementation

  • Systematic screening aims to detect more cases, and to detect them early
  • Innovations in diagnostic imaging make screening in high risk groups faster and more sustainable
  • Digital Chest X-ray (dCXR) with Computer Aided Detection (CAD) as a rapid and automated triage test before Xpert® MTB/RIF can significantly reduce screening cost
  • To scale up ACF in risk groups rapid and more affordable diagnostic pathways are required
  • Relatively high cost and 2 hours processing time still limit the uptake of Xpert® as point-of-care test
  • CAD4TB automatically scores chest radiographs between 0 -100 on abnormalities consistent with TB in 1 minute and surpasses trained human reader performance
  • These innovations support the Stop TB Partnership Paradigm Shift implementation through new diagnostic pathways to detect more TB cases earlier at much lower cost

PEPFAR on HIV/TB screening

U.S. President’s Emergency Plan for AIDS Relief (PEPFAR) launched in September 2023 an effort to detect two million active tuberculosis (TB) cases over the next five years and prevent at least 500,000 TB-related deaths among people living with HIV. According to estimates, more than half of the current TB cases among people living with HIV go undiagnosed. PEPFAR and its partners are committed to using new tools and new approaches that dramatically increase TB diagnosis and treatment coverage for people living with HIV.

The first phase of this initiative will prioritize PEPFAR-supported countries with a high burden of HIV and TB, such as South Africa, Nigeria, Kenya, Tanzania, and Mozambique. The effort will focus primarily on people living with HIV most at risk to die of TB such as children, older adults, and people with advanced HIV.

FY 2024 Technical Considerations

Intensified TB case-finding among all People Living with HIV

“All people living with HIV must be screened at every clinical encounter for TB symptoms. The new 2021 WHO recommendations on TB screening include symptom screening at each encounter, and given the sub-optimal sensitivity of symptom screening, consideration of adding Chest X-Ray (CXR), C-reactive Protein (CRP), or a molecular WHO-recommended Rapid Diagnostic (mWRD) test to the screening algorithm for PLHIV.”

FY-2024-PEPFAR-Technical-Considerations.pdf (state.gov).