Screening

The evaluations reviewed by the GDG demonstrated substantial variation in the diagnostic accuracy (sensitivity and specificity) of CAD programmes across settings, even when using the same technology set to the same threshold. Thus, it will be essential to calibrate the threshold to be used for any given software for each setting and population in which it will be used in order to ensure that the accuracy, predictive values, overall yield and requirements for further diagnostic testing are as expected.

Recommendations on specific tools to be used for screening people living with HIV are described in detail in Section 3. Persons living with HIV who have a positive or abnormal screening test should go on to have primary diagnostic tests to confirm or rule out TB disease including an mWRD and lateral flow urine lipoarabinomannan assay (LF-LAM), if eligible (8).

People living with HIV are approximately 19 times more likely to develop TB disease than those without HIV; in 2019, an estimated 44% of people living with HIV who also had TB disease did not reach care, and 30% of all HIV-related deaths were due to TB (1). Thus, ensuring early detection and treatment for TB among all people living with HIV is crucial for reducing morbidity and mortality in this group.

Populations with structural risk factors for TB are those that are at increased risks of TB and of poor health outcomes from TB due to structural determinants in their environment, defined as the conditions that generate or reinforce social stratification (e.g. socioeconomic inequalities, population growth, urbanization), and therefore give rise to an unequal distribution of key social determinants of TB epidemiology, such as poor housing, poverty and malnutrition, which in turn influence exposure to risk, vulnerability and ability to recover after developing the disease (16,17).

The magnitude and balance of desirable and undesirable effects vary according the epidemiological conditions (the prevalence of TB and of risk factors) and the intensity of the screening intervention being implemented (the coverage of the population and the sensitivity of the screening test and algorithm). There is currently no evidence that population-wide screening using less sensitive screening algorithms that begin with symptom screening are effective at reducing the population prevalence or transmission of TB.

Standard monitoring and evaluation procedures may be complemented by operational research aimed at improving the performance of screening in the local setting as well as research aimed at improving the global evidence base for screening. Topics that may be explored include:

Across all populations and tools, more research is needed to evaluate the accuracy and effectiveness of complete screening and diagnostic algorithms, including symptom screening, CXR, CRP and mWRDs used in various combinations with diagnostic evaluation. Research into their effectiveness should include measures of the impacts on patient-important outcomes, such as mortality and treatment success.

Contact screening should always be done when a person with TB has any of the following characteristics: bacteriologically confirmed pulmonary TB, proven or presumed multidrug-resistant TB or extensively drug-resistant TB, is a person living with HIV or is a child younger than 5 years. Among contacts of patients with bacteriologically confirmed TB, the weighted pooled prevalence of TB was 3.4% (95% CI: 2.9–3.8). Among contacts of patients with multidrug-resistant or extensively drug-resistant TB the weighted pooled prevalence of TB was 3.7% (95% CI: 2.4–5.3).

For people living with HIV in settings with different TB prevalences, more research is needed to evaluate the accuracy and predictive value of measuring CRP above any cut-off higher than 5 mg/L for TB screening, when it is used either alone or in combination with other screening tests.

Further evidence is needed about the performance of CAD software stratified according to the characteristics of the individual being evaluated (e.g. by smear status, HIV status, age cohort, history of TB, smoking status, sex) to allow for better setting-specific and patient-specific calibration of CAD programmes.

More research on users’ perspectives is needed about CAD technologies in TB screening and triage, including their perceived acceptability to patients, providers and other stakeholders.