This can only be possible when we confront one of the key obstacles to the treatment of childhood TB – underreporting of cases, particularly of latent TB infections (LTBIs), which evade detection due to the asymptomatic nature of infections. However, this obstacle is not insurmountable. By expanding our screening efforts, improving diagnostic services and leveraging technological innovations like Artificial Intelligence (AI), we can improve the way we track, test and treat both latent and active TB infections among our children.

Latent TB Infections (LTBIs) – where children are infected but asymptomatic – pose a major challenge in eradicating childhood TB. While LTBI rates in Indian adults range from 21% to 48%, the prevalence amongst under-five children was found to be 11.7% in a study conducted by the ICMR-National Institute of Research in Reproductive and Child Health (ICMR-NIRRCH)1. Overall, it still remains a blind spot that emerges from several key challenges pertaining to the identification and reporting of LTBIs in children.

Firstly, children with LTBIs are asymptomatic, making it difficult to identify their diseases without targeted screening programs. Secondly, the smaller number of TB bacteria in children’s lungs can evade detection by traditional diagnostic tests. Thirdly, testing methods like Tuberculin Skin Testing (TST) face limitations as variations in test administration and interpretation, as well as the need for multiple visits, can hinder accurate diagnosis and reporting using TST. More recent and superior alternatives to TST, like IGRAs, which are based on the detection of interferon-gamma released by lymphocyte cells upon contact with TB bacteria, have been introduced in the National Tuberculosis Elimination Programme (NTEP), a public health initiative of the Government of India. IGRAs are not yet uniformly implemented in all states and are not well accepted due to the invasive nature of the test.

Countering these challenges requires a new and innovative approach to TB diagnosis and testing, especially amongst vulnerable populations. Recent research in the field indicates the pivotal role of community screening in identifying LTBIs in children, especially those residing in areas with poor sanitation. This approach can also prove critical in understanding the general awareness around latent TB amongst people, making intervention policies more streamlined and effective.

Apart from community screening, there is also a need to incorporate a “test and treat” approach for LTBIs amongst children, which involves diagnosing at-risk children and promptly initiating treatment for those who test positive. This approach should be accompanied by the creation of treatment guidelines for children who have been in close contact with patients of drug-resistant TB and a mechanism to regularly retest children who live in high TB burden areas. There is also a need to scale up the use of alternative testing methods like a newer TB skin test, the C-TB test having similar sensitivity as that of IGRA in community screening, to identify LTBIs more accurately.

Targeting LTBIs will be a crucial step in eliminating TB since LTBI increases the risk of developing active TB infections. There is an underestimation of the true burden of active childhood TB as the majority of children are sputum smear negative microscopically. Artificial Intelligence and Machine Learning (AI/ML) technologies can aid in the diagnosis of active TB. ICMR-NIRRCH’s collaboration with IIT Bombay to develop efficient AI algorithms for early radiological diagnosis of childhood TB is a significant step forward in bolstering TB diagnosis and reporting efforts in India.

By bridging gaps in existing diagnostic technologies and leveraging technological innovations to enhance TB reportage across demographic groups, India can accelerate its march towards a TB-free India and promise a healthy future for its next generation.

This article is authored by Geetanjali Sachdeva, director, ICMR-National Institute of Research in Reproductive & Child Health (NIRRCH).