Community Health Officers Lead the Charge for Statewide AI-based TB Screening

Authored by: Priyam Nayak

In the remote, hilly districts of Lunglei and Aizawl in Mizoram, an AI application is playing a vital role in the state's efforts to combat tuberculosis. Mizoram, which saw a rise in TB-related mortality in 2023, is also contending with a significant number of multi-drug resistant cases. Multi-drug resistant TB (MDR-TB) is resistant to first-line drugs, making it harder to treat and increasing the risk of spread and mortality.?

Since May 2024, health camps supported by civil society organizations like Piramal Foundation and Jhpiego have accelerated TB screening efforts in the state.?

Community Health Officers (CHOs) working in these health camps have been increasingly using Cough Against TB developed by Wadhwani AI to screen individuals. Developed in partnership with the Central TB Division, with the support of the USAID-backed TRACE TB project, the application analyzes cough sounds and symptoms factored in by healthcare workers at Health and Wellness Centres to predict whether an individual has presumptive pulmonary TB.?

As a smartphone application that needs no extra equipment and can work in offline environments, Cough Against TB has emerged as a valuable ally for healthcare workers. Over the last few months, it has been embedded seamlessly into their existing workflows, with CHOs advocating for the statewide deployment of the solution.?

Insights from the Frontline

In Mizoram, Wadhwani AI trained 300 CHOs across 115 health camps. The training not only focused on familiarising them with the intricacies of Cough Against TB, but also offered insights on handling TB diagnostics, using the Ni-kshay portal, and other related topics.?

Many CHOs had never used an AI solution before, and the training reduced their initial resistance by improving their understanding of the solution and its practical benefits—thereby fostering acceptance. Our capacity-building efforts addressed the uncertainties usually associated with new technology, paving the way for its broader adoption.

Mimi Klalroliani, CHO at Mission Venghthlang Ayushman Arogya Mandir (AAM) in Aizawl West, says: "When screening was based on the 4S criteria—which entails ascertaining symptoms like cough, weight loss, fever, and night sweats—people would be unconvinced that their cough could indicate TB. Now, with the Cough Against TB app, they come back to check whether their cough might signal something abnormal. They are also curious about the application. The screening process is much easier, and when the app identifies a presumptive TB case, people are more willing to undergo further testing because it is a more reliable method of initial validation than just listing symptoms."

Mimi and her colleagues' experiences were the catalysts for the deployment of Cough Against TB at all health and wellness centers across Mizoram.?

"It’s much easier and takes less time because if the? application doesn't identify presumptive TB cases, we don't need to proceed with testing those patients. This helps us focus our resources and efforts only on those who need further investigation," shares RL Chungnungi, a CHO at Durtlang AAM in Aizawl West.

Since 2018, Wadhwani AI has been the official partner of the Central TB Division (CTD). Cough Against TB was developed under the guidance of the CTD in alignment with the National TB Elimination Program (NTEP) guidelines. The AI model is trained on data collected from health facilities across India, and uses a binary classification model that assigns a risk score based on cough sounds and symptoms to indicate whether the individual is likely to have TB.

A Reliable Ally in the Fight Against TB Across India

As of September 2024, Cough Against TB is deployed across 25 districts in 9 states where over 100,000 individuals have been screened for TB. The application plays a crucial role as a point-of-care screening tool in community settings in multiple states to address gaps in screening infrastructure in rural and semi-urban areas.?

Cough Against TB is deployed in AAMs in seven districts in Nagaland, whose mountainous terrain is an obstacle for seamless delivery of public health services. The lack of connectivity and infrastructure in remote areas makes testing and treating TB patients challenging for healthcare workers. The Cough Against TB application comes in handy for screening patients in outreach areas where transport facilities are limited, and collecting samples for laboratory testing and sending them back for initial screening may not be viable.?

Rarenungla Pongener, a CHO working at Aosetsu AAM in Mokokchung, says, "Earlier, conducting TB tests for people in remote villages was difficult. However, the process has become much easier with the introduction of the Cough Against TB app. We can now screen suspected TB patients at their doorsteps using the app and follow the necessary steps based on the app's predictions. This has significantly improved accessibility to TB screening and care, especially in hard-to-reach areas."

Bringing Together AI and Community Engagement to Eliminate TB

As per the World Health Organization's Global TB Report 2023, 0.36 million (13%) TB cases were missed in India. Undiagnosed TB cases are a significant driver for higher transmission rates. It is essential to identify every case within the community, ensure early diagnosis, and provide comprehensive treatment and support until the patient is TB-free.

Addressing the challenge of missed TB cases requires a comprehensive approach that places communities at the center of the effort. Engaging local populations, scaling up AI solutions, and improving program management are key to finding undiagnosed cases and ensuring better treatment outcomes.

Cough Against TB's deployment initiation in Mizoram exemplifies how AI solutions attuned to healthcare workers' needs can cut through the complexities of TB screening and streamline resource-draining tasks, allowing them to focus on what truly matters: timely, compassionate care. The solution's ease of use can also encourage individuals to get themselves screened when they notice symptoms. Every case flagged by Cough Against TB represents a potential diagnosis and a step forward toward a future where no patient falls through the cracks.?