Skip to main content

Scaling Intensified TB Screening in Tertiary Hospital Ecosystems

Bridging the Detection Gap through Intensified OPD/Inpatient Screening and AI-Driven Triage in Medical Colleges and District Hospitals

Medical colleges and district hospitals serve as a strong, yet underexplored, opportunity to identify missing cases, and integrate them into the care cascade. Medical colleges and district hospitals in India serve as tertiary referral hubs, witnessing massive daily footfalls in Outpatient Departments (OPDs) and high bed-occupancy in wards, contributing 15-20% of total notifications consistently for decades. Despite the high potential, medical colleges, for example, face massive pre-testing attrition, with 67% of symptomatic patients dropping out between the initial screening and the laboratory.

Current data also reveals a stark contrast in screening efficiency in both OPD and IPD settings. One study, using symptom-based screening in an OPD setting, showed a modest yield of 0.14% (NNS ≈ 737). However, targeted screening in IPD settings was highly effective, with a Number Needed to Screen (NNS) of only 48. Furthermore, targeted screening in diabetic clinics using the 4-symptom complex (cough, fever, weight loss, night sweats) yields an NNS between 65 and 147.

Along with the institutional fast-tracking, there is a compelling opportunity to deploy a robust Intensified Case Finding (ICF) operational model in OPD and Indoor facilities of the tertiary care settings. By transitioning from symptom-based screening to CAD-supported Chest X-Ray (CXR) screening, medical institutions can bypass the low sensitivity of symptom-only protocols and eliminate pre-testing attrition via dedicated "Fast-Track" channels.

 indicates a required field.

Section I: Contact Details

Section II: Organizational Eligibility

Note your organization’s ability to meet the following:

(1,000-character limit)
(1,000-character limit)

Section III: Proposal Description

(3,000-character limit)
(3,000-character limit)
(2,000-character limit)
(2,000-character limit)
(1,000-character limit)