A Prospective Feasibility Evaluation of an AI-Assisted Clinical Decision Support Intervention for Diagnosis and Triage in a Semi-Urban Healthcare Setting

DOI

https://www.doi.org/10.59256/ijire.20260704002

Published In

Volume 7, Issue 4 (2026)

Published Online

July-August 2026

Pages

07-12

Article Type

Original Article

Check for updates

Background: In semi-urban and rural India, a severe shortfall of specialist physicians leaves junior doctors and community health workers making high-stakes diagnostic decisions without senior clinical support. AI-assisted clinical decision support systems have been proposed to bridge this gap, yet real-world evidence of their diagnostic performance in resource-limited settings remains scarce. Objective: To evaluate the diagnostic accuracy of Bhairav AI against senior clinician judgement in active semi-urban practice, and to assess its utility as an AI-assisted support tool for community health workers operating under medical supervision. Methods: A one-month prospective feasibility study at Mahadev Nursing Home, Chhattisgarh, India. Participants were three senior consultants, five junior doctors, and ten community health workers. Bhairav AI was deployed on desktop devices for junior doctors and on mobile devices for CHWs at field and satellite clinics under medical supervision. Primary outcomes were diagnostic concordance with senior consultant assessments, consultation duration, and tool uptake. Results: Full diagnostic concordance with senior consultants was achieved in 85.0% of 100 cases, with partial concordance in 8.0% -a combined agreement of 93.0% -and discordance in 7.0%. All 13 critical findings were correctly flagged; no critical condition was missed. Mean consultation time fell from 9.53 (SD 0.78) to 6.19 minutes (SD 0.61), a 35.1% reduction (p<0.001). Junior doctor uptake rose from 25% on day one to 90% by the end of the month. CHWs using the mobile version under supervision reported improved confidence in identifying and communicating urgent presentations. Conclusion: Bhairav AI demonstrated clinically meaningful diagnostic concordance and significant workflow gains in real-world semi-urban practice. Junior doctors retained full clinical authority, using the tool as a diagnostic aid; CHWs benefited from structured triage support under medical supervision. These findings justify a larger, multi-site randomised study to confirm diagnostic performance, safety, and cost-effectiveness across a broader range of primary care settings in India.