Evaluating the Cost-Effectiveness of an AI-Driven Chatbot for Tuberculosis Screening in Primary Health Care Settings in Indonesia: A Decision-Analytic Modeling Study

Abstract

Background: Tuberculosis (TB) remains a major public health challenge in Indonesia, with persistent gaps in early detection and case finding at the primary care level. Digital health innovations, particularly artificial intelligence (AI)–based chatbots, offer a scalable approach to improve symptom screening and triage. However, evidence on their economic value within health systems remains limited. This study evaluated the cost-effectiveness of an AI-based chatbot for TB screening compared with standard symptom-based screening in primary health care settings in Indonesia. Methods: A decision-analytic model was developed to compare AI-based chatbot screening with conventional screening from the health system perspective over a one-year time horizon. Model inputs were derived from secondary data sources, including the Indonesian TB surveillance system (SITB), published literature, and pilot data from 50 primary health centers, covering a cohort of 120,000 individuals. Effectiveness was measured as additional TB cases detected and disability-adjusted life years (DALYs) averted. Costs included screening, diagnostic confirmation, and treatment initiation. Incremental cost-effectiveness ratios (ICERs) were calculated. Deterministic and probabilistic sensitivity analyses were performed. Results: Among 120,000 individuals screened, the AI-based chatbot identified 4,320 presumptive TB cases and 1,080 confirmed cases, compared to 3,150 and 840 under standard screening, yielding 240 additional cases detected (28.6%). The intervention averted 385 DALYs, compared with 290 DALYs with conventional screening, for a total of 95 additional DALYs. Total costs were USD 412,560 for the chatbot strategy and USD 398,200 for standard screening. The ICER was USD 151 per DALY averted, well below Indonesia’s willingness-to-pay threshold (USD 1,200). Sensitivity analyses confirmed robustness. Conclusion: AI-based chatbot screening is a highly cost-effective strategy to improve TB case detection in Indonesia. Scaling up this approach could enhance efficiency, accelerate diagnosis, and support national TB elimination targets.