The Developing a Tuberculosis Screening Model in Chai Nat Province

Abstract

            Tuberculosis (TB) remains a major public health problem, and screening based on the standard seven-risk-group definition may miss true cases, so that many patients enter care only after the disease has advanced. This study aimed to analyze the characteristics of TB patients outside the risk-group definition (Walk-in), to develop the TB Screening Chainat Model, and to establish an effectiveness baseline. Using participatory action research following the Kemmis and McTaggart cycle, it triangulated provincial registry data on 763 patients (fiscal years 2566-2568), an in-depth survey of 40 Walk-in patients, and interviews with 13 health staff across three service levels. Quantitative data were analyzed with descriptive statistics, chi-square, Cochran-Armitage trend, and exact binomial tests; qualitative data used content analysis.

            Walk-in patients comprised 60.7% of all cases and represented the largest infectious reservoir, with 1.49 times more sputum-positive cases in absolute number than the in-definition group, although the positivity rate did not differ (65.0% vs 67.3%, p = 0.509). They were predominantly male (75.4%), elderly (mean 57.2 years), of low socioeconomic status, and underweight. Their mortality reached 15.6%, significantly exceeding the national target (p < 0.001). Content analysis yielded four themes (risk-definition gaps, data silos, X-ray as a game changer, and a practical tool), producing a model built on three pillars: a context-specific risk checklist, proactive chest radiography, and seamless data integration. When piloted in the target districts, measured baseline effectiveness indicators showed 3,119 proactive screenings across 21 events, high satisfaction among service recipients (4.13 ± 0.62) and staff (4.54 ± 0.55), while the new-case yield remains under confirmation.

            The standard definition alone is insufficient; the model reframes screening from passive to proactive, context-based case finding and is suitable for policy scale-up.