Impact of demographic and clinical information on pulmonary tuberculosis diagnosis via chest X-ray interpretation: a cross-sectional study in an Australian tertiary hospital

Tuberculosis (TB) remains a global health concern. Diagnosis often relies on chest X-ray (CXR) interpretation, particularly in low-incidence countries such as Australia, where TB predominantly affects individuals from high-prevalence regions. The accuracy of CXR-based diagnosis may be influenced by the availability of clinical history and epidemiologic risk indicators, such as patient country of origin. This study evaluated the impact of clinical history and country of origin on clinicians’ diagnostic accuracy for TB on CXR. This cross-sectional descriptive study was conducted at a tertiary teaching hospital in Victoria, Australia, involving 52 doctors of varying specialties and experience. Participants reviewed 36 anonymized CXRs in three sequential scenarios: (1) image only, (2) image with clinical history, and (3) image with clinical history plus patient country of origin. Each set contained 4 confirmed TB cases and 8 non-TB pathologies. Diagnostic accuracy and confidence were analyzed using descriptive statistics and repeated-measures ANOVA. Diagnostic accuracy for TB as the primary diagnosis improved with added context: 21.2% (image only), 37.5% (image + history), and 60.1% (image + history + country of origin). Considering any of the top three diagnoses, accuracy increased from 44.2% to 78.8%. Clinicians with > 10 years’ experience and at consultant/registrar level demonstrated higher accuracy. Only 15.4% of participants reported high confidence in CXR interpretation. Including clinical history and country of origin significantly enhances TB diagnosis on CXR. These findings highlight the critical role of epidemiologic context and suggest the need for targeted education and structured reporting systems in multicultural healthcare settings.