Objective: Diagnosing extrapulmonary tuberculosis (EPTB) is difficult because of its paucibacillary nature and the need for invasive sampling. This study evaluated the diagnostic performance and inter-test agreement of commonly used methods and identified demographic and clinical features that may assist diagnosis. Methods: This retrospective cohort study was conducted at a tertiary care hospital (2018–2025). EPTB was defined as at least one positive result from acid-fast bacilli (AFB) smear microscopy, polymerase chain reaction (PCR), solid or automated liquid cultures. Demographics, comorbidities, symptoms, laboratory results, histopathological and radiological findings were obtained from medical records. The positivity proportions were calculated, and agreement was assessed using the kappa coefficient. Results: 46 adult patients with EPTB were included. The mean age was 51.6 ± 20.1 years, with female predominance (67.4%). The most frequent sites were the lymph nodes (32.6%), genitourinary tract (21.7%), and musculoskeletal system (21.7%). The highest positivity was observed with automated liquid culture (76.1%), followed by solid culture (71.7%) and PCR (59.5%); AFB smears had the lowest rate (19.6%). Granulomatous inflammation was detected in 63.2% of patients who underwent pathological examination and caseous necrosis in 36.8%. There was slight agreement between AFB smear microscopy, culture, and histopathology. PCR results showed poor agreement with solid and automated liquid cultures, including negative kappa values. Radiological examinations revealed pathological findings in 88.2% of patients, predominantly lymphadenopathy. Conclusion: Clinical and laboratory diagnosis of EPTB remains challenging. These findings support a multimodal strategy that integrates microbiological, molecular, histopathological, and radiological methods to improve diagnostic accuracy.
Güler et al. (2026) studied this question.