High Sensitivity of Ultrasound for the Diagnosis of Tuberculosis in Adults in South Africa: A Proof-of-concept Study
Overview
Authors
Affiliations
Background: There are limited data on the performance characteristics of ultrasound for the diagnosis of pulmonary tuberculosis in both HIV-positive and HIV-negative persons. The objective of this proof-of-concept study was to determine the sensitivity and specificity of ultrasound for the diagnosis of tuberculosis in adults.
Methods: Comprehensive thoracic and focused abdominal ultrasound examinations were performed by trained radiologists and pulmonologists on adults recruited from a community multimorbidity survey and a primary healthcare clinic in KwaZulu-Natal Province, South Africa. Sputum samples were systematically collected from all participants. Sensitivity and specificity of ultrasound to detect tuberculosis were calculated compared to a reference standard of i) bacteriologically-confirmed tuberculosis, and ii) either bacteriologically-confirmed or radiologic tuberculosis.
Results: Among 92 patients (53 [58%] male, mean age 41.9 [standard deviation 13.7] years, 49 [53%] HIV positive), 34 (37%) had bacteriologically-confirmed tuberculosis, 8 (9%) had radiologic tuberculosis with negative bacteriologic studies, and 50 (54%) had no evidence of active tuberculosis. Ultrasound abnormalities on either thoracic or abdominal exams were detected in 31 (91%) participants with bacteriologic tuberculosis and 27 (54%) of those without tuberculosis. Sensitivity and specificity of any ultrasound abnormality for bacteriologically-confirmed tuberculosis were 91% (95% confidence interval [CI] 76%-98%) and 46% (95% CI 32%-61%). Sensitivity and specificity of any ultrasound abnormality for either bacteriologically-confirmed or radiologic tuberculosis were 86% (95% CI 71%-95%) and 46% (95% CI 32%-61%). Overall performance did not appear to differ markedly between participants with and without HIV.
Conclusion: A comprehensive ultrasound scanning protocol in adults in a high TB burden setting had high sensitivity but low specificity to identify bacteriologically-confirmed tuberculosis.
Guido G, Nigussa W, Cotugno S, Kenate Sori B, Bobbio F, Gulo B Front Public Health. 2024; 12:1476866.
PMID: 39691654 PMC: 11650444. DOI: 10.3389/fpubh.2024.1476866.
Weber S, Wolf P, Wetzstein N, Kupper-Tetzel C, Vehreschild M, Suarez I Open Forum Infect Dis. 2024; 11(12):ofae651.
PMID: 39691283 PMC: 11651151. DOI: 10.1093/ofid/ofae651.