Table 3 Selected characteristics of studies investigating commercial tests for the diagnosis of extrapulmonary tuberculosis.
Reference | Data collection | Verification | Reference standard | Disease site | Country | Comparison group | Name of test* | No. of participants‡ | Sensitivity (95% CI) | Specificity (95% CI) |
---|---|---|---|---|---|---|---|---|---|---|
Banerjee (2003, b)42 | Retrospective | NR | Histology | Lymph node | India | Healthy | SEVA TB | 30/32 | 0.77 (0.58 to 0.90) | 0.88 (0.71 to 0.97) |
McConkey (2002)48† | Prospective | Complete | Culture | Meninges | Egypt | Mixed disease | AMRAD ICT | 56/74 | 0.48 (0.35 to 0.62) | 0.82 (0.72 to 0.90) |
Nsanze (1997, a)49 | Prospective | Differential | Culture and/ or smear | NR | UAE | Mixed disease and healthy | Pathozyme Myco | 35/35 | 0.51 (0.34 to 0.69) | 1.00 (0.90 to 1.00) |
Nsanze (1997, b)49 | Prospective | Differential | Culture and/ or smear | NR | UAE | Mixed disease and healthy | Pathozyme TB Complex Plus | 35/35 | 0.11 (0.03 to 0.27) | 1.00 (0.90 to 1.00) |
NR, not reported.
*All commercial tests detected immunoglobulin G antibody.
†Blinded study.
‡Number of participants with tuberculosis/without tuberculosis.