Subject(s)
Laboratories/standards , Mycobacterium tuberculosis/isolation & purification , Sputum/microbiology , Tuberculosis/microbiology , Argentina , Bacteriological Techniques , False Negative Reactions , False Positive Reactions , Laboratories/organization & administration , Quality Control , Tuberculosis/prevention & controlABSTRACT
IgG antibody to Mycobacterium tuberculosis antigen 5 and tuberculin purified protein derivative (PPD) was measured, by enzyme-linked immunosorbent assay (ELISA), in serum samples from 86 patients with active pulmonary tuberculosis and 91 non-tuberculous control subjects from Santa Fé, Argentina. The geometric mean titre for the tuberculosis patients was 74.6 with antigen 5 and 99.5 with PPD. In 91 control subjects the geometric mean titres were 3.6 and 15.6 respectively. Titres were not related to tuberculin reactor status or prior BCG vaccination. At a serum dilution end-point of 1:40, ELISA with antigen 5 had a sensitivity of 81.4% and a specificity of 93.4% for tuberculosis. At 1:40, ELISA with PPD showed a sensitivity of 82.6% and a specificity of 54.9% for tuberculosis. Applied at a serum dilution of 1:40 to a hypothetical model population with a tuberculosis prevalence of 2%, ELISA using antigen 5 would correctly classify 93.2% of persons and ELISA with PPD, 55.5%. At a dilution of 1:80, accuracy is increased to 99.3% with antigen 5 and 83.3% with PPD, but sensitivity decreases to 64.0% with antigen 5 and 72.1% with PPD. Thus, antigen 5 is more accurate than PPD for the diagnosis of tuberculosis using ELISA.