Diagnostic performance characteristics of a rapid field test for anthrax in cattle

Although diagnosis of anthrax can be made in the field with a peripheral blood smear, and in the laboratory with bacterial culture or molecular based tests, these tests require either considerable experience or specialised equipment. Here we report on the evaluation of the diagnostic sensitivity and specificity of a simple and rapid in-field diagnostic test for anthrax, the anthrax immunochromatographic test (AICT). The AICT detects the protective antigen (PA) component of the anthrax toxin present within the blood of an animal that has died from anthrax. The test provides a result in 15min and offers the advantage of avoiding the necessity for on-site necropsy and subsequent occupational risks and environmental contamination. The specificity of the test was determined by testing samples taken from 622 animals, not infected with Bacillus anthracis. Diagnostic sensitivity was estimated on samples taken from 58 animals, naturally infected with B. anthracis collected over a 10-year period. All samples used to estimate the diagnostic sensitivity and specificity of the AICT were also tested using the gold standard of bacterial culture. The diagnostic specificity of the test was estimated to be 100% (99.4-100%; 95% CI) and the diagnostic sensitivity was estimated to be 93.1% (83.3-98.1%; 95% CI) (Clopper-Pearson method). Four samples produced false negative AICT results. These were among 9 samples, all of which tested positive for B. anthracis by culture, where there was a time delay between collection and testing of >48h and/or the samples were collected from animals that were >48h post-mortem. A statistically significant difference (P<0.001; Fishers exact test) was found between the ability of the AICT to detect PA in samples from culture positive animals <48h post-mortem, 49 of 49, Se=100% (92.8-100%; 95% CI) compared with samples tested >48h post-mortem 5 of 9 Se=56% (21-86.3%; 95% CI) (Clopper-Pearson method). Based upon these results a post hoc cut-off for use of the AICT of 48h post-mortem was applied, Se=100% (92.8-100%; 95% CI) and Sp=100% (99.4-100%; 95% CI). The high diagnostic sensitivity and specificity and the simplicity of the AICT enables it to be used for active surveillance in areas with a history of anthrax, or used as a preliminary tool in investigating sudden, unexplained death in cattle.