Three novel diagnostic tests for visceral leishmaniasis (VL), namely FD-DAT, rK39 dipstick and KATEX, were evaluated under field conditions using 101 clinical cases suspected of having VL enrolled in a trial either by active (63 patients) or passive (38 patients) surveillance. VL was confirmed in 49 patients: 35 by both aspirate smear microscopy and NNN culture, 10 by NNN culture alone and 4 by aspirate smear microscopy alone. Based on tests performed in the field, sensitivity for FD-DAT, rK39 dipstick and KATEX was determined to be 95.3% (95% CI 82.9-99.2%), 71.7% (95% CI 56.3-83.5%) and 57.4% (95% CI 42.3-71.4%), respectively. Similarly, the specificity was determined to be 62.7% (95% CI 48.1-75.5%), 82.4% (95% CI 68.6-91.1%) and 84.3% (95% CI 70.9-92.5%), respectively. A higher sensitivity of KATEX (73.9% vs. 41.7%) and higher specificity of FD-DAT (100.0% vs. 48.6%) were demonstrated under passive case detection compared with active case detection. FD-DAT is recommended for confirmation of VL diagnosis in hospital settings, whereas its use in the field will be limited to exclude VL in clinical suspects. The sensitivity of KATEX and rK39 dipstick tests needs to be improved to promote their use as first-line diagnostic tests in the field setting of northwestern Ethiopia.

A fast agglutination screening test (FAST) for the detection of Leishmania antibodies in human serum samples was evaluated under harsh field conditions in northern Ethiopia. Test performance was compared with a standard serological test, namely the direct agglutination test (DAT), and with parasitology. In total, 103 suspected cases were recruited for the study. Based on parasitological examination, 49 patients were confirmed of having visceral leishmaniasis (VL) and the other 54 suspected cases were parasitologically negative. Field evaluation of FAST was possible in blood samples of 89 patients. FAST had 4 false negative results and 13 false positive results. DAT had 2 false negative results and 20 false positive results. A good degree of agreement (86.9%) was observed between FAST and DAT (kappa value 0.73). In this field-based evalauation, the sensitivity and specificity of FAST were found to be 91.1% (95% CI 77.9-97.1) and 70.5% (95% CI 54.6-82.8), respectively, compared with 95.3% (95% CI 82.9-99.2) and 62.3% (95% CI 47.9-74.9) for DAT. FAST had a high predictive value of a negative test, demonstrating that FAST could be utilised to exclude rapidly non-VL patients from a large population of suspects with fever and splenomegaly in endemic areas.