Background: We evaluated the diagnostic accuracy as well as the reproducibility of the urine latex agglutination test 'KAtex' in the diagnosis of kala-azar in patients recruited at a tertiary care centre in Dharan, Nepal, between November 2000 and January 2002.
Methods: All patients presenting with fever of 2 weeks or more and splenomegaly were consecutively enrolled. Bone marrow and--if negative--spleen aspirates were examined for Leishmania donovani. Serum and urine samples were taken in duplicate for the Direct Agglutination Test (DAT) and KAtex. The reference laboratory determined sensitivity and specificity of KAtex. Reproducibility between both laboratories was assessed.
Results: KAtex was performed on urine from 155 parasitologically confirmed kala-azar and 77 non-kala-azar cases (parasitology and DAT-negative). KAtex showed a sensitivity of 47.7% (74/155, 95% CI: 39.7-55.9) and a specificity of 98.7% (76/77, 95% CI: 93.0-100.0). Reproducibility of KAtex showed a kappa of 0.684 (P < 0.001, n = 232).
Conclusion: KAtex evaluation showed high specificity, low sensitivity and moderate reproducibility. A urine test for kala-azar could become a real breakthrough in kala-azar management if its reproducibility and sensitivity could be further improved.