Visceral leishmaniasis (VL) is a serious public health problem in Bangladesh where 20 million people (18% of the total population) are at risk with a trend of rising incidence . Diagnosis of VL still relies on clinical manifestations and microscopic confirmation of parasites from aspirates of lymph nodes, bone marrow, and the spleen. These invasive and painful techniques require skilled personnel and are difficult to implement in resource-limited settings. Several less-invasive serological tests, including indirect fluorescent antibody test (IFAT), enzyme-linked immunosorbent assay (ELISA), and an improved version of direct agglutination test (DAT) have been evaluated for the diagnosis of VL [2–4]. However, a rapid immune-chromatographic test (ICT) based on a recombinant 39-amino acid repeat antigen, conserved in the kinesin region of Leishmania chagasi and Leishmania donovani (rK-39 strip test), gained popularity for the field screening of kala-azar . The detection of soluble antigen and antibody in urine of VL patients has been reported . A urine-based ELISA method has also been developed to detect anti-Leishmania donovani immunoglobulin G (IgG) .
Recently, a low molecular weight, heat-stable, and carbohydrate based leishmanial antigen has also been detected in urine of VL patients . A latex agglutination test (KAtex) based on antigen detection in urine of VL cases has been evaluated in different field studies; however, the test showed lower sensitivity in some studies [9, 10]. So, the antibody detection tests especially DAT and rK-39 strip test, are still being extensively used in the field-screening of VL.
The study was conducted to determine the potential application of the rK-39 strip test for detecting anti-leishmanial antibody in urine for the preliminary diagnosis of VL infection compared with the serum-based rK-39 test to establish the value of the urine-based rapid test for the primary diagnosis of VL.
Study area and population
In total, 100 suspected VL patients, who were positive with the serum based rK-39 strip test and had fever for at least two weeks, along with other clinical signs , were enrolled in this study from Trishal Upazila (sub-district) Health Complex (UHC) in Mymensingh district, which is one of the most endemic VL regions in Bangladesh. All the VL subjects were treated free of charge in the UHC as per the National Guideline and the WHO recommendations. To investigate cross-reaction with other diseases, twenty five (25) subjects with malaria were enrolled from a malaria-endemic area. To investigate subclinical infection, twenty five (25) healthy controls were enrolled who lived in the endemic area (Trishal) but did not have a past history of VL. Twenty five (25) healthy controls from non-endemic area were also enrolled for assessing the specificity of the urine rK-39 strip test. The serum rK-39 test was performed again in the field setting, a small laboratory in Trishal with that is near about 300 meters form UHC, whereas the urine rK-39 test was performed in the Parasitology Laboratory, ICDDR,B in Dhaka.
The Ethical Review Committee (ERC) and Research Review Committee (RRC) of ICDDR,B approved the study.
Sample collection and methods
Finger- prick blood was taken in a capillary tube and transferred to a micro-tube (200 μm). Urine samples were also collected in a tube containing preservative (Na-azide) and stored at 4°C until transporting to the ICDDR,B. The blood sample was then centrifuged for separation of serum at the field laboratory (Trishal) where the rK-39 strip test (Kalaazar Detect™, InBios Inc., USA) was also performed as per the protocol of the manufacturer. Briefly, 1 drop of serum samples was applied to the base of nitro-cellulose strips impregnated with recombinant rK-39 antigen. After being air-dried, 3 drops of the test buffer (phosphate-buffered saline, plus bovine serum albumin) were added, and the strip was placed upright. The appearance of a lower red band (control) indicated the proper functioning of the test while the appearance of an upper red band indicated the presence of anti-rK-39 IgG, signifying a positive test. For urine assay, 3 drops of urine sample were applied directly to the strip without adding any test buffer. In both the cases the strip was observed after 10 minutes for the test band. A skilled laboratory technician performed the urine rK-39 strip test in the Parasitology Laboratory ICDDR,B who also monitored the serum rK-39 strip test at the field level.
Sensitivity and specificity were computed along with 95% confidence interval (CI) using the Epi Info software (version 6.02; CDC, Atlanta, GA, USA). Data were also analyzed by 2×2 contingency tables using the SPSS software (version 10.0) for Windows (release 10.0.1, standard version, 1999; SPSS Inc., Chicago, USA), which enabled us to calculate the kappa coefficient or κ value. Reproducibility was assessed between the field rK-39 serum test and the urine rK-39 test in laboratory settings followed by Landis Koch  based on κ value.