Objective: To evaluate the repeatability and reproducibility of the serological direct agglutination test (DAT) for visceral leishmaniasis (VL) with aqueous antigen in a multi-centre study in VL-endemic areas in Sudan, Kenya and Nepal.
Methods: Repeatability within each centre and reproducibility between the centres' results and an external reference laboratory (Belgium) was assessed on 1596 triplicate plain blood samples collected on filter paper.
Results: High kappa values (range 0.86-0.97) indicated excellent DAT repeatability within the centres. The means of the titre differences between the reference laboratory and the centres in Sudan, Kenya and Nepal (2.3, 2.4 and 1.1, respectively, all significantly different from 0) showed weak reproducibility across centres. 95% of the titre differences between the reference laboratory and the respective centres were accounted for by large intervals: 0.6-9 fold titre variation for Sudan, 0.7-8 fold for Kenya and 0.26-4 fold for Nepal.
Conclusion: High repeatability of DAT confirms its potential, but reproducibility problems remain an obstacle to its routine use in the field. Reproducibility was hindered by alteration of the antigen through temperature and shaking, especially in Kenya and Sudan, and by nonstandardization of the test reading. DAT handling procedures and antigen quality must be carefully standardized and monitored when introducing this test into routine practice.