Abstract
Three-hundred and eight patients with suspected visceral leishmaniasis (VL) were received at Doka Hospital (eastern Sudan) during the period September 2004 to October 2005. The sensitivity and specificity of a glycerol-preserved (GP) antigen for VL diagnosis was assessed against the results of repeated lymph node aspiration and readings from a direct agglutination test (DAT) employing standard formaldehyde-fixed (FF) or freeze-dried (FD) antigen. Despite 13 months of storage at ambient temperature (28-47 degrees C), the GP antigen mean titres obtained from these 308 patients were no different from those that were FD (P=0.945) and stored under similar conditions, but were significantly different (P=0.019) from those that were FF and kept continuously at the optimum temperature for storage (4-8 degrees C). Taking the parasitological result as the gold standard and using a pre-established titre of 1 : 3200 as the DAT cut-off, the GP antigen revealed a sensitivity (91/105, 86.7 %) and specificity (187/203, 92.1 %) comparable to that of FD antigen (92/105, 87.6 %, and 188/203, 92.6 %, respectively) and FF antigen (94/105, 89.5 %, and 188/203, 92.6 %, respectively). At a titre range of 1 : 400-1 : 800, statistically determined as the optimum cut-off for the three antigens, sensitivities of 92.4, 90.5 and 96.2 % and specificities of 90.6, 90.1 and 88.7 % were achieved for the GP, FD and FF antigens, respectively, at a peripheral hospital. Regardless of the antigen preparation used, DAT results obtained in the peripheral hospital were highly reproducible in the central laboratory in Omdurman (weighted kappa: GP=0.957, FD=0.979 and FF=0.936). With a diagnostic reliability comparable to formaldehyde fixation and stability under ambient conditions similar to freeze drying, glycerol preservation, by virtue of its high potential for reproduction, meets the requirements for the management of VL in developing countries.