Assessment of routine malaria diagnosis in the Venezuelan Amazon.

The quality of routine malaria diagnosis is a crucial topic of malaria control. The aim of this assessment was to monitor and evaluate the quality of routine malaria diagnosis in Amazonas (Venezuela) and to improve the quality control system. The traditional non-blinded quality control system was found to be overburdened with diagnostic samples. A modified sampling system with fewer samples to be tested was proposed. Expert microscopists blindly double-checked 1000 slides and 550 rapid diagnostic tests (RDT) (OptiMAL-IT) from health posts (HP). For Plasmodium vivax, HP microscopy and OptiMAL-IT showed sensitivies of 86% and 63%, respectively. For P. falciparum, HP microscopy and OptiMAL-IT showed sensitivities of 68% and 89%, respectively. Both methods lost accuracy when fewer parasites occurred in the sample. HP microscopists from different municipalities displayed significant differences in diagnostic quality. Overall, quality of routine malaria diagnosis in the Venezuelan Amazon is good but not optimal. The change from the traditional non-blinded quality control system to blinded cross-checking of a minimal selection of samples is - comparatively - a low cost intervention with possibly high impact on the quality of routine malaria diagnosis. The introduction of RDTs should be discussed carefully in order not to displace an existing network of HP microscopists.