Field evaluation of a polymerase chain reaction-based nonisotopic liquid hybridization assay for malaria diagnosis.

In a blind field evaluation of a nonisotopic liquid hybridization assay for detection of malaria parasites, 100 blood samples were tested from an area in which malaria is endemic; light microscopy was used as the reference test. Sensitivity, specificity, and positive and negative predictive values of the hybridization assay were 100%. One sample that was microscopy-negative and hybridization-positive was positive when reexamined. Another sample that was microscopy-positive and hybridization-negative was negative at reexamination. The detection limit of the test was > or = 0.0005% parasitemia. Four samples with mixed infections were misdiagnosed by microscopy as single-species infections. Four samples diagnosed as mixed infections by microscopy and single infection by the hybridization test had no evidence of a second Plasmodium species upon reexamination. The polymerase chain-reaction-based nonisotopic liquid hybridization assay was better than conventional light microscopy in detecting low-grade parasite infection and offers an exceptional advantage for detecting mixed infections.