Visual assessment of sporozoite and bloodmeal ELISA samples in malaria field studies

The accuracy of visually assessing positivity for samples of field-collected Anopheles tested by enzyme-linked immunosorbent assays (ELISA) for Plasmodium falciparum sporozoites and human bloodmeals was determined during malaria field studies in Kenya. Six observers familiar with ELISAs evaluated 5,344 sporozoite ELISA samples and four observers evaluated 360 bloodmeal samples as either positive or negative based on the presence and strength of green-colored peroxidase reactions relative to controls on each microtiter plate. Interobserver agreement ranged from 97.9 to 99.8% for sporozoite samples and from 90.3 to 96.1% for bloodmeal samples. For both assays, the mean sensitivity and specificity of visual readings, compared with spectrophotometric readings, exceeded 98% when absorbance values were greater than or equal to 0.4 (on a scale of 0.0 to 2.0). Most incorrect visual readings occurred for samples with absorbance values between 0.2 and 0.4. The total percentage of samples classified correctly by visual examination ranged from 97.7 to 99.5% for the sporozoite ELISA and from 95.0 to 96.7% for the bloodmeal ELISA. Thus, there was minimal error associated with visually determining positive reactions for the ELISA assays used in malaria field studies.