• Malaria;
  • sporozoite rate;
  • ELISA absorbance cut-off;
  • Anopheles gambiae;
  • Anopheles funestus;
  • Kenya

ABSTRACT. Malaria sporozoite infection rates in a mixed species group of 244 Anopheles gambiae Giles sensu lato and 115 An. funestus Giles wild female mosquitoes were compared using three methods to determine cutoff absorbance values for positivity of a Plasmodium falciparum Welch enzyme-linked immunosorbent assay (ELISA). Positive controls were based on P. falciparum circumsporozoite protein. As negative controls, four wild male Anopheles were included on each microtitre plate; tests were repeated on four consecutive days for each plate.

Infection rates were estimated at 13.1–22.8% using the mean absorbance value of negative controls plus three standard deviations, 11.7–12.8% using double the mean and 12.5–13.6% using the fixed cut-off value of 0.20 (allowing for 20% variation in negative control absorbance values).

Observed agreement for positivity or negativity among samples tested four times was 98.6% for the 2× mean method, 97.2% for the fixed cut-off 0.20 value, but only 82.7% for the mean +3 SD method. It was concluded that the 2× mean cut-off method is most reliable for field studies. P. falciparum sporozoite rates of 12.2% in An. funestus and 11.9% in An. gambiae s. l. were thus determined on the basis of the 2× mean cut-off method.

This comparative evaluation demonstrates that vector infectivity rates can be seriously over-estimated from sporozoite ELISA tests, by as much as 87% in one case considered here, depending on the absorbance cut-off method applied for negative controls.