Description:
<jats:title>Abstract</jats:title><jats:p>Comparisons were undertaken to investigate cost‐effective methods of implementing the enzyme‐linked immunosorbent assay (<jats:styled-content style="fixed-case">ELISA</jats:styled-content>) for sporozoite determination in anophelines when large numbers require processing. Comparisons between <jats:styled-content style="fixed-case">ELISA</jats:styled-content> plate reader and visual assessments were performed with <jats:italic><jats:styled-content style="fixed-case">A</jats:styled-content>nopheles funestus</jats:italic> and <jats:italic><jats:styled-content style="fixed-case">A</jats:styled-content>nopheles gambiae s.l.</jats:italic> (<jats:styled-content style="fixed-case">D</jats:styled-content>iptera: <jats:styled-content style="fixed-case">C</jats:styled-content>ulicidae), as were comparisons between whole‐body mosquito samples, heads and thoraces, and abdomens alone. Rates obtained from pools of five or 10 mosquitoes were compared with those for individual mosquitoes, as were rates obtained using different sampling methods. A total of 41 792 <jats:italic><jats:styled-content style="fixed-case">A</jats:styled-content>n. funestus</jats:italic> and 9431 <jats:italic><jats:styled-content style="fixed-case">A</jats:styled-content>n. gambiae s.l.</jats:italic> collected in light traps, and 22 323 <jats:italic><jats:styled-content style="fixed-case">A</jats:styled-content>n. funestus</jats:italic> and 6860 <jats:italic><jats:styled-content style="fixed-case">A</jats:styled-content>n. gambiae s.l</jats:italic><jats:italic>.</jats:italic> from exit collections were analysed. Visual assessments gave results similar to those of machine readings. Sporozoite rates were similar in both species, as were rates by collection method. The use of whole mosquitoes increased estimates of infection rate by 0.6%. Pool size did not affect infection rates of <jats:italic><jats:styled-content style="fixed-case">A</jats:styled-content>n. gambiae s.l.</jats:italic>, but rates were higher among individually tested <jats:italic><jats:styled-content style="fixed-case">A</jats:styled-content>n. funestus</jats:italic> than among those tested in pools. For large‐scale surveys, the use of whole mosquitoes in pools of 10 mosquitoes, with correction for overestimation, and the noting of results according to a simple three‐stage visual assessment of positivity is the most cost‐effective approach and is sufficient to obtain reliable data for comparative purposes.</jats:p>