Egg density as a guide for Aedes control: a controlled field comparison of four intervention strategies in a Brazilian city

Abstract Brazilian municipalities control Aedes aegypti with a routine of bimonthly vis-its to every premise, and when counts stay high despite that routine, agents have little evidence on what to add. We compared four additional actions, each directed by ovitrap counts, in Patos de Minas, Minas Gerais. Five of the city’s eleven operational zones were followed fortnightly for two years, all on the rou-tine programme throughout. In the second year four zones also received one additional action within 150 m of traps holding more than 50 eggs: extra vis-its for container elimination with focal Bacillus thuringiensis israelensis (Bti), the same plus ultra-low-volume (ULV) spraying, clean-up campaigns, or ULV alone. The fifth kept the routine only. We examined 4,670 ovitraps and counted 165,199 Aedes spp. eggs. In a difference-in-differences negative binomial model with autocorrelation-consistent errors, no action differed significantly from the routine alone. The largest reduction followed extra elimination with Bti (rate ratio 0.53, 95% CI 0.25–1.13) and partly predated the change; ULV alone had the least favourable estimate (1.12, 0.49–2.53). Rainfall accounted for more deviance than all four actions together. The population was resistant to all four pyrethroids tested. With one zone per action these results are exploratory, but they point to larval work rather than spraying as the addition to test first.