Background
A universal coverage campaign (UCC) with long-lasting insecticidal nets (LLINs) was implemented in four districts in Midwestern Uganda in 2009–2010. Entomological surveys were carried out to monitor changes in vector density, behaviour and malaria transmission following this intervention.

Methods
Anopheles mosquitoes were collected using CDC light traps quarterly and human landing catch twice a year in four sites. Collections were done at baseline before the campaign and over a three-year period following the campaign. Plasmodium falciparum circumsporozoite enzyme-linked immunosorbent assays were performed. A subset of anophelines were molecularly identified to species, and kdr L1014S frequencies were determined.

Results
The prevailing malaria vector in three sites was Anopheles gambiae s.l. (>97 %), with An. funestus s.l. being present in low numbers only. An. gambiae s.s. dominated (> 95 %) over An. arabiensis within A. gambiae s.l. In the remaining site, all three vector species were observed, although their relative densities varied among seasons and years. Vector densities were low in the year following the UCC but increased over time. Vector infectivity was 3.2 % at baseline and 1.8 % three years post-distribution (p = 0.001). The daily entomological inoculation rate (EIR) in 2012 varied between 0.0-0.98 for the different sites compared to a baseline EIR that was between 0.0-5.8 in 2009. There was no indication of a change in indoor feeding times, and both An. gambiae s.l. and An. funestus s.l. continued to feed primarily after midnight with vectors being active until the early morning. Kdr L1014S frequencies were already high at baseline (53–85 %) but increased significantly in all sites over time.

Conclusions
The entomological surveys indicate that there was a reduction in transmission intensity coinciding with an increase in use of LLINs and other antimalarial interventions in areas of high malaria transmission. There was no change in feeding behaviour, and human-vector contact occurred indoors and primarily after midnight constantly throughout the study. Although the study was not designed to evaluate the effectiveness of the intervention compared to areas with no such intervention, the reduction in transmission occurred in an area with previously stable malaria, which seems to indicate a substantial contribution of the increased LLIN coverage.