applied strategies to one in which local control personnel decide for themselves what is the most operationally and cost-effective strategy for their particular situation. The new approach will need to be designed to account for variation in dengue transmission at different geographic locations and at different times at the same place. Local control personnel will need to constantly evaluate their surveillance and response methods. Their goals will have to be spatially and temporally specific, accounting for local variation in ecology, epidemiology, and availability of intervention resources.
An example of this would be use of pupal productivity analysis to target vector control at containers producing most of the adult Ae. aegypti. In some places most Ae. aegypti production is associated with water storage, and those containers are easily identified and treated with larvicides. In contrast, at other locations most production comes from unmanaged containers that are transient and often missed in routine entomological inspections. Control campaigns for these two extremes would be noticeably different. In Iquitos during a severe 2002 DV-3 epidemic, local health officials deemphasized an entrenched pattern of uniform larvicide applications in preference of enhanced public awareness and container clean-up. The change was motivated by solid entomological surveillance data, which indicated that adult Ae. aegypti were being produced primarily from unmanaged containers rather than water storage containers.
The shift from prescribed to adaptable strategies will require application of translational research, basic and applied, to the development of novel products and strategies that reduce disease. For example, dynamic, operational tools like virus transmission models and decision support systems will be necessary to guide site- and situation-specific dengue control. For a meaningful conversion of research to improved public health, it is imperative that those responsible for preventing DV transmission use surveillance information to inform their control decisions.
For more than half a century dengue prevention programs focused on immature Ae. aegypti for surveillance and control (PAHO, 1994). There are theoretical and empirical reasons for no longer strictly following that approach. With regard to surveillance, immature indices of Ae. aegypti density have not proven to be good predictors of DV transmission risk. Moreover, goals for immature Ae. aegypti surveillance are often vague and do not account for temporal and spatial variation in transmission factors. With regard to control, killing larvae is expected to have a relatively small impact on a reduction in the number of new human dengue infections, compared to killing adults.
Refocusing dengue surveillance and control on adult Ae. aegypti would be a significant step forward. One of the major road blocks to improved dengue surveillance is our inability to directly monitor the vector life form that transmits